<%BANNER%>

Dose Assessment And Prediction In Tube-Current Modulated Computed Tomography

Permanent Link: http://ufdc.ufl.edu/UFE0042534/00001

Material Information

Title: Dose Assessment And Prediction In Tube-Current Modulated Computed Tomography
Physical Description: 1 online resource (168 p.)
Language: english
Creator: Fisher, Ryan
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: computed, ct, current, dose, modulation, organ, tomography, tube
Nuclear and Radiological Engineering -- Dissertations, Academic -- UF
Genre: Nuclear Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: DOSE ASSESSMENT AND PREDICTION IN TUBE-CURRENT MODULATED COMPUTED TOMOGRAPHY Computed Tomography (CT) technology has advanced rapidly since its inception allowing clinicians to obtain large amounts of useful anatomical data in very short amounts of time. These advances have led to an explosion in the use and utility of CT for a wide variety of medical applications. As the use of CT has expanded concerns over patient exposure to radiation from CT scans has risen. In order to reduce patient doses and maintain better image quality between exams and within slices in the same exam, manufacturers have developed tube-current modulation systems that change the output of the x-ray tube as it travels around the patient in response to changes in patient anatomy and attenuation. Current methods used to estimate patient organ doses from CT procedures are based on outdated technology and do not incorporate this newer tube-current modulation technology. Further, the methods of dose estimation do not take patient size into account, which can have a marked impact on dose, especially in tube-current modulated scans that adjust output in response to patient anatomy. For a given imaging technique using tube-current modulation, two different sized patients can receive drastically different doses. In order to better estimate patient organ doses in modulated CT procedures an existing, custom fiber-optic dosimetry system and anthropomorphic phantom representing a 50th percent by height and weight male were used to collect organ doses from a variety of modulated CT scans. Additionally, an adipose tissue equivalent substitute material was developed and used to build a phantom add-on to make the 50th percentile phantom a 90th percentile by weight male. Doses were measured in both phantoms on a Siemens and a Toshiba CT scanner, each with a different tube-current modulation scheme. The collected library of organ doses was compared to dose estimates from several current methods, both of which were deemed inadequate in predicting modulated doses. Using the collected data as an input, a CT dose estimate spreadsheet was created that takes patient size and scanner parameters into account in order to more accurately estimate specific organ doses from tube-current modulated CT exams.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ryan Fisher.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Hintenlang, David E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042534:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042534/00001

Material Information

Title: Dose Assessment And Prediction In Tube-Current Modulated Computed Tomography
Physical Description: 1 online resource (168 p.)
Language: english
Creator: Fisher, Ryan
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: computed, ct, current, dose, modulation, organ, tomography, tube
Nuclear and Radiological Engineering -- Dissertations, Academic -- UF
Genre: Nuclear Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: DOSE ASSESSMENT AND PREDICTION IN TUBE-CURRENT MODULATED COMPUTED TOMOGRAPHY Computed Tomography (CT) technology has advanced rapidly since its inception allowing clinicians to obtain large amounts of useful anatomical data in very short amounts of time. These advances have led to an explosion in the use and utility of CT for a wide variety of medical applications. As the use of CT has expanded concerns over patient exposure to radiation from CT scans has risen. In order to reduce patient doses and maintain better image quality between exams and within slices in the same exam, manufacturers have developed tube-current modulation systems that change the output of the x-ray tube as it travels around the patient in response to changes in patient anatomy and attenuation. Current methods used to estimate patient organ doses from CT procedures are based on outdated technology and do not incorporate this newer tube-current modulation technology. Further, the methods of dose estimation do not take patient size into account, which can have a marked impact on dose, especially in tube-current modulated scans that adjust output in response to patient anatomy. For a given imaging technique using tube-current modulation, two different sized patients can receive drastically different doses. In order to better estimate patient organ doses in modulated CT procedures an existing, custom fiber-optic dosimetry system and anthropomorphic phantom representing a 50th percent by height and weight male were used to collect organ doses from a variety of modulated CT scans. Additionally, an adipose tissue equivalent substitute material was developed and used to build a phantom add-on to make the 50th percentile phantom a 90th percentile by weight male. Doses were measured in both phantoms on a Siemens and a Toshiba CT scanner, each with a different tube-current modulation scheme. The collected library of organ doses was compared to dose estimates from several current methods, both of which were deemed inadequate in predicting modulated doses. Using the collected data as an input, a CT dose estimate spreadsheet was created that takes patient size and scanner parameters into account in order to more accurately estimate specific organ doses from tube-current modulated CT exams.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ryan Fisher.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Hintenlang, David E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042534:00001


This item has the following downloads:


Full Text

PAGE 1

1 DOSE AS SESSMENT AND PREDICTION IN TUBE CURRENT MODULATED COMPUTED TOMOGRAPHY By RY AN F. FISHER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR TH E DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Ryan F. Fisher

PAGE 3

3 To Also, to my wife and parents

PAGE 4

4 ACKNOWLEDGMENTS I w ould first like to thank my research advisor and chair of my committee, David Hintenlang for his guidance during the course of my graduate work at the University of Florida. His guidance, instruction, sense of humor, and friendship has helped shape my fu ture in the field and made my time at UF more enjoyable. I would also like to thank the other members of my committee, Manuel Arreola, Wesley Bolch, and Hans van Oostrom for their input, advice, and guidance through the course of this work. Additionally, I would like to thank the staff of the Nuclear and Radiological Engineering Department, as well as the technologists at Shands Hospital for their help through my studies as well. Additionally, I would like to thank my original classmates at Florida: Bob Am brose, Lindsay Lavoie, Bill Moloney, Chris Koenings, and Keelan Seabolt, for their friendship and easing the transition to a new town/school /field of study I would also like to thank Dan Hyer, James Winslow and Perry Johnson for their friendship and assis tance throughout my resear ch at the University of Florida, as well as anyone who I may have raced or ridden with on the cycling team at UF. All of you have contributed to making my time here memorable and enjoyable. Finally, I would like to thank my parent s for all of their support through the years (even though they still refuse to root for the gators) and for instilling the curiosity and work ethic in me required to stay in school as long as I have. I also want to thank my loving wife Kelly for her love and sacrifices through the years, including the years spent apart in grad school and her moving to a small college town and providing for me through my graduate studies.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPT ER 1 INTRODUCTION AND BACKGROUND ................................ ................................ 16 Introduction ................................ ................................ ................................ ............. 16 Background and Significance ................................ ................................ ................. 19 Automatic Tube current Modulation ................................ ................................ 20 Clinical Tube current Modulation Studies ................................ ......................... 25 Monte Carlo B ased CT Studies ................................ ................................ ........ 28 Current Standards of Dose Reporting in Clinical CT ................................ ........ 32 Objectives of this Research ................................ ................................ .................... 35 2 FIBER OPTIC DOSIMETRY SYSTEM & PHANTOM CONSTRUCTION ............... 46 Fiber Optic Coupled Dosimetry System ................................ ................................ .. 46 Fiber optic System Development ................................ ................................ ..... 46 FOCD System Calibration ................................ ................................ ................ 47 Fiftieth Percentile Phantom Construction ................................ ................................ 48 Computational Phantom and Mold Fabrication ................................ ................. 49 Tissue Equivalent Substitute Development ................................ ...................... 50 Organ Point Dose L ocations ................................ ................................ ............. 52 Ninetieth Percentile Phantom Construction ................................ ............................ 53 Adipose Tissue Equivalent Substitute Material Development ........................... 54 Phantom Construction ................................ ................................ ...................... 54 3 CREATION OF CT ORGAN DOSE LIBRARY ................................ ........................ 72 MDCT Exams ................................ ................................ ................................ ......... 72 Fixed Tube Current Scans ................................ ................................ ...................... 73 Organ Dose Measurements ................................ ................................ .................... 76 CTDI VOL and ImPACT Dose Estima tes ................................ ................................ ... 79 4 RESULTS ................................ ................................ ................................ ............... 87 Siemens Tube Current Modulated Scans ................................ ............................... 87 Chest Exam ................................ ................................ ................................ ...... 87

PAGE 6

6 Modulation plot ................................ ................................ ........................... 87 Organ doses ................................ ................................ .............................. 88 Abdominal Exam ................................ ................................ .............................. 89 Modulation plot ................................ ................................ ........................... 89 Organ doses ................................ ................................ .............................. 90 Pelvic Exam ................................ ................................ ................................ ...... 91 Modulation plot ................................ ................................ ........................... 91 Organ doses ................................ ................................ .............................. 91 Toshiba Tube Current Modulated Scans ................................ ................................ 92 Chest Exam ................................ ................................ ................................ ...... 92 Modulation plot ................................ ................................ ........................... 92 Organ doses ................................ ................................ .............................. 93 Abdomin al Exam ................................ ................................ .............................. 94 Modulation plot ................................ ................................ ........................... 94 Organ doses ................................ ................................ .............................. 94 Pelvic Exam ................................ ................................ ................................ ...... 95 Modulation plot ................................ ................................ ........................... 95 Organ doses ................................ ................................ .............................. 95 Scanner Comparison for Tube Current Modulated Scans ................................ ...... 97 Tube current Response ................................ ................................ .................... 97 Image Quality Comparison ................................ ................................ ............... 98 Comparison to CTD I VOL and ImPACT Dose Estimates ................................ ......... 100 Chest Exam ................................ ................................ ................................ .... 101 Abdominal Exam ................................ ................................ ............................ 104 Pelv ic Exam ................................ ................................ ................................ .... 106 Conclusions of Comparison Study ................................ ................................ 107 Effects of Beam Width on Modulated CT Scans ................................ ................... 108 Chest Exam ................................ ................................ ................................ .... 108 Abdominal Exam ................................ ................................ ............................ 109 Pelvic Exam ................................ ................................ ................................ .... 109 Surface Beam Profile Plots ................................ ................................ ............. 110 Surface Variability ................................ ................................ .......................... 111 Z axis Over ranging ................................ ................................ ........................ 113 Fixed Tube Current Dose Measurement Results ................................ .................. 113 Chest exam ................................ ................................ .............................. 115 Abdominal exam ................................ ................................ ...................... 116 Pelvic exam ................................ ................................ .............................. 117 5 CREATION OF PATIENT SIZE DEPENDANT CT DOSE ESTIMATION CALCULATOR ................................ ................................ ................................ ...... 149 Need For Patient Size Dependent Dose Estimation ................................ ............. 149 Methodology ................................ ................................ ................................ ......... 149 Limitations and Validation of Method ................................ ................................ .... 150 CT Dose Estimation Calculator ................................ ................................ ............. 153 Future Work to Improve Accuracy and Robustness of Method ............................. 153

PAGE 7

7 6 CONCLUSION ................................ ................................ ................................ ...... 158 Results of This Work ................................ ................................ ............................. 158 Potential Future Work ................................ ................................ ........................... 159 Final Thoughts ................................ ................................ ................................ ...... 160 APPENDIX: INDIVIDUAL POINT DOSE MEASUREMENTS ................................ ...... 161 LIST OF REFERENCES ................................ ................................ ............................. 164 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 168

PAGE 8

8 LIST OF TABLES Table page 2 1 Organs of interest for each exam ................................ ................................ ....... 62 3 1 Chest exam techniq ues ................................ ................................ ...................... 81 3 2 Abdominal exam techniques ................................ ................................ ............... 82 3 3 Pelvic exam techniques ................................ ................................ ...................... 83 3 4 HU standard deviations in the image quality phantom. ................................ ....... 83 3 5 Fixed tube current mAs values utilized to scale organ dose measurements ...... 84 3 6 Organs and their corresponding number of dose points for a chest exam ......... 85 3 7 Organs and their corresponding number of dose points for an abdominal exam ................................ ................................ ................................ ................... 86 3 8 Organs and their corresponding number of dose points for a pelvic exam ......... 86 4 1 Scanner image quality comparison ................................ ................................ ... 134 4 2 Mass fraction of red bone marrow used in whole body marrow calculations .... 134 A 1 Individual point dose measurements for chest exams ................................ ...... 161 A 2 Individual point doses for abdominal exams ................................ ..................... 162 A 3 Individual point doses for pelvic exams ................................ ............................ 163

PAGE 9

9 LIST OF FIGURES Figure page 1 1 CTDI head and body phantoms. ................................ ................................ ......... 36 1 2 Screenshot of ImpaCT dosimetry calculator ................................ ....................... 37 1 3 Increased image noise as a result of low tube current in the larger phantom ..... 38 1 4 Anisotropic noise as a result of photon starvation in the direction of highest attenuation in an ellip tical phantom. ................................ ................................ ... 38 1 5 Three types of automatic tube current modulation ................................ ............. 39 1 6 Angular Tube Modulation ................................ ................................ ................... 40 1 7 Angular Tube Modulation: Tube current is adjusted in real time to match patient attenuation ................................ ................................ .............................. 40 1 8 Z axis modulation ................................ ................................ ............................... 41 1 9 Z axis modulation ................................ ................................ ............................... 41 1 10 Kidney phantom ................................ ................................ ................................ .. 42 1 11 P atient organ doses versus patient perimeter from fixed tub e current CT scans ................................ ................................ ................................ .................. 43 1 12 P atient organ doses versus patient perimeter from tube current modulated CT scans. ................................ ................................ ................................ ........... 44 1 13 Scenario illustr ating two different sized patients receiving a CT scan with the same techniques ................................ ................................ ................................ 45 1 14 Scenario illustrating two different sized patients receiveing a CT scan with size adjusted techniques ................................ ................................ .................... 45 2 1 FOCD system schematic ................................ ................................ .................... 58 2 2 Complete fiber optic dosimeter ................................ ................................ ........... 59 2 3 Step s in phantom construction process ................................ .............................. 60 2 4 Phantom slice including ( a) STES, (b) LTES), and (c) BTE ................................ 61 2 5 Completed 50 th percentile ant hropomorphic physical phantom .......................... 62 2 6 Scanogram image of completed phantom. ................................ ......................... 63

PAGE 10

10 2 7 HU value of various samples of ATES material ................................ .................. 64 2 8 50 th and 90 th percentile, by weight, hybrid computational phantoms .................. 65 2 9 Completed 90 th percentile phantom mold ................................ ........................... 66 2 10 50 th percentile phantom section wrapped in plastic and centered in a mold. ...... 67 2 11 90th percentile phantom mold filled with ATES material. ................................ .... 68 2 12 Multiple phantom sections curing in molds. ................................ ........................ 69 2 13 Completed 90 th percentile phantom add on around a section of 50 th percentile phanto m ................................ ................................ ............................. 70 2 14 Completed 90 th percentile phantom. ................................ ................................ ... 71 3 1 Survey results of 25,000 CT scans at Shands Hospital at the University of Florida. ................................ ................................ ................................ ............... 80 3 2 Scan range for chest exam ................................ ................................ ................. 81 3 3 Scan range for abdominal exam ................................ ................................ ......... 82 3 4 Scan range for pelvic exam ................................ ................................ ................ 83 3 5 Dummy slice created for the assessment of image quality. ................................ 84 3 6 Catphan image quality module ................................ ................................ ........... 84 3 7 Slice images from image quality scan for 50 th (left) and 90 th (right) percentile phantoms ................................ ................................ ................................ ............ 85 4 1 Plot of tube current response fo r chest exam ................................ ................... 119 4 2 Average organ doses for Siemens chest exam ................................ ................ 119 4 3 Dose increases in larger phantom for average organ doses in th e Siemens chest exam ................................ ................................ ................................ ....... 120 4 4 Plot of tube current response in abdominal exam ................................ ............ 120 4 5 Average organ doses for Siemens abdominal exam ................................ ........ 121 4 6 Average in field organ doses for Siemens abdominal exam ............................. 121 4 7 Dose increases in larger phantom for in field average organ dose s in the abdominal exam ................................ ................................ ............................... 122

PAGE 11

11 4 8 Plot of tube current response in pelvic exam ................................ .................... 122 4 9 Average organ doses for Siemens pelvic exam ................................ ................ 123 4 10 Average in field organ doses for Siemens pelvic exam ................................ .... 123 4 11 Dose increases in larger phantom for in field average organ doses in the pelvic exam ................................ ................................ ................................ ....... 124 4 12 Plot of tube current response in Toshiba chest exam ................................ ....... 124 4 13 Average organ doses for Toshiba chest exam ................................ ................. 125 4 14 Average in field organ doses for Toshiba chest exam ................................ ...... 125 4 15 Dose increases in larger phantom for in field average organ doses in the Toshiba chest exam ................................ ................................ ......................... 126 4 16 Plot of tube current response in Toshiba abdominal exam ............................... 126 4 17 Average organ doses for Toshiba abdo minal exam ................................ ......... 127 4 18 Average in field organ doses for Toshiba abdominal exam .............................. 127 4 19 Dose increases in larger phantom for in field ave rage organ doses in the Toshiba abdominal exam ................................ ................................ ................. 128 4 20 Plot of tube current response in Toshiba pelvic exam ................................ ...... 128 4 21 Average organ do ses for Toshiba pelvic exam ................................ ................. 129 4 22 Average in field organ doses for Toshiba pelvic exam ................................ ..... 129 4 23 Dose increases in larger phantom fo r in field average organ doses in the Toshiba pelvic exam ................................ ................................ ......................... 130 4 24 Comparison of tube current response in chest exam ................................ ....... 130 4 25 Average in field organ doses for both Toshiba and Siemens chest exams ...... 131 4 26 Comparison of tube current response in abdominal exam ............................... 132 4 27 Average in field organ doses for both Toshiba and Siemens abdominal exams ................................ ................................ ................................ ............... 132 4 28 Comparison of tube current response in pelvic exam ................................ ....... 133 4 29 Average in field organ doses for both Toshiba and Siemens pelvic exams ...... 133

PAGE 12

12 4 30 Comparison of measured and predicted chest exam doses in 50 th percentile phantom ................................ ................................ ................................ ........... 135 4 31 Comparison of measured and predicted chest exam doses in 90 th percentile phantom ................................ ................................ ................................ ........... 136 4 32 Comparison of measured and predicted abdominal exam doses in 50 th percentile phantom ................................ ................................ ........................... 137 4 33 Comparison of measured and predicted abdominal exam doses in 90 th percentile phantom ................................ ................................ ........................... 138 4 34 Compar ison of measured and predicted pelvic exam doses in 50 th percentile phantom ................................ ................................ ................................ ........... 139 4 35 Comparison of measured and predicted pelvic exam doses in 90 th percentile phantom ................................ ................................ ................................ ........... 140 4 36 Effects of beam width on organ doses in chest exam ................................ ....... 141 4 37 Effects of beam width on organ doses in abdominal exam ............................... 141 4 38 Effects of beam width on organ doses in pelvic exam ................................ ...... 142 4 39 Plot of instantaneous dose rate versus time for various beam widths .............. 142 4 40 Dose rates for various beam widths scaled according to scan length .............. 143 4 41 Monte Carlo simulation of the average in plane dose distribution for an ATCM scan o f an anthropomorphic phantom ................................ ................... 144 4 42 Point dose variability due to beam start angle ................................ .................. 145 4 43 Effects of z axis over ranging on out of field point doses ................................ 145 4 44 Tube current modulation plot with fixed current lines for chest exam ............... 146 4 45 Comparison of fixed and modu lated doses in the Siemens chest exam ........... 146 4 46 Plot of tube current response and fixed tube currents in Siemens abdominal exam ................................ ................................ ................................ ................. 147 4 4 7 Doses from fixed and modulated abdominal scan ................................ ............ 147 4 48 Tube current plot of ATCM and FTC scans in a Siemens pelvic exam ............ 148 4 49 Com parison of pelvic organ doses in FTC and ATCM scans. .......................... 148

PAGE 13

13 5 1 Scatter plot of lung and breast doses for 32 women undergoing ATCM chest exams ................................ ................................ ................................ ............... 155 5 2 Reproduction of Figure 5 1 with dose points from this work added .................. 156 5 3 Screenshot of the created CT dose calculator ................................ .................. 157

PAGE 14

14 Abstract of Dissertatio n Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DOSE AS SESSMENT AND PREDICTION IN TUBE CURRENT MODULATED COMPUTED TOMOGRAPHY By Ryan F. Fish er December 2010 Chair: David E. Hintenlang Major: Nuclear and Radiological Engineering Computed Tomography (CT) technology has advanced rapidly since its inception allowing clinicians to obtain large amounts of useful anatomical data in very short am ounts of time. These advances have led to an explosion in the use and utility of CT for a wide variety of medical applications. As the use of CT has expanded concerns over patient exposure to radiation from CT scans has risen. In order to reduce patient do ses and maintain better image quality between exams and within slices in the same exam, manufacturers have developed tube current modulation systems that change the output of the x ray tube as it travels around the patient in response to changes in patient anatomy and attenuation. Current methods used to estimate patient organ doses from CT procedures are based on outdated technology and do not incorporate this newer tube current modulation technology. Further, the methods of dose estimation do not take pat ient size into account, which can have a marked impact on dose, especially in tube current modulated scans that adjust output in response to patient anatomy. For a given imaging technique using tube current modulation, two different sized patients can rece ive drastically different doses.

PAGE 15

15 In order to better estimate patient organ doses in modulated CT procedures an existing, custom fiber optic dosimetry system and anthropomorphic phantom representing a 50 th percent by height and weight male were used to co llect organ doses from a variety of modulated CT scans. Additionally, an adipose tissue equivalent substitute material was developed and used to build a phantom add on to make the 50 th percentile phantom a 90 th percentile by weight male. Doses were measure d in both phantoms on a Siemens and a Toshiba CT scanner, each with a different tube current modulation scheme. The collected library of organ doses was compared to dose estimates from several current methods, both of which were deemed inadequate in predic ting modulated doses. Using the collected data as an input, a CT dose estimate spreadsheet was created that takes patient size and scanner parameters into account in order to more accurately estimate specific organ doses from tube current modulated CT exam s.

PAGE 16

16 C HAPTER 1 INTRODUCTION AND BAC KGROUND Introduction Multi detector Computed Tomography (MDCT) technology has advanced rapidly have as many as 320 detector rows and can complete a full revolution in 300 milliseconds, allowing a large amount of patient anatomical information to be gathered in a short period of time. These technological advancements have lead to an increase in the clinical use of CT for application s such as angiography and multiphase cardiac and brain imaging. 1, 2 While the expanded clinical use of CT has numerous benefits, including faster patient throughput and fewer invasive patient procedures, it has als o sparked scrutiny into the levels of radiation dose these exams produce. In particular, recent studies by the New England Journal of Medicine 3 the American College of Radiology 4 and the National Council on Radiation Protection (NRCP) 5 have highlighted the increases in medical radiation exposures to the public. In a n effort to reduce CT patient dose in individual exams, automatic tube current modulation (ATCM) technologies have been developed by scanner manufacturers and are now in use clinically. Tube current modulation acts to adjust the x ray output of the scanner during a scan in order to maintain constant photon fluence at the detector. This modulation results in a decrease in tube current along anatomical areas with less photon attenuation and correspondingly higher tube current s though denser sections of anatom y. The mechanism by which tube current is modulated varies between scanner manufacturers, and variations include modulation along the z, and x y axes. When used properly, tube current modulation techniques can help to reduce patient doses during

PAGE 17

17 CT exams while ensuring uniform image quality between exams, and between slices with in a single exam. It should be noted that modulation systems can also lead to higher patient doses if they are used improperly, or without sufficient user knowledge and training. D ose concern in CT imaging is not a new concept however, and over the years several strategies have been developed to monitor exposure. Currently, the Computed Tomography Dose Index (CTDI) and its variations are the primary dose measurement concept in CT. Numerous variations on the traditional CTDI have been proposed over the years in order to make the metric more relevant as scanner technology progressed from single slice axial scans to multi detector helical scans. Its current iteration, CTDI vol is int axis, from a series of 6 and is measured using a 10 cm long pencil ion chamber from a single axial tube rotation in each of two circular ac rylic phantoms (Figure 1 1 ) that approximate the human body (32 cm diameter) or head (16 cm diameter). Despite these adaptations, the CTDI dose metric has received a fair share of criticism for not adequately predicting patient doses in modern CT exams. 7 12 Despite not accurately reflecting patient doses, CTDI retains a practical use as a quality assurance tool due to its widespread use and utility in measuring tube output for given parameters 8 It is worth noting that the current CTDI phantoms, combined with the single slice, axial acquisition, do not challenge the A TCM systems in modern scanners. As a result, the same CTDI vol value is curre ntly used to predict patient doses for both fixed tube current and ATCM scans, though the two exams can deliver

PAGE 18

18 significantly different patient doses. In fact, the concept of CTDI breaks down with ATCM scans, as the tube current is constantly changing thro ughout the exam In order to estimate patient organ and effective doses from CT procedures, several approaches have been implemented. Many groups have used computational phantoms in order to model patient organ doses in CT imaging with Monte Carlo method s though all only modeling fixed tube current techniques 13 18 Several of these computational models have been compared to TLD measurements in anthropomorphic Rando phantoms, with varying levels of success. 17, 18 Additionally the ImpaCT group of the UK has compiled a Monte Carlo based spreadsheet (Figure 1 2 ) for use with clinical CT procedures (ImpaCT, London, UK). This commercially available software incorporates user in putted CTDI values and imaging protocol techniques, coupled with built in scanner specific factors including beam and geometric characteristics, in order to estimate specific organ doses resulting from CT imaging procedures. The ImPACT spreadsheet calcula tes dose based on Monte Carlo data that is based on axial, fixed tube current scans performed on geometric phantoms close to 20 years ago and do es not include the tube current modulation currently in widespread clinical use. Users are forced to use an aver age or maximum tube current used in a scan in order to utilize such software, and the accuracy of estimated organ doses can suffer as a result. Additionally, such computational models necessarily rely on measured CTDI values as a benchmark, and such values have been proven to underestimate actual delivered doses. 9 11 Given that current methods for estimating patient organ doses from MDCT exams utilizing ATCM technology are sparse and rely on outdated models, it is the purpose of

PAGE 19

19 this work to catalogue a library of organ doses for various MDCT scans on phantoms representing two different patient sizes in order to determine the effect of patient size on patient organ doses. Background and Significance Concern over th e clinical doses from CT exams has been around for some time, but recent reports have brought the issue to the forefront of national media and underscored the issue. A 2007 review article in the New England Journal of Medicine (NEJM) highlighted the radia tion risks posed by CT imaging and concluded that too many CT scans were being performed in the US. 3 The NEJM article drew national media attention to the issue of CT dose and estimated that 1.5% to 2% of all cancers in the US coul d be attributed to CT studies. It is also noted that while the average dose from a chest or abdominal CT scan is on the order of 15 25 mGy, multiple scans are commonly performed in a single imaging series, and patients commonly undergo multip le imaging series over t he course of their treatments. In order to reduce CT related radiation exposure to patients, the authors of the NEJM study recommended a three part approach: reducing the CT dose in individual exams, replacing CT use with alternate options such as ultrasound or MRI, and reducing the number of CT exams that are prescribed Additionally, a recent article in the Journal of the American Medical room visi ts for injury related conditions increased significantly, without an equal increase in the prevalence of life 19 Most recently, the National Council of Radiation Protection (NCRP) i ssued report 160 stating that in 2006, Americans were exposed to more than seven times as much ionizing radiation from medical procedures as compared to 1980 5 The NCRP attributed

PAGE 20

20 the majority of this increase to advances and increase in clinical CT imaging as w ell as nuclear medicine scans. attempt to quantif y the associated health risks nor specify the actions that should be taken in light of these latest data." 5 The American Association of Physicists in Medicine (AAPM) responded with a press release noting that such exams have revolutionized American medicine and save thousands of lives daily. The AAPM further questioned the value of populat ion averaged doses in estimating individual health risks or effects and pointed out that while more CT scans are being performed today, the average dose per scan has fallen over the past decades. The AAPM also acknowledged risks posed by CT, but maintained that the benefits of an appropriately ordered CT exam far outweigh those risks. Automatic Tube current Modulation In response to these dose concerns in computed tomography, numerous solutions have been suggested, including a general lowering of tube curr ent techniques for all exams. 20 Although this suggesti on would result in lower patient doses, it would come at the expense of noisier images, since image noise is proportional to the number of photons incident on the detector. This increase in noise in illustrated in Figure 1 3 which shows the two CTDI phant oms imaged with the same tube current. As expected, the larger phantom shows a much larger degree of image noise. A decrease in tube current could thus compromise image resolution a nd low contrast delectability. As an alternate approach, CT scanner manufa cturers have developed technologies for lowering patient doses without sacrificing imag e quality. These technologies are referred to as automatic tube current modulation, and act to adjust x ray tube output during the CT scan in either the x y plane (angul ar or in plane modulation) or along the z axis (z axis modulation) in response to changes in patient anatomy and attenuation in

PAGE 21

21 order to reduce patient dose and maintain a constant image quality throughout the scan. 21 In computed tomography exams, selectable techniques such as tube current and tube potential determine the photon flu ence output of the x ra y tube. This fluence, along with the attenuation characteristics of the patient, determines both patient dose as well as the number of photons reaching the detectors, which in turn determines reconstruct ed image noise characteristics. If all other variable s are held constant, a reduction in tube current leads to a reduction in patient dose, but an increase in quantum noise or mot tle in the reconstructed image. Images with too much noise may obscure low contrast lesions or tumors that would normally be visib le in less noisy images and could lead to misdiagnoses or the need to rescan the patient, exposing them to unnecessary radiation In conventional CT, a technologist selects the tube current and tube potential based on patient characteristics such as size a nd weight, as well as the image quality requirements of the pa rticular exam being performed. These techniques are held constant for each slice throughout the exam. Since patients are neither homogeneous in tissue composition, nor circular in exterior body circumference, these fixed techniques lead to variabl e attenuation though the body. This results in a variable number of photons reaching the detectors on the opposite side of the patient for different projection angles. 22 Certain anatomical areas, such as the shoulders, are problematic in that lateral views can have much higher attenuation, up to three orders of magnitude higher, than anterio r po sterior views of the same area. In these instances, tube current must be increased to allow more photons to reach the detector in order to minimize

PAGE 22

22 reconstruction artifacts due to photon starvation through the higher attenuating projection angles. This artifact is illustrated in Figure 1 4 showing anisotropic noise in the direction of the highest attenuation in an elliptical phantom. The result of this increase in tube current is an increased patient dose throughout the entire scan area of the exam. In response to these imaging and dose issues, CT manufacturers have developed automatic tube current modulation techniques that allow the tube current to be automatically adjusted during a CT examination in order to provide lowered patient doses and constan t image noise characteristics. Tube current in low attenuation projections can be greatly reduced without a loss of image quality, thus reduc ing patient dose for the exam. There are currently two major strategies employed by manufacturers to accomplish this task; angular and z axis modulation, which are illustrated in Figure 1 5 along with traditional means of adjusting fixed tube currents for larger patients. 23 Angular modulation was first developed by GE Medical Systems and works to modulate tube current in the x y plane within a single rotation of the tube. 21 Photon fluence is increased through areas of higher attenuation, such as lateral views though the shoulders and decreased in areas with lower attenuation, such as AP v iews through the chest. The first angular modulation system from GE appeared in 1994 (SmartScan), and used two localizer radiographic images, AP and lateral, to determine atte nuation values of th e patient. The tube current was then modulated in a preprogrammed sinusoidal pattern that matched the attenuation characteristics from the scout images These systems attained a dose reduction of up to 20% while maintaining a relatively constant level of i mage noise 21 More recent offerings of the technology by

PAGE 23

23 Siemens employ an online real time anatomy ad apted system ( CARE Dose 4D ) that automatically adjusts the tube current for a given projection based on the attenuation calculated from the previous rotation. 24 In the CARE Dose 4D system, tube current is input from localizer radiographic images, and is adjusted by attenuation information provided from the previous 180 degree projection as seen in Figures 1 6 and 1 7 The CARE Dose 4D system has been shown to produce tube current reductions of up to 90% for the anter i o r posterior projection in regions such as the shoulders with marked asymmetry. 3 Philips Medical Systems also utilizes an angu lar tube current modulation system (Dose Right Dose Modulation) in their CT scanners. The Philips technology modulates current within a single tube rotation according to the square root of the attenuation measured during the previous rota tion. This modula tion technique is based on the fact that image noise is inversely related to the square root of the number of photons captured. 21 The second major type of tube current modulation technique is known as z axis modulation. This technology also acts to adjust the photon output of the x ray tube according to patient specific attenuation characteristics, but unli ke angular modulation, it does not alter tube current within a single rotation of the x ray source. Instead, a scout radiographi c image is taken of the patient and the system calculates the photon flux required in each slice in order to maintain a user des ignated noise level in the reconstructed image. 21 The tube current remains constant for each rotation around the patient, but is altered along the length of the patient as the table translates through the beam. Lower current values are thus used in lower attenuation regions, such as the

PAGE 24

24 chest, lowering patient dose in comparison to higher attenuation regio ns such as the pelvis, as seen in Figures 1 8 and 1 9 Z axis modulation is currently offered by GE medical systems (AutomA) as well as Toshiba and Siemens medical systems, and acts to keep image noise the same in each slice image in an exam The scanners to be utilized in this study utilize two different systems of ATCM. y modulation in the imaging plane as well as z axis modulation along the length of the patient. In the CARE D value according to the imaging needs of the particular scan. A higher reference mAs setting will result in less image noise, but at the expense of higher patient dose. This metric is adjusted only in order to change image quality in the reconstructed image, and not as a result of larger or smaller patients. During an exam, the system uses a single scout image to set the baseline tube current levels along the z axis, while a feedback system is used to adjust the in plane modulation during the course of the scan based on attenuation information from the previous 180 degrees of tube rotation. The system employed by the Toshiba scanners is called Sure Exposure which until recently comprised of z axis modulation only with no changes in tube output within a single tube rotation. Recently, Sure Exposure has been expanded to include in plane modulation as well, which modulates tube current within a slice in a pre planned manner based on information from both AP and lateral scanogram images. In the Sure Exposure system, the user inputs a target image standard deviation, which determines the standard deviation of pixels in the reconstructed image The system also utilizes minimum and maximum tube curre nt values to avoid under and overexposures. It

PAGE 25

25 should be noted that since choice of reconstruction kernel impacts standard deviation in reconstructed images, the choice of reconstruction on the Toshiba system will result in va riations in the mA of the scan Clinical Tube current Modulation Studies Numerous studies have been published testing the effectiveness of ATCM systems in clinical practice, with a majority of them coming out of the radiology department at M assachusetts General Hospital. These studies, headed by Mannudeep Kalra, made use of 16 slice GE scanners equipped with AutomA, a z axis tube current modulation system. The AutomA system is similar to first version of Sure Exposure in that it operates off a user inputted noise index level as well as maximum and minimum tube current thresholds. The first such clinical study involved utilizing tube current modulation in abdominal and pelvic CT exams. 25 Sixty two patients underwent follow up abdom inal CT scans using Z axis modulation. Images from these follow ups were then compa red with previous images obtained using fixed tube current techniques from the same patients, but otherwise using identical imaging parameters; with t he mean interval betwee n the scans wa s 5 months (range 2 8 months). The two sets of images were graded by subspecialty radiologists on the basis of diagnostic acceptability and image noise. Images were graded on a five point scale with 1 being unacceptable, 3 being acce ptable a nd 5 being excellent. Images at five anatomic levels in the abdomen and pelvis were used, and t he total mAs for each exam was also recorded for comparison. The results of the abdominal pelvic study showed no significant differences in the scores for image noise and diagnostic acceptabil ity between the two techniques, and a ny lesions detectable on the manual tube current scans were also detectable o n the Z

PAGE 26

26 axis modulation scans. An average overall mAs reduction of 31.9% (range 18.8% to 87.5%) was found for Z axis modulated scans as compared to fixed tube current techniques. The use of Z axis modulation resulted in an mAs reduction in 87% of all exams. An increase in mAs was found in 13% of exams using Z axis modulation and is attributable to the larger mean weight of patients in this category as compared to those patients that had reduced mAs. Although mAs increased in these exams, a significant (p = 0.1) improvement in image noise and diagnostic accept ability was noted for all cases, a fact that indicates pr evious scans with overweight patients possibly utilized inappropriately low tube current s resulting i n below average image quality. Overall, the study showed that significant dose reduction is possible with Z axis modulation and highlighted the need for pr oper technique settings of both the minimum and maximum current limits in order to ensure appropriate noise levels in images. Another clinical test performed by the same group evaluated dose reduction and scanner performance in the detection of urinary tra ct stones using Z axis ATCM modulation on both patients and imaging phantoms. 26 In the phantom study, sixteen calcium oxalate or calcium phosphate kidney stones with sizes ranging from 2.5 19.2 mm, were embedded in the collecting systems of two freshly harvested bovine kidneys. The kidneys were placed in an elliptical Plexiglas container that was then filled with a saline solution. The phantom was scanned once with a fixed tube current technique, and then once w ith Z axis tube current modulation at 5 different noise index s ettings; 14, 20, 25, 35, and 50, with t he remaining scanning and reconstruction parameters remaining the same. The phantom images were then viewed by two separate radiologists who were blinded to the scanning techniques and graded using the same 5

PAGE 27

27 point scale described above in order to describe the detectability of the stones in the images. In addition to the phantom study, patient studies were conducted in 22 patients using a method similar to that of the abdominal study. The same radiologists evaluated both sets of patient images using the same five point scale for conspicuity and margins around the stones. In the phantom study, both radiologists identified all 16 stones in the fixed tube curr ent technique images, as well as in the Z axis modulated images at no ise indexes of 14, 20, and 25. Three stones each smaller than 5 mm were not identified by either radiologist in images from Z axis modulation with a noise index of 35 and 50. There was no significant difference (p > 0.5) in stone conspicuity, image noise, or diagnostic acceptance between fixed current and Z axis modulated images at a noise index of 14, 20 o r 25. Dose reduction, by means of reduced mAs, was found in all Z axis modulated s cans ranging from 51% in the noise index 14 case, to 92% in the noise index 50 case, although as previously mentioned, not all stones were localized with the noisier techni que. An mAs reduction of 76% was found in the noise index 25 case, the highest noise index where all stones were detected. Kidney phantom images at each noise level are shown in Figure 1 10 Similar dose reductions were found in the patient studies, although only noise indexes of 14 and 20 were used. All stones located with the fixed tu be current scans were also located with Z axis modulation, and no statistically significant difference (p > 0.7) was found in the radiologists ratings for conspicuity, image noise, or diagnostic acceptability in patient images An average reduction in mAs of 43% was found with a noise index of 14, and a 66% reduction was found at a noise index of 20.

PAGE 28

28 Overall the study demonstrated that doses in urinary tract stone detection exams can be greatly reduced by utilizing Z axis modulation techniques. These redu ctions can be higher than those seen in the chest or abdominal imaging due to the high contrast nature of urinary tract stones in soft tissue, allowing for slightly noisier images than required for low contrast detection cases. It was also found that for t hese exams, a 5% reduction in the noise index correlates to a dose (mAs) increase of approximately 10%. The same research group at Massachusetts General Hospital conducted a similar patient study of chest CT studies comparing fixed tube current methods to Z axis modulation. 27 Very similar methodologies were used as in both the previous studies. The study used 53 patients, with two radiologists reading the results of CT exams and rating them based on image noise, diagnostic acceptability and streak artifacts. Again the study found that Z axis modulation provided acceptable imaging of the chest with no significant differences in image noise, diagnostic acceptability, or s treak artifacts, as compared to fixed tube current techniques. The use of Z axis modulation showed average mAs reductions of 18%, 26%, and 38%, with a noise index setting of 12, 12.5, and 15 respectively, when compared to fixed tube current protocols. Agai n tube current was found to increase slightly in heavier patients, which is a result of the system attempting to keep image noise constant throughout the thicker area s of the scan. It was noted that care should be taken by physicians to pay special attenti on to noise requirements from larger patients to avoid unnecessary increases in dose. Monte Carlo Based CT Studies Several researchers have attempted to use Monte Carlo methods to estimate patient CT doses from clinical exams. A 2005 study by DeMarco et al attempted to utilize Monte Carlo to estimate CT dose in MDCT scanners. 28 They used a GE

PAGE 29

29 LightSpeed 16 slice CT scanner along with both an acrylic CTDI body phantom and ATOM anthropomorphic phantom in conjunction with MOSFET and pencil ion chamber de tectors to measure scan doses. The researchers performed both axial and helical scans over the length of the CTDI phantom and used MOSFETS on the anterior surface to measure be am profiles during CT imaging. Monte Carlo simulations were run for identical scans and the results were compared. For a single slice axial dose profile, the researchers were able to match simulation to physical measurements to within 3.4% a cross a range of tube voltages. For axial contiguous and helical scans along the length of the CTDI phantom, it was found that surface doses varied by position by a factor as high as 2 along the z axis, with simulated doses showing good agreeme nt with physical measurements. Similar results were found in scans of the ATOM phantom, though with less agreement betwee n projected and measured dose. The researchers also noted that Monte Carlo systematically underestimated doses in all scans, and that the magnitude of surface dose for the ATOM phantom was higher than tha t observed in the CTDI phantom. Overall, the DeMarco paper demonstrated that Monte Carlo could be fairly accurate in simulating dose profiles and surface doses for fixed tube current MDCT scans, though no attempts were made to benchmark organ doses or mode l ATCM systems. A 2008 study by Deak et al. also used physical measurements to validate Monte Carlo MDCT simulations. 18 This study utilized a Siemens 64 sli ce scanner as well as a pencil ion chamber and TLDs to measure single slice dose profiles in a variety of phantoms, including CTDI head and body, a Rando phantom, a liver phantom, and adult and pediatric thorax phantoms. Simulations of weighted CTDI like m easurements in the

PAGE 30

30 circular and other phantoms were found to be in agreement with measured values to within 10%. Scan dose profiles measured with TLDs and compared to Monte Carlo simulations were also found to be in agreement. This study again showed phy sical measurements validating Monte Carlo simulation of MDCT imaging. It is presumed that similar methods could be used to obtain patient organ dose information from clinical scans, but no such efforts were taken in this work, nor was ATCM addressed. A 2 008 Medical Physics paper by Perisinakis and Tzedakis attempted to determine the appropriateness of Monte Carlo derived patient dose information in CT imaging. 13 This study examined the errors associated with the e stimation of patient doses in the ImpaCT spreadsheet; namely in using data derived from axial acquisitions to model helical scans, in calculating all doses for a single patient size, and in estimating doses from a CT exposure performed on a specific scanne r using Monte Carlo data der ived from a different scanner. The effects of changes in patient body size were also explored as two mathematical computational phantoms were used; both representing an approximately 75 th percentile by height hermaphroditic pati ent of 55 th and 95 th percentiles in weight. This study found that organ doses varied by up to 15% when comparing axial to helical scans on the same techniques and scan volumes. Further, when investigating the effects of patient body size, simulations sho wed variation of up to 34% between the two mathematical phantoms. The authors determined that scanner specific Monte Carlo data sets are only needed if the desired level of accuracy in patient effective doses is below 10%. A point is also made that when us ing software such as ImpaCT to model helical scans, care must be made to include added volume on either end of the scan to account for z overscanning. All

PAGE 31

31 simulated scans in the paper were conducted with fixed tube current scans, as the authors note that currently available Monte Carlo tools can not accurately model ATCM function. Several recent studies out of UCLA have begun modeling the effects of patient size on CT doses and have even begun incorporating ATCM scans into their models. A 2007 study by De Marco et al. 29 used Monte Carlo methods to examine the eff ects of patient size on radiation doses from CT scans. They utilized eight different voxelized 3 d phantoms, including six adults, and simulated both full body and thoracic CT exams with fixed tube current techniques. They found that patient effective dose s for the same techniques decreased fairly linearly when plotted against body weight, but that individual organ doses did not always follow recognizable patterns. Erin Angel, also of the UCLA group, has recently published several Monte Carlo based studies that incorporate ATCM. 30, 31 In one study, retrospective dose reconstructions were done on thirty patients of various sizes. Actual patient CT scan data was used and the lungs and glandular breast tissue organs wer e segmented out for study. Using the know ATCM output specific to each scan from a Siemens 16 slice scanner, Monte Carlo dose reconstructions were completed for all patients and the effects of patient size on lung on breast dose were investigated. Fixed t ube current scans were also simulated. The study showed similar results to the Demarco paper 29 in regards to reduced patients doses as patient size increases in fixed tube current scans, though the author found better correlation by plotting organ doses versus patient perimeter as opposed to weight. These results are shown in Figure 1 11 This trend was reversed in the tube current modulated scans, with larger patients having generally

PAGE 32

32 higher doses for both lung and breast as the ATCM system increased tube current in order to maintain image quality. The results of the ATCM simulations are shown in Figure 1 1 2 It should be noted that it is possible to fit a curve to the dose vs. perimeter plots that could used to estimate patient organ doses for any generic patient based only on their size. Overall, several resear ch groups have used Monte Carlo methods to model CT scanner doses and output, but much of the research has been focused on surface doses and single slice dose profiles, which do little to aid in e stimating patient organ doses. The UCLA group is currently t he only research group incorporating tube current modulation into their simulations, hinting at the effects of patient size on specific organ doses. Even these simulations are retrospective and require specific scanner output information from the patient s simulation packages that can be used to model generic phantoms Current Standards of Dose Reporting in Clinical CT The current method of measuring clinical scanner output and dose is the CTDI vol, whi along the z 6 In order to obtain CTDI vol pencil ion chamber measurements are taken at the cente r and top peripheral points on the acrylic CTDI head and body phantoms (Figure 1 1). These measurements are taken for a single axial slice through the center of the phantom for tube voltages and collimation settings matching those of clinical exa minations. The ion chamber measurements are weighted (2/3 x peripheral dose + 1/3 x center dose) to arrive at the weighted CTDI (CTDI w ) for the scanning techniques. CTDI w is then divided by the pitch used in a helical scan to arrive at the CTDI vol value.

PAGE 33

33 The CTD I vol metric has come under fire for not accurately representing dose in modern MDCT scanners. Robert Dixon in particular has published numerous papers on its inadequacies 9 12 He has pointed out that in modern scan ners utilizing beams of 2 cm or more, the 15 cm long CTDI phantoms and 10 cm ion chambers do not collect the entirety of the scatter tail from a single axial slice. 9 This problem leads to an underestimation of the equilibrium center dose by up to 20% in a body phantom, and 10% in a head p hantom for a 20 mm beam width, with effects expected to be larger for wider beam widths. For reference, the 16 slice Siemens scanner to be utilized in this study has a maximum beam width of 24 mm while the Toshiba scanner use d has a maximum beam width of 1 60 mm, which is used for specialized procedures, but only widths of 8 mm, 14 mm, and 32 mm were used for clinical scans. Further, other researchers have questioned the validity of the weighting of the peripheral and center measurements when calculating CTD Iw, proposing that an equal weighting is more realistic. 32 Another disadvantage of the current CTDI metrics is that the y do not take the size of the patient being examined into account, as a single 32 cm diameter phantom is used tube current s, a CTDI vol value for the same techniques used on two patients of drastically different sizes would b e the same, even though the doses incurred by those patients would be drastically different (to say nothing of differences in image quality). This concept is illustrated in Figures 1 1 3 and 1 1 4 which are reproduced with permission from a presentation by Michael McNitt Grey to the AAPM in July of 2010

PAGE 34

34 Many modern CT scanners report post scan CTDI vol values based on either the average or maximum tube current used during the scan. These values can be helpful indicators of general patient dose, but attenti on must be paid to how a system is coming to these values as different manufacturers may utilize differing methodology for arriving at the parameter. O ne must also keep in mind that individual anatomical areas within a scan volume can be subject to differi ng tube current s, and as such, differing doses. The Toshiba scanner used in this work reports a CTDI vol value based on the maximum mAs value expected during a scan, but reports the DLP value based on the mean mAs value during a scan. The Siemens scanner re ports CTDI vol based on the mean mAs value utilized during the scan. used in for CT quality assurance and dose optimization bear an increasingly distant relationship to organ do 10 He suggests replacing the CTDI with a metric that is a direct surrogate of cancer risk, specifically measured organ doses in realistic phantoms. In lieu of organ dose measurements, Dixon has also proposed a method involving small volume ion chambers in a 40 cm long version of the standard CTDI body phantom that can accurately measure the equilibrium dose in modern CT techniques and scanners. 11 The FOC dosimetry system to be utilized in this work is directly applicable to the CT dose m easurements described by Dixon. Recently, t he American association of Physicists in Medicine (AAPM) has released a draft of a proposed change in methodology for measuring and reporting dose in CT that is similar to that described by Dixon 33 AAPM Report 111 outlines the use of small volume ion chambers and helically scanning the entire length of a multiple CTDI

PAGE 35

35 phantom set up instead of the 10 cm pencil chambers and single slice procedure currently utilized. It r emains to be seen if this new methodology will overtake traditional CTDI and we worked into governmental regulations as the primary dose metric in CT, but it promises to be a slightly more realistic dose metric for modern scanners. In summary, the current metric used to estimate patient doses from CT exams is not accurate and is based on scanner technology that is more than 20 years old. Modern scanners utilize helical acquisition, much wider beams, and most importantly tube current modulation schemes in o rder to obtain clinical images, and neither CTDI vol nor spreadsheet based dose calculators based on Monte Carlo simulations account for these advances in imaging technology. Only recently have Monte Carlo simulations begun to address the A TCM systems used in most clinical protocols, and initial results indicate a correlation between patient organ dose and patient size. It wa s the goal of my work to produce a library of specific organ doses for multiple sized patients on multiple scanners utilizing ATCM tech nology. These doses were then used to determine the effects of patient size on organ doses and address the feasibility of patient size dependent weighting factors to be applied to CTDI vol values in order to more realistically predict pa tient doses from CT procedures. Objectives of this Research The objective of this research was to investigate the effect of patient size on specific organ doses in tube current modulated CT exams and develop the tools and methods necessary to more accurately quantify the radi ation doses to patients resulting from these procedures. In order to accomplish this objective, several specific tasks were carried out.

PAGE 36

36 1. Development of an adipose tissue equivalent substitute material and fabrication of a 90th percentile phantom add on fo r an existing 50th percentile by weight anthropomorphic phantom. 2. Creation of a library of specific organ doses for clinical CT scans performed using ATCM. Exams were performed on both a 50 th and 90 th percentile by weight male phantom, and on both a Siemens and Toshiba CT scanner. Further, the Toshiba exams were repeated with three different beam widths (8 mm, 16 mm, and 32 mm) in order to investigate the effects of beam widths on patient organ doses. 3. Development of a method to accurately estimate patient c linical organ doses for a range of MDCT procedures. Using the dose information collected, a method for mor e accurately estimating patient size specific organ doses for tube current modulated MDCT procedures was developed. 4. Collection of data relevant to th e creation of Monte Carlo based method for modeling automatic tube current modulation. As a corollary to the primary focus of this research, both fixed and tube current modulated dose information was collected on the Siemens CT scanner in order to provide dose information useful for the possible creation of Monte Carlo methods to model the Siemens ATCM system for retrospective epidemiological patient dose reconstructions. Figure 1 1. CTDI head and body phantoms

PAGE 37

37 Figure 1 2. Scr eenshot of ImpaCT dosimetry calculator

PAGE 38

38 Figure 1 3. Increased image noise as a result of low tube current in the larger phantom ( page 7 Figure 1 )] Figure 1 4. Anisotropic noise as a result of photon starvation in the direction of highest attenuation in an elliptical phantom. [Reprinted with permission from M. Geis, tube current modul Phys 26 (1999) (page 2245, Figure 9)]

PAGE 39

39 Figure 1 5 Three types of automatic tube current modulation: a) patient size modulation: higher mA is used for larger patient, b) z axis ATCM: higher mA used through more attenuating z axis positions, c) angular ATCM: the degree of modulation depends on asymmetry at each position along the z axis. d) combination of angular and z axis ATCM. [Reprinted with permission from N. MHRA Report 05016 (200 5 ) (p age 11 Figure 3 )]

PAGE 40

40 Figure 1 6 Angular Tube Modulation: Tube current remains constant for the first 180 degrees rotation about the shoulder phantom before being modulated to match attenuation. [Reprinted with permission from C Suess Dose optimizatio n in pediatric CT: current technology and future innovations Pediatr Radiol : 32 729 734 (200 2 ) (page 73 2 Figure 1 )] Figure 1 7 Angular Tube Modulation: Tube current is adjusted in real time to match patient attenuation. Larger currents are neede d through the shoulder region with less required through the thorax due to lower attenuation. [Reprinted with Dose optimization in pediatric CT: current technology and future innovations 32, 729 734 (2002) (pag e 733, Figure 5 )]

PAGE 41

41 Figure 1 8 Z axis modulation: A scout radiographic image is used to determine the appropriate mAs per rotation based on patient attenuation characteristics Dose optimization in pediatric CT: current technology and future innovations 32, 729 734 (2002) (page 733, Figure 4)] Figure 1 9 Z axis modulation: Tube current varies per slice as determined by a scout radiograph Automatic tube current modulation in CT A comparison between different solutions, Prot Dosim: 114, 308 312 (2005) (page 311, Figure 3)]

PAGE 42

42 Figure 1 10 Kidney phantom. Kidney stones visualized in bovine kidneys scanned with a) fixed tube current technique b) Z axis modulation noise index =50, c) noise index=35, d) noise index=25, e) noise index=20, f) noise index=14 Detection of urinary tract stones at low radiation dose CT with Z axis automatic tube current mo dulation: Phantom and clinical studies, Radiology : 235 523 529 (2005) (page 526 Figure 1 )]

PAGE 43

43 Figure 1 11 Plot of patient organ doses versus patient perimeter from fixed tube current CT scans. Glandular breast dose is represented by diamonds, with lung doses represented by circles. [Reprinted with permission from E. Angel, Tube current 193 1340 1345 (2009) (page 1343, Figure 2)]

PAGE 44

44 Figure 1 1 2 Plot of patient organ do ses versus patient perimeter from tube current modulated CT scans. Glandular breast dose is represented by diamonds, with lung doses represented by circles. [Reprinted with permission from E. Angel, ffects of Tube current 193 1340 1345 (2009) (page 1343, Figure 3)]

PAGE 45

45 Figure 1 1 3 Scenario illustrating two different sized patients receiving a CT scan with the same techniques. The CTDI value for each is the same, even though the d ose to the smaller patient is greater than that to the larger patient. [used with permission from M. McNitt Gray from a presentation to the American Association of Physicists in Medicine at the 2010 annual meeting in Philadelphia, PA, July 2010.] Figure 1 1 4 Scenario illustrating two different sized patients receiveing a CT scan with size adjusted techniques. The CTDI value for the smaller patient is half that of the larger patient, even though the dose received by each patient is similar. [used with pe rmission from M. McNitt Gray from a presentation to the American Association of Physicists in Medicine at the 2010 annual meeting in Philadelphia, PA, July 2010.]

PAGE 46

46 CHAPTER 2 FIBER OPTIC DOSIMETRY SYSTEM & P HANTOM CONSTRUCTION Fiber Optic Coupled Dosimetry System Fiber optic System Development Dose measurements in this work were taken using a custom fiber optic coupled dosimetry (FOCD) system as described in Hyer et al. 34 The system utilizes a 2 mm long, 500 m diameter cylindrical plastic scintillating element (BCF 12, Sa int Gobain Crystals, Nemours, France) mechanic ally coupled to a 400 m diameter, 2 meter long fiber optic transmission cable (400 UV, Ocean Optics, Dunedin, FL) Both the scintillator and the transmission fiber were wrapped in heat shrink tu b ing to prevent any ambient light leakage which could affect measurements. The fiber optic cable wa s then coupled to a photomultiplier tube (PMT) (H7467, Hamamatsu Corporation, Bridgewater, NJ) Radiation incident on the scintillator produces visible light with a peak a t 400 um wavelength. This light i s transmitted through the fiber optic cable to the PMT where the light signal is amplified and converted into an electrical signal which is routed through a serial to USB hub (UPort 1610 8, Moxa Inc., Brea, CA). This resul ting signal is read out using a laptop running a custom Matlab computer program, providing real time dose and dose rate information that can be viewed during the course of a scan. A schematic of the FOCD system is shown in Figure 2 1. The completed system makes use of 5 PMTs, allowing for two simultaneous dose measurements (each measurements requiring one PMT for signal readout, and one PMT for background readout). During the course of development, it was discovered that the glass fiber optic transmission f ibers used in the system also fluoresced in the presence of ionizing radiation in the diagnostic energy range This inherent fluorescence in the fibers meant

PAGE 47

47 that dose readings made for the same exposure could differ depending how much of the fiber was in the primary radiation beam. In order to address this issue, a second was employed in order to subtract out the light signal contributed by the fiber itself. This dummy fiber was covered in the same heat shrink tuning as the signal fiber, but did not have a scintillation element coupled to the end. The dummy fiber was run to its own PMT, and its signal was subtracted from the total signal of the scintillator fiber, thus eliminating the inherent fiber fluorescence and leaving only the signal fro m the plastic scintillation element. The dummy fiber is shown in Figure 2 1, though in the finished fiber it is housed within the same heat shrink tubing as the signal fiber, and not separately as shown in the illustration A photo of a finished fiber is s hown in Figure 2 2. FOCD System Calibration Because the final output of the fiber optic system is a total number of counts (signal fiber counts dummy fiber counts), it was necessary to calibrate the system to a known detection sys tem. Prior to calibratin g the fibers, it was also necessary to account for differences in the response of the five different PMTs used in the system. In order to accomplish this, a single fiber was exposed to the same techniques from a Toshiba Aquilion One CT scanner with the tub clock position. The full beam width of 16 cm was used and the exposure was repeated five times to account for variations in both the x ray tube output and the fiber readout. These five exposure measurements were repeated wit h the same fiber connected to each of the five PMTs in the FOCD system and the data was used to create scaling ratios to account for differences in the response of the PMTs. Once the difference in PMT response were accounted for, it was necessary to creat e calibration factors in order to convert the total number of counts output by the

PAGE 48

48 FOCD system into a dose measurement. For the purpose of this work, a 0.6 cm 3 small v olume ion chamber ( 10x5 0.6 Radcal Corporation, Monrovia, CA ) was used in a CTDI acrylic body phantom 15 cm in length and 32 cm in diameter The ion chamber was placed in both the center and top holes of the CTDI phantom for measurements, and the phantom was placed at isocenter in the CT scanner. The calibration procedure was repeated on b oth the Siemens Somatom Sensation 16, and Toshiba Aquillion One scanners used for this work in order to determine if differences in filtration and beam quality affected the response of the FOCD system The full 15 cm length of the phantom was helically scanne d to ensure that the entire active area of the ion chamber was exposed. Five consecutive measurements were taken at each location with the ion chamber, and then with the FOCD fibers. A system integration time of 150 ms was utilized on the FOCD control prog ram for the calibration process, and due to the nature of the system it was necessary to use this same integration time throughout the data collection in this work. The average of the exposure measurements and fiber counts obtained from the calibration pr ocedure were used to produce a calibration factor for each fiber. A calibration factor was calculated for both the peripheral and center positions in the CTDI phantom to account for differences in beam quality as a function of depth in phantom and increase d contribution of scatter radiation. Both the center and peripheral calibration factors were averaged to produce a single calibration factor per fiber, though the differences between the two were minor, and on the order of less than 2%. Fiftieth Percentile Phantom Construction In order to determine the effects of patient size on organ doses during CT exams it was necessary to utilize physical anthropomorphic phantoms of multiple sizes. A

PAGE 49

49 physical phantom representing a 50 th percentile by height and weight m ale was previously developed for fiber optic coupled organ dose measurements in radiological imaging 35 This 50 th per ce ntile phantom consists of 186 slices, each 5 mm in thickness, and constructed of lung, bone and tissue equivalent materials The tissue equivalent materials were developed to mimic the mass energy attenuation properties of human lung, bone, and soft tissue in the d iagnostic imaging energy range. Computational Phantom and Mold Fabrication The physical anthropomorphic phantom was developed from a computational University of Florida. 36 This model was one of the first of a series of hybrid computational models; in that it incorporates the anatomical accuracy and realism of a voxelized tomographic phantom based on actual CT data with the ease of use and scalability of a geometric, stylized model. The or iginal data set came from a full body CT scan of a 36 year old Korean male. The outlines of major organs and tissues in the individual slice images were segmented by hand, and adjusted as needed in order to match reference organ mass data for a 50 th percen tile by height and weight male. A simplified computational dataset was created from the fully segmented data, with delineations only for bone, lung, and soft tissue. Based on this dataset, individual bitmap images were created for each slice and these ima ges were used as the basis for the construction of the anthropomorphic phantom. Extruded p olystyrene f oam insulation board was used along with a Vision 1624 engraving system and VisionPro7 software ( Vision Engraving, Phoenix, A Z ) to create a mold for each phantom slice. An outline of the slice making process is shown in Figure 2 3

PAGE 50

50 Tissue Equivalent Substitute Development Once the mold had been created, a soft tissue equivalent substitute (STES) material based on the commercially available urethane mater 121/30 (Smooth On Inc. Easton, PA) was poured into it and allowed to cure overnight. The 121/30 consistency that is durable, clean, and easy to work with. The addi tion and thorough mixing of 2.8% by weight of powdered calcium carbonate to the uncured urethane brings the attenuation and the density of the soft tissue equivalent material equal to that of human soft tissue. Adipose tissue was originally not specificall y modeled in the 50 th percentile phantom. The STES material was designed as a homogenous soft tissue analogue that encompasses skeletal muscle, internal organs, connective tissues, and intra abdominal adipose tissue. Once the STES material had cured in t he molds, it was removed from the foam, leaving voids in the positions of both bone and lung tissue. A bone tissue equivalent substitute (BTES) material was then filled into the corresponding voids and allowed to harden. This BTES material was originally d eveloped by Jones et al 37 and consists of a mixture of Araldit e and Jeffamine epoxies (Huntsman Corp., Woodlands, TX), with silicon dioxide a nd calcium carbonate additives. The BTES material was designed to represent a homogeneous mixture of cortical and trabecular spongiosa. After the BTES had cured, the individual slices were sanded down to ensure they were flat. This was necessary as the BTES material is thick and difficult to pour evenly while ensuring full coverage within the voids.

PAGE 51

51 After the BTES material was sanded, a lung tissues equivalent substitute (LTES) was used to fill the remaining voids representing lung tissue. The LTES material was created from a mixture of the STES material and poly fil polystyrene micro beads (Fairfield Processing, Danbury, CT), mixed in a 10:1 ratio. The density of the LTES was c hosen to match the density of a fully inspired lung (0.33 g/cm 3 ), as most radiological exams are performed while a patient holds their breath. Once all tissue equivalent materials were cured in each slice, the individual slices were glued together into s ections of variable thickness, ranging from one to five centimeters. Gaps were left at the positions of organs of interest for dose measurements and t he original fully segmented image files from the computational model were used to trace the specific locat io ns of organs within each slice. A photograph of a completed phantom slice is shown in Figure 2 4 While the entire computational phantom consists of 353 slices covering the entire body, the absence of radiosensitive organs in the legs and time consuming nature of phantom construction led to the decision to only fabricate from the top of the head to the mid thigh. Arms were built along with the phantom, but were removed for measurements in order to closer replicate the actual exam conditions of a CT scan, where the patient commonly holds their arms above their head. It should be noted that simply removing the arms from a phantom designed with arms at its side will create slight anatomical differences as compared to if the dataset had been taken with the ar ms raised above the head. These anatomical differences include the inclusion of several centimeters of the humerus in the chest section which would not normally be in the field, as well as minor differences to the angles of the clavicles and scapu lae. The

PAGE 52

52 effect of these anatomical differences is expected to be minor and should only impact dose points for the lungs, and esophagus. The completed phantom is 186 slices total, with each slice being 5 mm thick. The original dataset represents a 50 th percentile by height and weight adult male, which equates to a 5 foot 8 inch male, weighing 172 lbs. A photograph of the majority of the completed 50 th percentile phantom is shown in Figure 2 5 The lower portion of the pelvis is not shown in the picture due to stabi lity issues while stacking the phantom vertically. Organ Point Dose Location s As previously mentioned, the outlines of specific organs of interest were traced onto the physical slices for the purpose of proper positioning of fiber optic dosimeters used t o measure organ doses. The organs investigated in this work for each exam (chest, abdominal, pelvic) are listed in Table 2 1. These organs were selected based on their radiosensitivity as defined in ICRP Report 103 indicating that they are more susceptibl e to radiation damage from CT scans 38 All organs with the exception of the kidneys, small intestines, and prostate are assigned an organ weighting factor by the ICRP based on epidemiological and empirical data related to their radiosensitivity. The aforementioned organs fall b ut represent organs that are either wholly or partially irradiated in the common chest, abdomen, and pelvic procedures. For small organs such as the gonads, thyroid, and prostate, a single dose point loca ted at the centroid of the organ was used to represent the average organ dose to the organs from a particular CT procedure. Large and diffuse organs such a s the esopha gus, liver, and intestines required multiple point dose locations within and between slic es in order to accurately capture the variation in organ dose Skin surface

PAGE 53

53 measurements were taken at two points, one on the anterior and one on the lateral surface of the patient at the midpoint of each scan length. Breast dose was measured as a single d ose point 1 cm below the skin surface at the location of each breast on the The breast measurements were intended to show an approximation of breast dose for CT scans. The measured doses at each point in the larger organs were averaged together Many organs were only partially irradiated in a particular exam procedure. In these instances, where dose points were located both inside and outside the primary beam during the course of a scan, dose measurements were taken for every dose point compromising the organ. The field field and out of field dose points could and give misleading results. A scanogram 6 as a ref erence for the location and distribution of dose points. Ninetieth Percentile Phantom Construction In order to measure the effects of patient size on organ doses in adaptive tube current modulated CT exams, it was necessary to have physical phantoms of m ultiple sizes. Due to the time consuming nature of the phantom construction described in the previous section it was decided to build an adipose tissue equivalent add on to the existing phantom in order to model a 90 th percentile by weight male. Construct ing an add on instead of an entirely separate phantom allowed for greatly reduced production time and materials cost needed to build the phantom, with minimal downside.

PAGE 54

54 Adipose Tissue Equivalent Substitute Material Development In order to build an add on to the existing phantom, an adipose tissue equivalent material has to be developed. The same urethane material base utilized in the development of the STES material was used as a base for the adipose tissue equivalent substitute (ATES). As with the STES a nd LTES, the ATES material was designed to match the attenuation of it s human analogue, in this case subcutaneous abdominal fat, in the diagnostic energy range. A 1999 Radiology paper by Yoshizumi et al. was used as a reference for determining an appropria te HU target for the material. 39 According to their results based on a survey of 120 patients over a range of ages and body compositions, the average HU of subcutaneous abdominal fat was found to be 93 +/ 25. Us various concentrations of phenolic microspheres (System Three, Auburn, WA) were created and subsequently scanned in a Siemens CT scanner at 120 kV. The HU values of the various s amples were recorded, and further samples were created based on this initial data. Using this iterative method a concentration of 2% by weight phenolic producing a HU value of approximately 100. A graph showing the results of the iterative sample testing is shown in Figure 2 7 with each point on the graph representing a sample of a given concentration of microspheres, and its corresp onding HU value after scanning. The densi ty of the final ATES material was measured to be 0.88 g/cm 3 which is in line with published values for subcutaneous fat (0.9 g/cm 3 ). Phantom Construction Once the ATES material was developed, work on building the 90 th percentile phantom add on was initiat ed. A hybrid computational anthropomorphic phantom

PAGE 55

55 created by Perry Johnson of the University of Florida representing a 50 th percentile by height but a 90 th percentile by weight male was used as a template for the physical phantom construction. 40 This dataset was based on target parameters for standing height, sitting height and total body mass, as well as dimensions such as waist, arm, buttocks, and thigh c ircumference, as defined by the National Center for Health Statistics databases The 90 th percentile hybrid computational phantom was based on the original 50 th percentile by height and weight male phantom used in the construction of the 50 th percentile ph antom but with alterations made to increase organ size and to add additional adipose tissue to reach the weight target. The 90 th percentile by weight phantom has a total weight of 214 lbs, as compared to the 172 lbs of the 50 th percentile phantom. A pictu re of both the 50 th and 90 th p ercentile by weight male hybrid computational phantoms is shown in Figure 2 8 Because the 90 th percentile physical phantom was being constructed as an add on, the changes made to organ masses from the 50 th to the 90 th percent ile computational phantoms were ignored, leaving organ size and position unchanged between the two physical phantoms. This decision could potentially affect organ dose measurements, as the exact size and placement of organs would likely differ in the large r phantom, but was a necessary concession to building a 90 th percentile phantom. Further, specific organ placement can vary between individuals of similar size and even within the same individual depending on digestion patterns, so specific organ placemen t inherently has a degree of error that is unavoidable. Due to these circumstances, the choice to build a 90 th percentile add on was deemed appropriate and its effects on final dose measurements minimal for the purposes of this work

PAGE 56

56 As previously mentione d, the 90 th % phantom add on was not built in 5 mm sections, but in larger sections of variable thickness (ranging from 1.5 8.5 cm) matching the 18 glued together sections of the 50 th percentile phantom. The 90 th percentile add on was only built from the neck down; as head exams were not considered for this study and increases in body fat do not typically produce significant changes in head size. The decision to build the add on in sections was made in order to make use of the phantom simpler, as well as to make the fabrication process more efficient. The resulting phantom has a lower z axis resolution as compared to the 50 th percentile phantom, because of the larger sections, but was still adequate for testing A TCM response to changes in patient size. Bi tmap images of the 90 th percentile hybrid computational phantom slices were used to create molds for phantom construction. Unlike the 50 th percentile phantom construction, no attention was paid to the segmented internal organs and only an outer body contou r was used. In cases where the outer body contour varied throughout a 1 cm section (the thickness of the foam used for construction) the largest outer body contour found in the section was used throughout. Eighteen sections total were created. Molds were created out of the same extruded polystyrene foam insulation board used to create the 50 th percentile phantom, but instead of milling molds 5 mm into the foam, the outer body contour of each section was cut entirely though the foam. In larger phantom secti ons, multiple pieces of foam were used and then held together and sealed to a base using masking tape. A photograph of a completed mold is shown in Figure 2 9 Once all eighteen molds were complete, the 50 th percentile phantom section associated with each was wrapped tightly in plastic wrap to protect it from the liquid

PAGE 57

57 ATES. Each phantom section was then appropriately centered in the mold, using bones and the segmented 90 th percentile dataset as landmarks. A centered slice in a mold is shown in Figure 2 10 The adipose tissue equivalent substitute material was then prepared, mixing a 1:1 ratio of both parts of the urethane base material with 2% by weight phenolic microspheres, taking care to ensure proper mixing to avoid separation. The ATES material was th en poured into the molds around the existing 50 th percentile phantom sections and allowed to cure. Photographs of the process are shown in Figures 2 11 and 2 1 2 After allowing the ATES material to cure overnight, the phantoms were removed from their molds The 90 th percentile add on sections were cleaned of any remaining foam by scraping with a r azor. T he plastic wrap was removed from the 50 th percentile phantom sections as well and any ATES that may have leaked though was removed. A completed section of t he 90 th percentile phantom is shown in Figure 2 1 3 and the entire completed phantom is shown in Figure 2 1 4

PAGE 58

58 Figure 2 1. FOCD system schematic. Power supplies for the serial to USB hub and [Reprinted with permission from D. Hyer, Characterization of a water equivalent fiber optic coupled dosimeter for use in diagnostic radiology 1711 1716 (2009) (page 1712, Figure 2)]

PAGE 59

59 Figure 2 2. Complete fiber optic dosimeter. The single fiber splits into two SMA connectors for separate PMT readout of the signal and dummy fibers. Characterization of a water equivalent fiber optic coupled dosimeter for use in diagnostic radiology Phys: 36 1711 1716 (2009) (page 1712, Figure 1)]

PAGE 60

60 Figure 2 3 Steps in phantom construction process: (top left) segmented CT image; (top right) soft tissue bitmap image; (bottom left) VisionPro software showing engraving path; (bottom right) engraving system milling a soft tissue mold. [Re Construction of anthropomorphic phantoms for use in dosimetry studies Phys: 10 195 204 (2009) (page 199, Figure 1 )]

PAGE 61

61 Figure 2 4 Phantom slice including (a) STES, (b) LTES), and (c) BTES. Heart and liver organ locations are also shown. [Reprinted with permission from J. Winslow, Construction of anthropomorphic phantoms for use in dosimetry studies Appl Clin Med Phys: 10 195 204 (2009) (page 200, Figure 2)]

PAGE 62

62 Figu re 2 5 Completed 50 th percentile anthropomorphic physical phantom Table 2 1. Organs of interest for each exam Chest Exam Abdomen Exam Pelvis Exam Thyroid Liver Colon Lungs Stomach Small Intestines Stomach Kidneys Bladder Liver Colon Prostate Breast S mall Intestines Gonads Esophagus Bone Marrow Bone Marrow Bone Marrow Skin Surface Skin Surface Skin Surface

PAGE 63

63 Figure 2 6. Scanogram image of completed phantom showing individual point dose locations. Additional dose locations were added to the pe lvic region after this image was created in order to better model small intestines and colon doses.

PAGE 64

64 Figure 2 7. HU value of various samples of ATES material as a function of their concentration of microspheres. 200 180 160 140 120 100 80 60 40 20 0 0 1 2 3 4 5 6 HU % Microspheres by Weight Adipose Tissue Equivalent Substitute Material Developsment

PAGE 65

65 Figure 2 8 50 th and 90 th percentile b y weight hybrid computational phantoms. Hybrid Patient Dependent Phantoms Covering Statistical Distributions of Body Morphometry in the U.S. Adulst and Pediatric Population 2060 2 075 (2 009) (page 2068, Figure 2 )]

PAGE 66

66 Figure 2 9 Completed 90 th percentile phantom mold. Multiple 1 cm sheets of foam were cut out and held together and sealed with masking tape.

PAGE 67

67 Figure 2 10 50 th percentile phantom section wrapped in plastic and centered in a mold.

PAGE 68

68 Figure 2 11 90th percentile phantom mold filled with ATES material.

PAGE 69

69 Figure 2 1 2 Multiple phantom sections curing in molds.

PAGE 70

70 Figure 2 1 3 Completed 90 th percentile phantom add on around a section of 50 th percentile phantom. Fiber optic coupled dosimeters are in place for measurement of doses to the kidney and colon.

PAGE 71

71 Figure 2 1 4 Completed 90 th percentile phantom.

PAGE 72

72 CHAPTER 3 CREATION OF CT ORGAN DOSE LIBRARY MDCT Exams After completion of the 90 th percentile ad ipose phantom add on, work began on the collection of specific organ doses for MDCT scans of both phantoms A fter reviewing a survey of mor e than 25,000 billable CT exams, shown in Figure 3 1, performed at Shands Hospital at the University of Florida over a 5 month period in the later part of 2007 a standard chest, pelvic, and abdominal exam were chosen for this study After the survey information was sorted and compiled, it was found that these three exam categories accounted for more than half the CT sca ns prescribed and thus would be the most useful for patient do se measurements and estimation. Head scans were ignored due to the fact that tube current modulation is not typically employed in head exams because of the generally symmetrical shape of the hea d. There is also typically less variation in adult head sizes as compared to body size, and head size does not vary as much with increases in body mass. 41 It has been shown that It should be noted that each category contains multiple speci fic types of exam, including angiography and scans both wi th and without added contrast. For the purpose of this was performed with no repeated scan s as are used in angiography and contrast scenarios, as i t was assumed that doses will scale linearly with increased mAs used in multiple scan protocols and that the estimation of patient dose s could be similarly scaled in such scenarios. All exams were performed on Siemens Sensation 16 and Toshiba Aquilion ONE CT scanners. The S iemens scanner is a 16 slice scanner with 1.5 mm detector widths and a 24 mm total maximum beam width. The Toshiba scanner offers up to 320 slices

PAGE 73

73 and 0.5 mm detector widths, for a 160 mm beam width; but this configuration is only utilized for specific ang iography sequences in brain and cardiac imaging and traditional helical scans are limited to smaller beam widths All exams in this study are clinically acquired with 64 slices (0.5 mm detector width) for a beam width of 32 mm. The 64 slice, 32 mm beam wi dth was used for the collection of organ doses in this study, with additional measurements being collected with both 32 and 16 slices (16 and 8 mm beam widths respectively), in order to investigate the effects of beam width on patient organ doses. Each of the three MDCT exams were performed on both the 50 th and 90 th percentile phantoms using the standard clinical protocols and anatomical markers available online at S hand s Department of Radiology website. 42 Exams were performed current modulation system ( CARE Dose 4D on Siemens, and SureExpsoure on Toshiba) and t he same image quality setting (Quality Reference mAs on the Siemens scanner and standard deviation ( SD ) value on the Toshiba) was utilized for both the 50 th and the 90 th percentile phantoms for each exam. The choice of ATCM image quality setting was made usi ng the standard parameters used in clinical practice. The anatomical clinical scan regions utilized for each exam are shown in Figures 3 2 through 3 4, while t he specific scan protocols for each exam are listed in Table 3 1 through 3 3 Fixed Tube Current Scans In addition to the ATCM scans perf ormed in this study, fixed tube current (FTC) scans were performed on the Siemens scanner in order t o provide information and dose data for future use in Monte Carlo modeling of the Siemens CARE Dose 4D system. The same three CT exams were investigated as in the ATCM scan portion of this work,

PAGE 74

74 but utilizing traditional fixed tube current scans instead of a modulated beam. Because patient dose is directly related to image quality in diagnostic imaging, in order to acc urately compare doses it was necessary to ensure image quality remained similar 5 and matching the outer contours of the 50 th percentile phantom, was created for the measurement of image quality. The image quality slice is constructed in such a way as to house a circular image quality module from a Catphan imaging phantom (The Phantom Laboratory, Salem NY). The module (Figure 3 6) contains high res olution line pairs to allow measurement of spatial resolution in the imaging plane, as well as low contrast and uniformity objects. The Catphan module is comprised of 2.5 cm thick acrylic and provides a uniform area to measure deviation of CT number values at multiple points. This phantom section is important, as the 5 mm slices that the phantom is comprised of often lead to partial volume averaging effects in the reconstructed images, making the anthropomorphic phantom ill suited for image quality measurem ents. A corresponding ATES add on was created for the image quality slice in order to also match the out er contours of the mid section of the 90 th percentile phantom This image quality slice was placed in both the 50 th and 90 th percentile phantoms and sc anned normally for each of the three CT exams using the same anatomical markers in Figures 3 2 through 3 4 I mages taken from these portions of the phantom were to provide a quantitative way to ensure comparable image quality between scans, while not affe cting the functionality of the ATCM systems. Image quality was originally to be objectively evaluated utilizing Image J software to determine spatial resolution,

PAGE 75

75 contrast, and uniformity values images from each scan In reality, neither the high resolution nor low contrast modules of the Catphan were useful due to the level of noise in the reconstructed images. This increased noise is due to the scatter produced by the phantoms surrounding the image quality module, which is normally scanned by itself. The i ncreased noise rendered comparisons between images based on the low and high contrast parameters impossible as t he low contrast modules were completely obscured in the reconstructed images, and none of the high contrast line pairs were able to be resolved A photo of a slice image from the image quality slice is shown in Figure 3 7. Due to the limitations on matching image quality between the fixed and ATCM scans, only CT number (also referred to as Hounsfield Units (HU)) standard deviation values were u sed to compare the two methods. Both phantoms were originally scanned with ATCM for the appropriate Quality Reference mAs setting for the exam being performed. The standard deviation of CT values was then measured for several regions of interest in the uni formity portion of the Catphan module and then averaged to provide a CT number uniformity value for each scan The effective mAs values used by the scanner to produce these images was also recorded. These values for each e xam and are shown in Table 3 4. Fo llowing the collection of image quality data, o rgan doses were measured for FTC scans with a fixed tube current of 300 mAs for the chest, abdominal and pelvic exams This was done to ensure good counting statistics in the collected data, as final dose valu es could be scaled to match th e image quality data due to the linear nature of the mAs to dose relationship for FTC scans. The scaling of the FTC dose values to match the ATCM was done in two ways. The first was to match the image quality level

PAGE 76

76 of the FTC scans to those measured in the ATCM scans. This was accomplished by scanning the image quality section of both the 50 th and 90 th percentile phantoms at a range of FTC mAs settings, recording the standard deviations, and then matching these HU standard devi ation values to those from the ATCM scans. The second method for scaling the FTC dose values was performed by matching the fixed mAs value to the average mAs value used by the ATCM system in the ATCM scans. A plot of the tube response per slice in the recon structed ATCM scans was created, and the average of the effective mAs per slice was used for the FTC matching. The mAs values derived from each of these methods is shown in Table 3 5. The two methods proved to produce similar results with minor differences (within 10% variation in mAs values) for both the abdominal and pelvic exams, but differed greatly for the che s t exam. This difference was attributable to a large increase in tube current though the asymmetrical shoulder region of the chest phantom, leadi ng to higher average mAs values for this exam as compared to the relatively more symmetric abdominal section of the phantom where the image quality phantom section was imaged The impact of these differences will be further discussed in the results and dis cussion portions of this work. Organ Dose Measurement s For the purpose of this study, all organ locations remained constant for all exams and with both phantoms. Again, realistically it is likely that organ sizes would differ in a larger patient, and it ha s been shown that organ sizes vary with patient size 40, 43 but the relationship shows a wide distribution and organ size can vary even within patients of similar body dimensions. This variability, coupled with the fact that changing specific organ dose location s was not logistically feasible with the current phantom set up led to the decision to leave organ point dose location fixed between the 50 th and 90 th

PAGE 77

77 percentile phantom measurements Any resulting changes in dose values are expected to be minimal or less than the magnitude of uncertainty of organ size and location on a patient to patient basis. As such, the only difference between the 50 th and 90 th p ercentile phantom dose measurements was the additional laye r of extra cutaneous adipose tissue. Doses were measured at discrete locations representing organs of interest within each scan volume. The specific organs measured in each exam along with the number of dose points utilized for each organ, are listed in T able s 3 6 through 3 8 For small organs, a single dose was s centroid, while larger organs involve d averaging doses from multiple locations within the organ area to account for do se gradients across the organ. Breast dose was measure d, despite the phantom being male, with a single dose point just below the skin surface at the location of the breasts. Marrow doses were measured at multiple marrow locations milled into the bone within each scan volume. A skin surface dose was measured for each exam at a medial anterior and lateral position on the phantom. In cases where only a partial volume of an organ was within the scan volume, measurements were made at all dose points for that organ, even those outside the primary beam coverage. For such organs, results were presented both as a total average organ dose, including both in and out of field dose points, and as an in field average organ dose where out of field dose points were ignored so as not to dilute the dose information of those po ints receiving higher doses from the primary beam. Specific organ doses for each exam were reported for this work instead of reporting an effective dose (E) for each exam, as the ultimate goal of this project wa s to

PAGE 78

78 produce a method of better estimating o rgan doses for a clinical patient undergoing CT examinations. the whole body that would have a similar risk of aggregated health detriment to the exposure received by a reference perso 44 As such, effective dose does not represent a sp ecific patient, but a hermaphroditic reference individual for which risks have been assessed based on the average for all ages and both sexes for an entire population. Effective dose is an appropriate metric for comparing the health detriment to a referen ce patient for various types of medical examination, as well as for optimizing exam protocols and in the justification of medical exposures. However, it was not the goal of this work to optimize protocol s or justify medical radiation exposure, but instead to provide a method for predicting patient organ doses as a result of these exams. Further, been established can support, and this could lead to decisions being made o n the basis 44 The uncertainty in the calculation of effective dose for medical exposures in a reference patient have been estimated at approximately +/ 40%, and can increase greatly once age, gender, and physical dimensions of specific patients and included. Multiple factors are involve d in the calculation of the uncertainty in E, including uncertainties in the derivation of the organ dose weighting factors, the assumption of a linear no threshold risk model, and uncertainties in the location, size and tissue densities of the organs bein g measured. In order to assess individual risk from CT procedures, it has been recommended that individual organ doses and organ specific absolute risk data based on gender and age be used instead of effective dose. 44, 45

PAGE 79

79 CTDI VOL and ImPACT Dose Estimates After the library of organ doses fo r MDCT exams was collected, comparisons were made to conventional methods of estimating patient dose with CTDI VOL and the ImPACT CT dosimetry spreadsheet. CTDI VOL values were collected on both scanners for mAs values relating to the average effective mAs f rom each scan. As with the image quality portion of this work, these average effective mAs values were calculated by going through reconstructed image sequences of each exam, on each scanner, and recording the effective mAs of each slice. These values were then averaged for each exam to produce an avera ge effective mAs value from the ATCM scans. An average eff mAs value is the best available approximation of scanner output during an ATCM CT exam, as the actual mAs per slice varies, sometimes greatly. CTDI VO L values were collected using each of the three beam widths utilized on the Toshiba scanner. The measured organ doses for each CT exam, both fixed and ATCM, were compared to the CTDI VOL ent doses. In addition to comparing measured organ doses to CTDI VOL values, the measured CTDI VOL for each exam was used as an input for the commercially available ImPACT CT dose assessment spreadsheet. This dose predictor is based on single slice, axial fixed tube current Monte Carlo CT simulations performed using a mathematical MIRD phantom Correction factors are used in the spreadsheet in order to better represent modern CT scanners and procedures. The spreadsheet allows users to select a specific scan ner, beam width, acquisition techniques, and scan volume and reports organ doses for all ICRP reference organs It should be noted that the ImPACT spreadsheet does not account for variation in patient size or the actions of tube current modulation Users c an input the higher techniques used by the scanner for larger

PAGE 80

80 patients in ATCM scans but this dose is delivered to the same sized patient as would be used to model a smaller patient. Comparison was made between the doses predicted with the ImPACT spreadsh eet and those directly measured in the phantoms. The most recent version of the ImPACT spreadsheet does not have a specific model for the Toshiba Aquilion ONE scanner. It does contain a model for an Aquilion 16 slice scanner and provides options for a beam width of up to 32 mm, matching that used in the 64 slice scans performed on the Aquilion ONE scanner. In the absence of a specific Aquilion ONE model the 16 slice scanner was used to produce the ImPACT estimated doses for all Toshiba scans. Figure 3 1 Survey results of 25,000 CT scans at Shands Hospital at the University of Florida. Abdominal 21% Chest 17% Pelvis 17% Head 25% All Other 20% CT Exams at Shands Over 5 Month Period

PAGE 81

81 Figure 3 2. Scan range for chest exam. Thoracic inlet to the top of the kidneys. Table 3 1. Chest exam techniques Scanner Slices Detector width Beam wi dth kVp Pitch Rot. t ime Image quality setting Siemens 16 1.5 mm 24 mm 120 0.75 0.5 s ref mAs 200 Toshiba 16 0.5 mm 8 mm 120 0.938 0.6 s SD 10 Toshiba 32 0.5 mm 16 mm 120 0.844 0.6 s SD 10 Toshiba 64 0.5 mm 32 mm 120 0.828 0.6 s SD 10

PAGE 82

82 Figure 3 3. Scan range for abdominal exam. Dome of diaphragm to the top of the iliac crest Table 3 2. Abdominal exam techniques Scanner Slices Detector width Beam width kVp Pitch Rot. t ime Image quality setting Siemens 16 1.5 mm 24 mm 120 0.75 0.5 s ref mAs 240 Toshiba 16 0.5 mm 8 mm 120 0.938 0.6 s SD 10 Toshiba 32 0.5 mm 16 mm 120 0.844 0.6 s SD 10 Toshiba 64 0.5 mm 32 mm 120 0.828 0.6 s SD 10

PAGE 83

83 Figure 3 4. Scan range for pelvic exam. Illiac crest to the lesser Trochanter. Table 3 3. Pelvic e xam techniques Scanner Slices Detector width Beam width kVp Pitch Rot. t ime Image quality setting Siemens 16 1.5 mm 24 mm 120 0.75 0.5 s ref mAs 240 Toshiba 16 0.5 mm 8 mm 120 0.938 0.6 s SD 10 Toshiba 32 0.5 mm 16 mm 120 0.844 0.6 s SD 10 Toshiba 64 0.5 mm 32 mm 120 0.828 0.6 s SD 10 Table 3 4. HU standard deviations in the image quality phantom for each ATCM exam. Siemens 50th % Phantom 90th % Phantom Exam Quality Ref mAs HU Std Dev. eff mAs HU Std Dev. eff mAs Chest 200 14.2 1 180 15.8 280 Abdomen 240 13.24 210 14.8 333 Pelvis 240 13.24 210 14.8 333

PAGE 84

84 Table 3 5. Fixed tube current mAs values utilized to scale organ dose measurements methods. Match t o SD value Match to average ATCM mAs Exam 50th 90th 50th 90th Chest 150 220 211 250 Abdomen 170 275 181 267 Pelvis 180 290 176 270 Figure 3 5. Dummy slice created for the assessment of image quality. Figure 3 6 Close up of the Catphan image quality module with line pairs on the left and low contrast test object on the right.

PAGE 85

85 Figure 3 7. Slice images from image quality scan for 50 th (left) and 90 th (right) percentile phantoms. Increased scatter from the surrounding phantom rendered the high and low contrast modules useless for the purposes of image quality comparison Table 3 6. Organs and their corresponding number of dose points for a chest exam Chest exam Number of dose points Thyroid 1 Lungs 6 Stomach 4 Liver 4 Breast 2* Esophagus 6 Bone Marrow 8 Skin Surface 2** single point at each 'breast' ** anterior and lateral surface

PAGE 86

86 Table 3 7. Organs and their corresponding number of dose points for an abdominal exam Abdomen exam Number of dose points Liver 4 Stomach 4 Kidneys 2* Colon 8 Small Intestines 6 Bone Marrow 4 Skin Surface 2** single point at the centroid of each kidney ** anterior and lateral surface Table 3 8. Organs and their corresponding number of dose points for a pelvic exam Pelvis exam Number of do se points Colon 13 Small Intestines 10 Bladder 1* Prostate 1* Gonads 2* Bone Marrow 6 Skin Surface 2** single point at the centroid of each organ ** anterior and lateral surface

PAGE 87

87 CHAPTER 4 RESULTS The results of the collection of organ doses in both the 50 th and 90 th percentile phantoms are provided below. Results are first broken out by CT scanner, with the results from the Siemens scanner being discussed first, followed by the Toshiba scanner. A comparison of doses between the two man ufacturers follows. For the purposes of comparison, only average and in field average organ doses are reported in this section. Tables of the individual doses to each of the specific organ point doses are provided in the appendix. Siemens Tube Current Modu lated Scans The organ dose data collected on the Siemens Somatom Sensation 16 slice scanner is provided below. Results are broken down by specific exam showing the tube current response, measured organ doses, and the difference in dose between the 50 th an d 90 th percentile phantoms. Chest Exam Modulation plot A plot of the tube current response to the two phantoms overlaid on a scout image of the 50 th percentile phantom, is provided in Figure 4 1, with the effective mAs per reconstructed slice along the y axis, and slice number along the x axis. These values were obtained by going through the reconstructed scan volume slice by slice and recording the effective mAs value for each image. It should be noted that this provides the average mAs value that went in to that reconstructed slice, and as such the graph does not illustrate any in plane changes in the tube current through the scan. Rather,

PAGE 88

88 the line depicts an average mAs along the phantom, smoothing out the in plane modulation changes. As expected, the mA s plots for both the 50 th and 90 th percentile phantoms remain fairly similar though the initial slices, as the additional adipose material of the add on is not very thick in the upper torso. Both plots increase through the shoulder region, where bony anato my and patient asymmetry between the anterior posterior and lateral projections push the tube current s higher in order to maintain image quality. Once past the shoulders, tube current drops off through the lung region as the low density tissue attenuates f ewer x ray photons and require s a lower tube output to maintain image quality. The tube responses to the 50 th and 90 th percentile phantoms diverge in the lower portion of the lung and through the remainder of the scan volume as the additional adipose tissu e in the 90 th percentile phantom increases, requiring higher tube output in order to maintain image quality. Organ doses A plot of the average organ doses with associated error bars for the standard deviation of point dose measurements, for both phantoms in the Siemens chest exam is shown in Figure 4 2. As expected from the shape of the modulation plot, the largest differences in dose between the two phantoms appears in organs in the lower portion of the phantom, with the liver seeing the largest increase in dose between the two phantoms. These are areas of the phantom where the additional adipose layer is substantially thicker, which in turn causes the ATCM system to increase the tube current in order to maintain image quality, which causes organ doses to rise. Conversely, organs such as the esophagus and breasts, show only slight increases in dose as there is not much difference between the two phantoms at the level of the dose

PAGE 89

89 points for those organs. It should be noted that the thyroid dose measurement p oint was just on the edge of the scan volume. This point was in the primary beam for the Siemens scans, but slight positioning changes can easily move it out of the primary beam and make a large impact on its dose, as will be seen in the Toshi ba measuremen ts. A graph of the percentage increase in doses seen in the 90 th percentile phantom is shown in Figure 4 3. Overall the larger phantom caused average chest exam doses to increase 10%, with the liver seeing the largest single increase, at 18.6%. In addition to averaging all the dose points for a given organ in order to determine average organ dose, an average of only dose points in the primary beam was done in For the Siemens chest exam, all dose points were located within the primary beam, so this additional data is not presented here, though the importance of the in field average organ dose will be evident in later sections. Abdominal Exam Modulation plot The plot of tube current response for the abdominal exam is shown in Figure 4 4. The scan range for the abdominal exam is from the lower dome of the diaphragm to the top of the iliac crest. With the exception of a slight dip in response to the air in the stomach, the tube current for the 50 th percentile ph antom remained fairly constant throughout the exam, while the tube current for the 90 th percentile phantom rose steadily in response to the increases in the additional adipose tissue on the larger phantom.

PAGE 90

90 Organ doses A plot of the average organ doses, wit h associated error bars for the standard deviation of point dose measurements, for both phantoms in the Siemens abdominal exam is shown in Figure 4 5 Again, doses are seen to increase in the 90 th percentile phantom for all organs in the exam, with the lar ger dose increases coming in organs such as the colon and skin measurements, which are located in the areas where the difference between the two tube current plots in Figure 4 4 are the greatest. Also of note is the large deviation of dose values in organs such as the colon and small intestines. This is a result of the method of point dose acquisition, where all dose points for large diffuse organs are collected for an exam, even though many individual points in this case were out of the primary beam This averaging in of multiple dose points outside of the primary beam artificially lowers the average organ dose and contributes to a large range of doses incurred by the specific organ. In order to address the issue of multiple out of field dose points for l arger organs, a separate in field average organ dose was calculated, in which only dose points in the primary beam were considered. A graph of the in field average organ doses for the abdominal exam is shown in Figure 4 6. The largest difference between F igures 4 5 and 4 6 is the increased in dose and decreased variation in those doses, to the colon and marrow, as multiple very low out of field doses were removed. Organs such as the liver and kidneys see no change between the two methods as all measured do se points are located within the primary scan area. A graph of the organ dose increases seen in the larger phantom is provided in Figure 4 7. Again, the largest increases in dose are seen in organs located in the regions of the tube current response where the 50 th and 90 th percentile phantoms most

PAGE 91

91 diverge. In the case of the abdominal exam, the kidneys, colon, bone marrow, and skin doses were seen to increase the most, with each increasing on the order of 30 40%. Overall, the larger phantom lead to an avera ge dose increase of 23% for measured organs Pelvic Exam Modulation plot A plot of the tube current response in the pelvic exam is shown in Figure 4 8. The tube current in the 50 th percentile phantom remains fairly stable, and is the highest through the b oney anatomy of the pelvis before dropping off through the groin area. The 90 th percentile phantom has significantly higher tube current through the boney area of the pelvis, and the tube current appears to be maxed out. The only way to test whether the tu be has maxed out during the 90 th % pelvic scan would be to change the maximum mAs level allowed by the system for the exam, which is not readily done from the exam protocol menu and requires behind the interface adjustments Raising the allowed mAs level a nd seeing the effect on the modulation plot in the scan would show whether the 90 th % scan had maxed out the tube for the given settings in current clinical protocol. The upper limit on mAs in the scanner is set to prevent patient overdoses in overly large patients. Organ doses A graph of the average organ doses in the pelvic exam for both phantoms is shown in Figure 4 9. Again doses are seen to increase in the larger phantom for all organs except the colon, which is a spurious result owing to slight variat ions in positioning of the phantom between the two phantom exams. Several dose points were on a slice at the edge of the scan field, and happened to be in field for the 50 th

PAGE 92

92 percentile phantom, and out of field for the 90 th Once only in field average orga n doses are taken into account, as shown in Figure 4 10 doses are seen to rise in the larger phantom for all organs. The largest increases in dose are seen in the marrow (65%), where most of the dose points are located in the upper portion of the pelvis, w here the disparity between the tube current modulations plots are the greatest. The skin dose was also seen to in increase by close to 30% in the larger phantom. Overall organ doses were seen to increase by 22% in the larger phantom as compared to the 50 t h percentile phantom for the pelvic exam. Toshiba Tube Current Modulated Scans The results for the scans performed on the Toshiba Aquilion ONE scanner are discussed below. For clarity of this section, only the exams utilizing the 64 slice (32 mm total bea m width) acquisition are discusse d as this is the beam width utilized for clinical exams. The effects of the multiple beam width s measured on ATCM exams will be further discussed in a later section. Chest E xam Modulation p lot A plot of the tube current response in the Toshiba scanner to the 50 th and 90 th percentile phantoms for the chest exam is shown in Figure 4 12. As seen in the Siemens scanner plot, the tube current in the 90 th percentile phantom continues to ride in response to the added attenuating material as compared to the 50 th percentile phantom. Unlike the Siemens scanner however, the Toshiba system does not seem to make adjustments for the lower density areas of the lung, and tube current also does not rise as much in response to the asymmetry of the shoulder region, making the Toshiba Sure Exposure system less dynamic in response to both phantoms.

PAGE 93

93 Organ doses A plot of the average organ doses from the chest exam in the Toshiba scanner is shown in Figure 4 13, with the average in field organ d oses shown in Figure 4 14. The general trend of higher organ doses in the larger phantom continues in the Toshiba scans. Figures 4 13 and 4 14 illustrate the role patient positioning can play in specific organ doses. In the average organ doses, both the th yroid and stomach doses were found to be higher in the 50 th percentile phantom. This is a result of all or portions of both organs being located along the edges of the scan field, where small changes in position can mean the difference between a specific d ose point being in or out of field. When only the in field organ doses are used, as in Figure 4 14, the stomach dose is seen to be higher in the larger phantom, as expected, but the thyroid dose remains lower. This is a result of the thyroid being only rep resented by a single dose point. Presumably if the thyroid dose point were entirely in the field for both exams, the dose values would be similar or slightly higher in the larger phantom, based on other trends seen in this work, and the tube current respon se plot of the two phantoms. This example also highlights the importance of clinical patient positioning, as one could potentially drastically reduce thyroid dose from CT imaging by taking special care to set exam boundaries so that it is not in the primar y beam, assuming of course that thyroid imaging is not an integral part of the scan. Figure 4 15 illustrates the increases in dose seen in the larger phantom for the chest exam on the Toshiba scanner. As previously mentioned, slight changes in positioning between the two phantoms resulted in a 12% dose reduction for the thyroid. Otherwise, doses were seen to increase by an overall average of 29%, with the stomach, liver and breasts seeing the largest increases, having increases of 44%, 40%

PAGE 94

94 and 58% respectiv ely. The esophagus was seen as having the smallest increase in dose, most likely attributable to its central location in the body, as compared to more peripheral organs such as the stomach and breast. Abdominal Exam Modulation plot The tube current respon se to the two phantoms in an abdominal exam on the Toshiba scanner is shown in Figure 4 16. Much as seen in the Siemens modulation plot, the response to the 50 th percentile phantom is relatively flat, which is to be expected as the outer body contour chang es very little throughout the scan range and the boney anatomy of the spine remains fairly constant. The response to the 90 th percentile phantom shows a steady increase to the increasing outer body contour of the larger phantom, eventually leveling off in response to the phantom size. Organ doses Graphs depicting the average and in field average organ doses from the Toshiba abdominal exam are shown in Figures 4 17 and 4 18. Again, larger doses are seen in the 90 th percentile phantom, and the largest increas es in doses correspond to organs located where the difference between the tube current outputs is the greatest. A plot of the dose increases for each in field organ between the two phantoms is shown in Figure 4 19. Overall doses increased on average 32%, w ith the largest increases seen in the colon (43%, marrow (37%) and skin (60%). The large increase in skin dose is a result of the much higher tube current s used in the larger phantom, coupled to the larger diameter of the phantom putting the skin surface c loser to the x ray tube.

PAGE 95

95 Pelvic Exam Modulation plot A graph of the tube current response to each phantom for a pelvic exam is shown in Figure 4 20. As has been the trend, the larger phantom produces a higher modulation curve as the system responds to th e larger phantom and tries to keep image quality constant. It is evident in the pelvic plot that the difference between the 50 th and 90 th percentile tube current responses is not nearly as drastic as it was in the Siemens scan, which has implication for th e dose measurements in the Toshiba pelvic exam. Organ doses A graph showing the average and in field organ doses in the Toshiba pelvis exam are show n in Figures 4 21 and 4 22. Overall doses were generally lower than those seen in the abdominal exam, and the trend of increased dose in the larger patient did not hold true. Even in the in field average organ doses, the doses to the colon and small intestines were seen to drop slightly in the larger phantom, while doses to the bladder and prostate were more o r less the same between the two phantoms. Doses to the skin, gonads, and marrow did increase in the larger phantom. This lowering of the dose in the large phantom is believed to be a result of the aforementioned tube current response plots. The differenc e between the tube current between the 50 th and 90 th percentile phantoms is not drastic, and as a result the added attenuation of the adipose add on causes doses to be the same or lowered for many organs in the larger phantom. This is especially evident in organs such as the colon and small intestines, where the majority of the dose points are surrounded by the boney area of the pelvis, which also contains the thickest areas of the adipose add on. Organs

PAGE 96

96 such as the gonads and marrow contain more inferior d ose points, where the adipose add on is not as thick, an d therefore not as attenuating. It is worth noting that the difference in tube current response between the two phantoms is more pronounced in the abdominal exam as compared to the pelvic exam. As a result the doses for the 90 th percentile phantom were found to be larger in the abdominal exam, while the pelvic exam saw some decreases in dose as compared to the 50 th percentile phantom This is noteworthy because the outer body contours of the phantom are very similar between the abdominal and pelvic sections, with the primary difference being the additional bone present in the pelvic region It is presumed that the presence of the pelvis in the scanogram image initiates some form of tube current limita tion to prevent patients from being over exposed, though this has yet to be confirmed. Further tests with phantoms with variable level of bone or a conversation with the programmers responsible for the tube current modulation algorithm on the Toshiba scann er would be needed in order to confirm this hypothesis. In order to test this theory, t he pelvis section of the phantom was scanned using an abdominal protocol to see if there was some difference in the pelvic protocol that was causing the tube current to be limited The modulation plots for the pelvic phantom section scanned using an abdominal protocol were very similar to that obtained with a pelvic exam lending credence to the theory that the limitation is innately built into the way the tube current i s preprogrammed based on the scanogram image. The pelvis section was also scanned using a chest protocol, which did produce a modulation plot with lower effective mAs values, but the chest scans utilizes difference reconstruction algorithms as compared to the abdomen and pelvis scans, which further complicates direct comparison.

PAGE 97

97 Scanner Comparison for Tube Current Modulated Scans Tube current Response A tube current response plot for both scanners during a chest exam is shown in Figure 4 24. For the chest e xam, the Siemens CARE Dose 4D system seems to be more dynamic in response to changes in the phantom, with a much larger increase in effective mAs through the shoulder region and drop off in the lungs before rising again though the mediastinum and torso. T he Toshiba Sure Exposure system maintains a much flatter response to both the 50 th and 90 th percentile phantoms. The average effective mAs for the Siemens scans were 195 and 247 mAs for the 50 th and 90 th percentile phantoms, which are both slightly higher than the average eff mAs of the Toshiba scans, which are 174 and 226 for both phantoms. A comparison of the average i n field doses in the chest exam from each scanner i s shown in Figure 4 2 5. Error bars have been removed from the comparison graphs for cla rity. For the majority of the organs, the Toshiba scanner has higher doses than the Siemens, with the exception of the thyroid and the lungs. The phantom positioning issue with respect to the thyroid has been discussed, and as a result the comparison betwe en scanners is not really a valid one. The lung doses for both the Siemens and Toshiba scanners are similar, likely a result of the higher and lower peaks of the Siemens modulation plot averaging out to be close to the flatter response of the Toshi ba scann er over the same range, though the beam spectrum for the two scanners are slightly different, with the Toshiba having an HVL of 6.02 mm at 120 kVp, and the Siemens having an HVL of 7.92 at the same energy. For the abdominal exam, the Toshiba and Siemens sc anners had similar outputs over the scan range, with the Toshiba having slightly higher eff mAs values for both

PAGE 98

98 phantoms. A plot of the tube current response is shown in Figure 4 26. The average effective mAs for each phantom in the Siemens scans was 181 a nd 267 mAs respectively, compared to 218 an d 331 mAs in the Toshiba scans. A comparison of average in field organ doses is provided in Figure 4 2 7 Again the Toshiba scanner produced higher doses for all organs. The tube output for the both scanners for t he pelvic exam is shown in Figure 4 28. As previously discussed the differences in response between the two phantoms was much greater in the Siemens exams as compared to the Toshiba. The average effective mAs values from the Siemens scans were 177 mAs and 270 mAs, as compared to 165 mAs and 207 mAs in the Toshiba. Unlike the chest exam, the Toshiba scanner has the more dynamic response to the pelvic section of the phantom. A comparison of average in field organ dose s for the two scanners is shown in Figure 4 29. There is no overall pattern to the organ doses as was present in the chest and abdominal exams. Doses for the colon, small intestines, bladder, prostate and gonads are similar for both scanners for both sized phantoms. The Toshiba scanner produced h igher skin and marrow doses for both sized phantoms. Image Quality Comparison In addition to comparing the measured doses and tube current responses of each ATCM system, image quality measurements were taken in each phantom using the image quality dummy p hantom section discussed in Chapter 3 As mentioned, only the standard deviation of HU values was examined as the low contrast and line pair modules were not visible when imaged with standard clinical protocols. The image quality phantom section was placed in the upper abdominal region of the phantom and both the 50 th and 90 th percentile phantoms were imaged using the image quality settings

PAGE 99

99 used for the clinical dose acquisition scans (quality reference mAs of 200 and 240 for the Siemens scanner, and an SD value of 10 for the Toshiba). The image quality data for both scanners is presented in Table 4 1 As seen in Table 4 1 the image quality, as measured by the standard deviation of Hounsfield units (HU) in the central portion of the Catphan image quality mo dule, for the Toshiba scanner was better than both settings used in the Siemens acquisitions. A lower HU standard deviation indicates lower variation in pixel values through a uniform region of phantom and lower image noise in the reconstructed images. In general, lower image noise allows for better low contrast detectability, but at the expense of higher patient dose The lower image noise seen in the Toshiba scanners coincides with the generally higher organ doses seen in the chest, abdominal, and pelvic exams on the Toshiba scan ner as compared to the Siemens. A direct correlation of image quality to mAs values between the two scanners was not seen, as the beam quality for each scanner are different. When comparing the change in image quality between the t wo phantoms for the same image quality setting on the same scanner, an increase in image noise in the 90 th % phantom is observed in both scanners. This decrease in image quality was similar in magnitude (a decrease of ~ 11%) for both quality reference mAs settings on the Siemens scanner, while the Toshiba scanner saw a dec rease in image quality of only 5.6% Based on these results, the Toshiba Sure Exposure system seems better at maintaining a constant image quality for a given quality setting than the Sie mens CARE Dose 4D system.

PAGE 100

100 Comparison to CTDI VOL and ImPACT Dose Estimates After collecting the library of specific organ doses for all exams on both scanners, a comparison between the measured doses was made to doses predicted by both CTDI VOL and the ImPA CT CT dose spreadsheet. For CTDI VOL value calculations, the average mAs value for each exam on each scanner was used along with the matching beam collimation and pitch for the exam. This CTDI VOL value was then used as an input in the ImPACT CT dose spreads heet, along with selecting the proper scanner, specific exam techniques, and scan range. At times selection of a matching scan range was exactly resemble a human patient. In such instances, skeletal markers were used to match the scan range to the anthropomorphic phantom as best as possible. As previously mentioned, the ImPACT spreadsheet does not have a specific model for the Toshiba Aquilion ONE scanner, but does have a mode l for an Aquilion 16 slice scanner with a 16 x 2 mm beam width configuration, which results in the same 32 mm beam width used for exams in this project. This scanner matching model was employed for the purposes of comparison in the work. The comparison to CTDI VOL and ImPACT predicted doses is broken down by exam and by phantom for the purposes of clarity. Doses are compared for the in field average organ doses for all organs except the bone marrow and skin For these two highly distributed organs, results a re presented for both the in field average doses, measured as previously described, and as whole body skin or marrow doses. The whole body skin and bone doses were calculated as these metrics are utilized in some risk estimate equations, and are the output of the ImPACT spreadsheet. This method allows the spreadsheet to calculate total effective doses for

PAGE 101

101 exams, but the measurements lose utility for predicting organ specific doses for the calculation of risk factors or estimation of deterministic effects. T he whole body skin dose was calculated by multiplying the measured average in field skin dose for each exam by the ratio of the irradiated skin area to the total body surface area for each exam and phantom. The associated areas were estimated using the com putational models from which the physical phantoms were based. The total skin area of the 50 th percentile phantom was found to be 1.9 m 2 while the 90 th percentile phantom was estimated at 2.08 m 2 The area of skin irradiated for the chest, abdomen and pel vic exams in the 50 th percentile phantom were estimated at 0.3198, 0.291, and .3397 square meters respectively. For the 90 th percentile phantom, the values were 0.362 for the chest, 0.323 for the abdomen, and 0.379 for the pelvis. The whole body marrow dos es were calculated by multiplying the measured doses at specific marrow sites by the mass fraction of red bone marrow at each location, and then summing the product for all marrow sites within the scan range. A table of the Mass fractions utilized for the calculations is provided in Table 4 2 Differences in measured dose and the CTDI or ImPACT dose for all exams were calculated using the following equation: (Measured Dose Estimated Dose) / (Estimated Dose) Chest Exam A comparison of measured and predic ted doses in the chest exam for the 50 th and 90 th percentile phantoms is provided in Figures 4 30 and 4 31. The CTDI VOL values for each scanner are provided on the far left in green, with the measured doses for each organ shown in orange and the ImPACT pre dicted doses shown in blue. In the 50 th percentile phantom, the Toshiba CTDI VOL value differed from the measured organ doses by an average of 88 %, with the breasts and in filed skin showing

PAGE 102

102 the highest deviation at 135 % and 125 % respectively. The Siemens C TDI VOL value did a better job of predicting organ doses and differed by an average of 9.8 %, with the largest differences seen in the breast ( 29 %) and in field bone marrow ( 28 %) It should be noted that the CTDI VOL value is just a single value describing average doses in a scan volume and is not necessarily designed to represent organ doses, though it has come to be used that way in clinical practices. For the ImPACT predicted doses in the 50 th percentile phantom, the results were mixed. On average, ImPAC T underestimated doses by 219 % in the Toshiba scans and by 194 % in the Siemens scans. In the Toshiba scan, ImPACT greatly underestimated the doses to the stomach ( 656 %), and liver ( 385 %), while overestimating lung and esophag eal doses by 23 % and 33 % respec tively. Similar results were seen in the Siemens scan, with vast underestimation of the stomach (658%), and liver (319%), but with a larger overestimation of esophageal doses ( 45 %). The large variations in doses for the liver and stomach are likely attribu table to organ placement in the MIRD vs the anthropomorphic phantom, with the majority of those organs seemingly outside the scan volume in the MIRD phantom when the scan range is set by boney anatomical markers. When comparing Overall, ImPACT does not see m to be a reliable predictor of organ doses for either scanner and at times produced much wider variations from measured doses than CTDI VOL When comparing predicted and measured skin doses, ImPACT overestimated the whole body dose but was within 25% for both scanners. It should be noted that the skin dose value presented in ImPACT is indeed a whole body average and should not be mistaken for a peak skin dose in the irradiated area. If it were to be interpreted this way,

PAGE 103

103 the ImPACT estimated doses would un derestimate peak skin dose by close to 400% for both scanners, so care must be taken to ensure a full understanding of how ImPACT, or any commercial dose calculator is calculating doses. The whole body marrow doses were underestimated in ImPACT by 62% on t he Toshiba scanner, and 23% on the Siemens. As with skin doses, if one were to mistakenly utilize the ImPACT marrow doses as peak dose to in field bone marrow, the underestimation would jump to 190% for the Toshiba scanner and 130% for the Siemens. The r esults of the comparison of measured and predicted doses in the 90 th percentile phantom are shown in Figure 4 31. Again the CTDI VOL value underestimated all doses in the Toshiba scanner, this time by an average of 88.7 % with the stomach, liver, skin, marro w and breast all being underestimated by over 100 %. The Siemens scanner overestimated all organ doses by an average of 22 %, with variations as high as 41 % in the breast and 37 % in the marrow, and variation as low as 5% for the skin. This again shows how th e CTDI VOL metric is not an accurate predictor of individual organ doses within a scan volume, as these doses will necessarily differ depending on their location within a patient and cannot be accurately modeled by a single average dose value. It is possibl e that an organ specific scaling factor could be used in order to better relate the CTDI VOL value to the specific placement of an organ within a patient. The ImPACT estimated doses in the 90 th percentile phantom showed similar results to those in the 50 t h with wide variability in the accuracy of predicted doses. For the Toshiba scanner, doses to the stomach (822%) and liver (477%) were again greatly underestimated. Meanwhile, doses to the lung were overestimated by 17 % and dose to

PAGE 104

104 the esophagus was over estimated by 38 %. In the Siemens scan, the overall doses also showed a high level of variability, with esophageal doses overestimated by 58 %, breast dose overestimated by 44 %, and stomach and liver dose underestimated by 500 % and 259% respectively Average skin dose was overestimated for both scanners, by 5% on the Toshiba and 40% on the Siemens, while in filed skin doses were again vastly underestimated by 445% and 250%. Whole body marrow doses were underestimated by 69% on the Toshiba, but overestimated by 12% on the Siemens scanner. Overall, neither the CTDI VOL nor the ImPACT spreadsheet do es a consistent job of predicting patent organ doses in the chest exam for either phantom. CTDI VOL values consistently over or underestimate d organ doses to varying ex tents, while t he ImPACT spreadsheet produce d wild variations in predicted doses, at times vastly overestimating some organs while vastly underestimating others. The ImPACT spreadsheet seems to do a better a job of estimating doses to organs that are entire ly within the scan volume, while at times being off by orders of magnitude for organs partially irradiated or at the edge of the scan volume. Even organs within the scan volume showed variability in upwards of 60 70% in the chest exam. Abdominal Exam A gra ph showing a comparison of measured doses and doses predicted by CTDI VOL and the ImPACT spreadsheet for an abdominal exam in the 50 th percentile phantom is shown in Figure 4 32. CTDI VOL underestimated measured doses in the Toshiba scanner by an average of 75 %, with the largest variation coming from the skin ( 124 %) and the smallest in the bone marrow ( 0.1 %). For the Siemens scan, CTDI VOL overestimated organ doses by an average of 23 %, with the bone marrow overestimated

PAGE 105

105 by 47 %, and the small intestines by 34 % The in field skin dose was underestimated by only 2.5% with the CTDI VOL metric for the Siemens scanner. For the ImPACT predicted doses in the Toshiba scanner, doses again varied greatly with colon dose underestimated by 152 %, and kidney doses overestimat ed by a 45% The small intestines dose was also underestimated by more than 100% Whole body marrow and skin doses matched to within 15%, while infield equivalents were underestimated by 133% and 467%. Similar varying results were seen in the Siemens scan, with colon and small intestines doses being underestimated by 201% and 139%,. K idney dose was overestimated by 25 %, while l iver and stomach doses both matched to within 10% for the Siemens scan Whole body marrow and skin doses matched to within 20% with in field doses being underestimated by 231% and 662% were the ImPACT results to be misinterpreted. Similar results were seen in the 90 th percentile phantom as well. CTDI VOL underestimated organ doses in the Toshiba scanner by an average of 53%, and overe stimated doses on the Siemens scanner by 35 % on average, with differences ranging from 2.7% in the skin to 53% in the in field marrow The ImPACT predictor under estimated colon and small intestines dose s on the Toshiba scanner by 149%, and 60 % respective ly. Doses to the liver, stomach, and kidneys were over estimated by 37%, 42%, and 56 %. Whole body marrow and skin doses matched to within 20%. On the Siemens scanner, dose to the colon and small intestines were similarly underestimated, while doses to the k idneys stomach and liver were overestimated by 37% 34 %, and 31 %. Whole body skin doses were under

PAGE 106

106 estimated by 10% while whole body marrow doses were over estimated by a similar amount. Again no clear pattern is observed in the discrepancies between ImPAC T predicted and measured organ doses that would make it possible to try to correct. Doses for some organs are grossly overestimated while doses to other organs located within the scan range are underestimated by a similar amount. Pelvic Exam A graph showi ng a comparison of measured organ doses to those predicted by CTDI VOL and the ImPACT spreadsheet for a pelvic exam in the 50 th percentile phantom is shown in Figure 4 34. The same general trend or lack of trend, seen in the previous two exams is again see n for the pelvic exam. The doses from the Toshiba scan are underestimated by CTDI VOL by an average of 76 %, with in field skin dose being underestimated by 182 %. For the Siemens scanner, CTDI VOL overestimates dose by an overall average of 13%, with the blad der, prostate, and in field marrow being overestimated by approximately 40 % each, and the colon dose being underestimated by 15 %. Examining the ImPACT predicted doses on the Toshiba scanner, doses to the bladder, prostate and gonads were all overestimate d by around 55 %, while doses to the Whole body marrow were underestimated by 93 %. Whole body skin dose was over estimated by 30%, while colon and small intestines doses matched to within 10%. For the Siemens scanner, similar results were seen with doses t o the bladder and prostate are overestimated by 46 % and 41 %, while doses to the colon and small intestines were underestimated by 38% and 22%. Whole body marrow doses were underestimated by 55% while whole body skin dose was underestimated by 24%.

PAGE 107

107 The resu lts of the comparison study in the 90 th percentile pelvic study are shown in Figure 4 35. Again, doses in the Toshiba scanner are underestimated by CTDI VOL this time by 60 % on average. Doses in the Siemens scanner were overestimated by CTDI VOL by an avera ge of 33%, with the bladder and prostate being overestimated by around 55% each, and the infield skin dose being within 7%. Examining the results of the ImPACT predicted results; once again wide variability is seen. On the Toshiba scanner, doses to the who le body marrow were under estimated by 93 %, while doses to all other organs were overestimated. T he gonads, prostate, and bladder are under estimated by 50% to 65% while d oses to the colon and small intestines were under estimated by 28 % and 35 % respectively Similar results are observed on the Siemens scanner, with whole body marrow doses underestimated by 92 %, and bladder and pro state doses overestimated by around 60 % each. Doses to the colon and small intestines were off by less than 10%. Conclusions of Co mparison Study On the whole, neither CTDI VOL nor the ImPACT spreadsheet is an accurat e or consistent predictor of patient organ doses for ATCM CT exams. CTDI VOL suffers in that it is a single average dose value measured in a homogenous circular phantom, a ttempting to represent organ values that will necessarily differ in a heterogeneous patient with varying dimensions. CTDI VOL also fails to account for ATCM utilized in modern CT scans, and any variation in patient size, further limiting it s utility as a do se descriptor. Again, CTDI VOL was not designed to be a dose descriptor, but is often used as such clinically. The ImPACT CT dose spreadsheet also suffers in accuracy. Its methods of determining skin and marrow doses are not useful for predicting the avera ge or

PAGE 108

108 maximum infield doses experienced by these tissues during a CT exam. Estimates of other organ doses vary wildly, even for similar organs within the same scan volume. Further, the method of matching scanners to the existing Monte Carlo dataset is susp ect, as the same organ simulated for each of the scanners often times showed drastically different predicted results, even when the measured doses were similar. An example of this can be seen in the 90 th percentile pelvic scan, where the gonad doses for ea ch scanner were similar (19.01 mGy on the Toshiba, and 19.41 mGy on the Siemens), but the ImPACT predicted doses for each scanner varied greatly (37 mGy on the Toshiba, and 27 mGy on the Siemens). The ImPACT spreadsheet also does not account for any differ ences in patient size or the use of ATCM which are likely contributing fa ctors to it s inability to accurately predict doses from these exams. Effects of Beam Width on Modulated CT Scans The effect that beam width has on doses in A TCM CT scans was investi gated on the Toshiba scanner. For each exam and each phantom, organ doses were recorded for additional beam widths of 16 mm (32 x 0.5 mm detector configuration) and 8 mm (16 x 0.5 detector configuration), in addition to the standard 32 mm (64 x 0.5 mm dete ctor configuration) utilized for clinical scans. In addition to the point dose measurements, several of the real time outputs of the dosimeters were saved as well in order to better visualize the effects of beam width on patient doses. Chest Exam A graph s howing the effects of beam width on individual in field, organ doses in each phantom in the chest exam is shown in Figure 4 36. Though not visible for every organ, a general trend of reduced doses for exams with wider beam widths is visible. When investi gated further organ doses were reduced by an average of 4.4 % from the

PAGE 109

109 16 slice to the 32 slice acquisitions in the 50 th percentile phantom, and then by an average of 10.6% further when going from 32 slice to 64 slice acquisition. Overall, changing acquisi tion from 16 to 64 slices resulted in an average dose reduction of d ose to 14% for all organs in the 50 th percentile chest scan. A similar result was seen in the 90 th percentile chest scan, where going from 16 slice to 64 slice acquisition resulted in an a verage reduction of organ dose of 8.4% Abdominal Exam A graph showing the effects of beam width on individual, in field organ doses for an abdominal scan is shown in Figure 4 37. A similar trend is seen in the 50 th percentile phantom measurements, with a s tepwise dose reduction for most organs when going from smaller to larger beam widths. T his trend is not seen in the 90 th percentile phantom however where the doses from the 32 slice acquisition are higher than either the 16 or 64 slice exams. The reason for this increase is not entirely known, as the general trend is observed in all other instances. In the 50 th percentile phantom, an average dose reduction of 9.5% is seen when changing from 16 to a 64 slice scan, while an increase in dose of 2.3% is seen for the same change in the 90 th percentile phantom. Again, it is not readily apparent why the trend is not seen in the 90 th percentile phantom, though it may have to do with phantom positioning during the exam, or the increased scatter from the larger phan tom. Pelvic Exam A graph for the same data in a pelvic exam is shown in Figure 4 38. The overall trend is once again visible in the 50 th percentile phantom, and in some but not all organs in the 90 th percentile phantom. In the 50 th percentile phantom an o verall dose reduction

PAGE 110

110 of 16.9% was seen when changing from 16 to 64 slice acquisition. A reduction of 11.2 % was seen for the same change in the 90 th percentile phantom. Surface Beam Profile Plots The FOCD system used to measure doses in this work allows f or real time imaging of the exposure to the individual dosimeter elements For most of this project, only the total integral dose from each point was recorded, but it is possible to break down these dose values into their instantaneous readouts to get a be tter picture of how the dose is delivered. A graph depicting the instantaneous dose of a surface dose point for each of the investigated beam widths is shown in Figure 4 39. In this plot, the instantaneous dose rate of each scan is plotted against the time it took to complete the scan. The total dose for each scan is also indicated above each curve, showing the previously discussed general trend of lowered doses for wider beams. From the graph, it is immediately apparent that wider beam widths require fewer revolutions of the tube, and less time to perform as compared to exams with thinner beams, as the 64 slice scan requires approximately 1/3 rd the time to complete as the 16 slice scan. This fact makes wider beam acquisitions more desirable for exams re quir ing patient breath holds, as a patient is more likely to be able to hold their breath and remain still for a shorter exam, reducing the possibility of motion blur introduced into the scan. The primary reason for the reduction of doses in wider beam widths is that because there are no perfect square waves, every beam has a profile and only a portion of this profile is captured by the detector and used in the creation of images. The scatter tails from each profile add to patient doses, but do not contribute t o image formation. In a wider beam, the percentage of the total profile that contributes to image formation is larger than in a smaller beam. This fact, coupled with the thinner beam requir ing more

PAGE 111

111 rotations about the patient than a wider beam results in t he larger doses seen in scans performed with smaller beams. This effect can be seen in Figure 4 40 in which the instantaneous dose rates from Figure 4 39 have been scaled against the scan distance as opposed to time. The higher peaks of the smaller beam w idth scan are clearly visible, which add up to contribute to the higher dose for this scan. Surface Variability Due to the helical nature of the CT scans, the specified pitch of the scan, and the random nature of the beam start angle for these scans, the a ctual point at which the beam passes by a small point dose can vary, and in turn lead to increased variability between measurements. This effect is averaged out over large scan volumes, but a single dose point can fall between peaks or troughs in exposure depending on how the helical beam path intersects the point. This helical dose distribution is illustrated in Figure 4 41, which is reproduced with permission from a 2009 Medical Physics paper by Marcel van Straten 46 The image depicts the relative dose distribution in a Rando phantom from an ATCM h elical CT scan, and shows how surface doses (the comb shaped distributions along the periphery of the phantom) can vary substantially over the course of a scan. These peaks and troughs in dose will shift along the patient z axis depending on the random sta rt angle of the x ray tube. The decrease in dose variability while moving from the periphery to the center of the patient is also visible, which is attributable to an increase in scatter radiation with depth in the phantom. This phenomenon of variable ab sorbed dose on the surface of a patient is illustrated in Figure 4 4 2 which shows the dose response from two consecutive scans of the same dose point ON THE Toshiba scanner having 11% variability between measurements. The total counts collected for each s can are indicated beneath the scan

PAGE 112

112 number. T he specific positioning of the fiber in relation to the helical path of the beam differs between the two scans as a result of the variable start angle of the x ray tube leading to a single, centered large peak i n the first scan, and an off centered larger peak, and smaller secondary peak in the second scan. This level of variability of point doses was found to be larger for dose points located closer to the surface of the phantom than for those located more cent rally. In order to characterize this trend, the coefficient of variation (COV) was calculated for the total number of fiber counts for each of the three scans taken at each point dose location during organ dose acquisition. The point dose locations were di vided into two groups for each exam: those points within 5 cm of the surface of the phantom, and those points located deeper than 5 cm. An average COV was calculated for each of these two groups in order to investigate the impact of point dose depth on the level of variation seen at a single dose point between successive scans. For the chest exam the average COV for surface points was found to be 6.7%, compared to 3.2% for those points located more centrally. A similar trend was seen in the pelvic exam, in which point doses within 5 cm of the surface produced an average COV of 6.1% compared to 2.9% for those deeper within the phantom. The abdominal exam showed an even larger difference between surface level and inner dose points, with the surface points sho wing an average COV of 13% and inner points again averaging to 3%. The results of the abdominal exam were skewed by a combination of the surface level dose variability and the z axis over ranging effect that will be discussed in the next section, with seve ral dose points being located right at the edge of the sc an volume and thus showing a

PAGE 113

113 wider variability ( on the order of 20 35%). This variability is due to the point dose being either entirely in or out of the primary beam depending on the beam start ang le The effects of this start angle position on point dose variability are being further investigated by another member of the research group, Chris Tien. Z axis Over ranging An additional consequence of variations in beam width in helical scanning is ex posure to organs outside the scan volume. In order to reconstruct images from a helical scan, exposures must be made outside of the selected scan volume. While this z axis over ranging is usually less than a full rotation added to the selected scan volume, as beam widths increase, the amount of extra volume tissue exposed also increases. The effects of z axis over ranging are shown in Figure 4 4 3 which shows the average dose delivered to both the colon and the small intestines in a single slice plane just outside the scan volume for the pelvic exam. Whereas overall average organ doses were generally seen to decrease as beam width increased, the doses to these specific point dose locations increased with beam width due to z axis over ranging. As beam widths increased, the primary beam came in closer proximity to these out of field points, causing an increase in scatter and a subsequent increase in dose. While these effects are usually minimal and often unavoidable, knowledge of z axis over ranging could pote ntially reduce dose to critical organs such as the thyroid in a chest exam. Fixed Tube Current Dose Measurement Results As previously mentioned, fixed tube current dose measurements were also collected on the Siemens scanner for all exams. The tube voltag e, beam width, and pitch used in the ATCM scans were duplicated in the fixed tube current exams. The tube current levels for each phantom and exam were set based on the average tube

PAGE 114

114 current utilized in each ATCM scan. As discussed in Chapter 3 the image q uality phantom section was used to match image quality for the FTC and ATCM scans. The mAs values derived from this method matched closely with the average mAs values from the ATCM scans for both phantoms for the abdominal and pelvic exams, but differed fo r the chest exam. The reason for the difference in the chest exam is the large variability seen in the ATCM response to the phantom though the shoulders and lung region. The spike in tube current through the shoulders raises the average mAs value for the A TCM scan significantly, while the image quality phantom section was images in the abdominal region which saw lower mAs values. This situation highlights the need for ATCM in CT exams, as a single fixed value will necessarily either suffer from increased i mage noise though the shoulde r region, or unnecessarily over irradiate the torso regions in order to maintain image quality through the chest. For the purposes of this study, all fixed tube currents were matched to the average mAs of the corresponding ATCM scan. This method was chosen due to the difficulties associated in matching image quality in the FTC scans to those in the ATCM scans, as image quality in the fixed scans changes depending on phantom anatomy. Because of these changes in phantom anatomy an d the resulting changes in image quality in FTC scans, selection of which section of the phantom to match image quality to would be an arbitrary decision that would affect dose levels. Setting mAs values to match image quality through the boney and asymmet rical shoulder region of a chest exam would result in higher doses, and better image quality, in the torso region. While matching image quality to the torso region would result in lower doses and lower image quality in slices through the shoulders. Because of these tradeoffs, it was decided that matching to the

PAGE 115

115 average eff ective mAs value from the ATCM scan would be the best method for comparing dose values. Chest exam A reproduction of the ATCM response plot is shown in Figure 4 4 4 with the two fixed tub e current lines superimposed. A plot of the average in field organ doses for both the modulated and fixed exams in each phantom is shown in Figure 4 4 5 In general, doses were seen to increase in both phantoms in the fixed tube current scans as compared to the modulated scans. The average dose increase in the FTC scan was 27% for the 50 th percentile phantom and 11% for the 90 th percentile phantom. In particular, dose increases on the order of 30% were seen in the thyroid, which is expected in a FTC scan, as the thin region of the neck where the thyroid is located is imaged with the same current used to image the thicker regions of the chest. The CARE Dose 4D system allows the tube current to be reduced through the thinner regions. Breast doses were also cons iderably higher in the FTC scans for both phantoms. This increase is due to the location of the breast dose points, which lie in the trough in the effective mAs values though the lung region seen in Figure 4 4 4 The ATCM lowers the tube current through thi s region in response to lower attenuation though the lungs, resulting in lower tube currents than the overall average which was used for the FTC scan. Additionally, the in plane modulation likely contributed to the lowered breas t dose in the ATCM scan, as the anterior posterior current is typically lower than the lateral current due to patient dimensions. As expected, the organs with the largest change in dose from fixed to ATCM scans were located in regions where the effective mAs values in Figure 4 4 4 de viated the most. Exceptions to this observation were seen in organs like the lungs, where

PAGE 116

116 portions of the organ in the ATCM scan saw higher doses than the FTC scan, while other portions saw lower doses. This range of doses was effectively canceled out in the averaging of dose points to get the average organ dose, though the FTC scan doses were higher in both the 50 th and 90 th percentile phantoms. Abdominal exam The modulation plot of the abdominal exam with FTC lines added is shown in Figure 4 4 6 and a co mparison of fixed and modulated doses is shown in Figure 4 4 7 As in the chest exam, in general, doses were seen to increase in the FTC scan as compared to the ATCM scan. Both the 50 th and 90 th percentile phantoms saw an average increase in in field organ doses of around 33% with the majority of individual organs seeing dose increases along those lines The largest increases in dose was seen in the liver and stomach of the 90 th percentile phantom (72% and 61% increases respectively), which coincide with t he larger dip in the eff mAs values seen in Figure 4 4 6 around slice # 5. The larger drop off in the 90 th percentile phantom lead to the largest discrepancy between the fixed and ATCM mAs values, and thus the largest increase in dose seen between the two e xams. Outside of the aforementioned dip in eff mAs values, the tube current for both exams, and the 50 th percentile phantom especially, remained fairly constant in the ATCM scans. Because of this, it is assumed that the majority of the dose reduction seen in the ATCM scan as compared to the FTC scan is a result of the in plane modulation, which can act to lower doses to organs in AP projections, though these changes are not visible in the modulation plots.

PAGE 117

117 Pelvic exam The modulation plot for the pelvic ex am with the FTC overlay is provided in Figure 4 4 8 and a comparison of ATCM and FTC average in field organ doses is provided in Figure 4 4 9 The pelvic exam exhibited several interesting changes in organ doses, and no overall trend was visible. This lack of overall trend is explained by the high degree in variation of the tube current response to each phantom seen in Figure 4 4 8 Tube current is greatly increased though the boney regions of the pelvis, which is also the location of the largest portion of t he adipose add on in the 90 th percentile phantom, and then drops off significantly in the lower region of the phantoms. This modulation of the tube current allows for adequate and consistent image quality, whereas the constant mAs of the FTC scan would nec essarily suffer from fluctuations in image quality throughout the scan Because of this fluctuation in tube current organs located primarily in the upper pelvis region, such as the colon and small intestine, saw higher doses in the ATCM scans in both phan toms, while doses in the lower portion of the phantom, such as the gonads and prostate, saw higher doses from the FTC procedure. In the 50 th percentile phantom, doses to the colon were decreased by 45% in the FTC scan as compared to the ATCM scan, with a similar reduction of 40% was seen in the small intestines. Similar results were seen in the 90 th percentile phantom with reductions of 38% and 31% for the two organs respectively. Conversely, in doses to the bladder, prostate, and gonads were found to incr ease by 11%, 18%, and 18% in the FTC scan in the 50 th percentile phantom. These increases were significantly higher in the 90 th percentile phantom as the bladder dose increased 57%, the prostate dose 52%,

PAGE 118

118 and the gonad dose by 81%. These results are easily explained by noting the differences in tube current between the fixed and modulated scans in Figure 4 4 8 Skin doses were found to be slightly higher for the ATCM scan for both phantoms, though skin doses were measured at a single point (approximately sl ice # 20 in Figure 4 4 8 ), and doses to the skin would be expected to vary in these scans depending on where in the phantom they were collected. Overall, change in dose between fixed and modulated MDCT procedures can be estimated based on the differences i n tube output through similar phantom or patient sections. Areas that contain larger differences in patient attenuation and asymmetry will show larger fluctuations in tube current throughout a scan range, and as such will show a larger degree of dose discr epancy compared to a fixed tube current scan. It follows that such exams would also show a larger variation in image quality from slice to slice through the scan, which is the initial problem that ATCM technologies set out to address. Specific doses from the fixed tube current scans to each individual point dose location are provided in Appendix A.

PAGE 119

119 Figure 4 1. Plot of tube current response for chest exam Figure 4 2. Average organ doses for Siemens chest exam 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 60 70 Effective mAs per Slice Slice # Chest Exam Siemens 50th% 90th% 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00 28.00 30.00 Dose (mGy) Average Organ Doses in Chest Exam Siemens 50th % 90th %

PAGE 120

120 Figure 4 3. Dose increases in larger pha ntom for average organ doses in the Siemens chest exam Figure 4 4. Plot of tube current response in abdominal exam 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0% 20.0% Thyroid Lung Stomach Liver Breast Esophagus Marrow Skin Percent Increase in Dose Dose Increase in Larger Phantom 0 50 100 150 200 250 300 350 0 10 20 30 40 50 Effective mAs per Slice Slice # Abdominal Exam Siemens 50th% 90th%

PAGE 121

121 Figure 4 5. Average organ doses for Siemens abdominal exam Figure 4 6. Average in field organ doses for Siemens abdominal exam 0.00 5.00 10.00 15.00 20.00 25.00 30.00 Liver Stomach Kidneys Colon Small Intestine Marrow Skin Dose (mGy) Average Organ Doses in Abdominal Exam Siemens 50th % 90th % 0.00 5.00 10.00 15.00 20.00 25.00 30.00 Liver Stomach Kidneys Colon Small Intestine Marrow Skin Dose (mGy) Average In Field Organ Doses in Abdominal Exam Siemens 50th % 90th %

PAGE 122

122 Figure 4 7. Dose increases in larger phantom for in field average organ doses in the abdominal exam Figure 4 8. Plot of tube current response in pelvic exam 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0% 45.0% Liver Stomach Kidneys Colon Small Intestine Marrow Skin Percent Increase in Dose Dose Increase in Larger Phantom (In Field Doses Only) 0 50 100 150 200 250 300 350 0 10 20 30 40 50 Effective mAs per Slice Slice # Pelvic Exam Siemens 50th% 90th%

PAGE 123

123 Figure 4 9. Average organ doses for Siemens pelvic exam Figure 4 10. Average in field organ doses f or Siemens pelvic exam 0.00 5.00 10.00 15.00 20.00 25.00 30.00 Colon Small intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) Average Organ Doses in Pelvic Exam Siemens 50th % 90th % 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Colon Small Intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) Average In Field Organ Doses in Pelvic Exam Siemens 50th % 90th %

PAGE 124

124 Figure 4 11. Dose increases in larger phantom for in field average organ doses in the pelvic exam Figure 4 12. Plot of tube current response in Toshiba chest exam 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% Colon Small intestine Bladder Prostate Gonads Skin Marrow Percent Increase in Dose Dose Increase in Larger Phantom Siemens (In Field Doses Only) 0 50 100 150 200 250 0 10 20 30 40 50 60 70 Effective mAs per Slice Slice # Chest Exam Toshiba 50th% 90th%

PAGE 125

125 Figure 4 13. Average organ doses for Toshiba chest exam Figure 4 14. Average in field organ doses for Toshiba chest exam 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 Dose (mGy) Average Organ Doses in Chest Exam Toshiba 50th % 90th % 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 Dose (mGy) Average In Field Organ Doses in Chest Exam Toshiba 50th % 90th %

PAGE 126

126 Figure 4 15. Dose increases in larger phantom for in field average organ doses in the Toshiba chest exam Figure 4 16. Plot of tube current response in Toshiba abdominal exam 20.0% 10.0% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% Thyroid Lung Stomach Liver Breast Esophagus Marrow Skin Percent Increase in Dose Dose Increase in Larger Phantom (In Field Doses Only) 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 Effective mAs per Slice Slice # Abdominal Exam Toshiba 50th% 90th%

PAGE 127

127 Figure 4 17. Av erage organ doses for Toshiba abdominal exam Figure 4 1 8 Average in field organ doses for Toshiba abdominal exam 0.00 10.00 20.00 30.00 40.00 50.00 60.00 Liver Stomach Kidneys Colon Small Intestine Marrow Skin Dose (mGy) Average Organ Doses in Abdominal Exam Toshiba 50th % 90th % 0.00 10.00 20.00 30.00 40.00 50.00 60.00 Liver Stomach Kidneys Colon Small Intestine Marrow Skin Dose (mGy) Average In Field Organ Doses in Abdominal Exam Toshiba 50th % 90th %

PAGE 128

128 Figure 4 19. Dose increases in larger phantom for in field average organ doses in the Toshiba abdominal exam Figure 4 20. Plot of tube current response in Toshiba pelvic exam 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% Liver Stomach Kidneys Colon Small Intestine Marrow Skin Percent Increase in Dose Dose Increase in Larger Phantom (In Field Doses Only) 0 50 100 150 200 250 300 350 0 10 20 30 40 50 Effective mAs per Slice Slice # Pelvic Exam Toshiba 50th% 90th%

PAGE 129

129 Figure 4 21. Average organ doses for Toshiba pelvic exam Figure 4 22 Average in field organ doses for Toshiba pelvic exam 0.00 10.00 20.00 30.00 40.00 50.00 60.00 Colon Small Intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) Average Organ Doses in Pelvic Exam Toshiba 50th % 90th % 0.00 10.00 20.00 30.00 40.00 50.00 Colon Small Intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) Average In Field Organ Doses in Pelvic Exam Toshiba 50th % 90th %

PAGE 130

130 Figure 4 23. Dose increases in larger phantom for in field average organ doses in the Toshiba pelvic exam Figure 4 24. Comparison of tube current response in chest exam 15.0% 5.0% 5.0% 15.0% 25.0% 35.0% 45.0% 55.0% Colon Small intestine Bladder Prostate Gonads Skin Marrow Percent Increase in Dose Dose Increase in Larger Phantom Toshiba (In Field Doses Only) 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 60 70 Effective mAs per Slice Slice # Chest Exam Comparison 50th% Toshiba 90th% Toshiba 50th% Siemens 90th% Siemens

PAGE 131

131 Figure 4 25. Average in field organ doses for both Toshiba and Siemens chest exams 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 Dose (mGy) In Field Organ Dose Comparison Chest Exam 50th% Toshiba 50th% Siemens 90th% Toshiba 90th% Siemens

PAGE 132

132 Figure 4 26. Comparison of tube current response in abdominal exam Figure 4 27. Average in field organ doses for both Toshiba and Siemens abdominal exams 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 Effective mAs per Slice Slice # Abdominal Exam Comparison 50th% Toshiba 90th% Toshiba 50th% Siemens 90th% Siemens 0.00 10.00 20.00 30.00 40.00 50.00 60.00 Liver Stomach Kidneys Colon Small Intestine Marrow Skin Dose (mGy) In Field Organ Dose Comparison Abdominal Exam 50th% Toshiba 50th% Siemens 90th% Toshiba 90th% Siemens

PAGE 133

133 Figure 4 28. Comparison of tube current response in pelvic exam Figure 4 29. Average in field organ doses for both Toshiba and Siemens pelvic exams 0 50 100 150 200 250 300 350 0 10 20 30 40 50 Effective mAs per Slice Slice # Pelvic Exam Comparison 50th% Toshiba 90th% Toshiba 50th % Siemens 90th% Siemens 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 Colon Small Intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) In Field Organ Dose Comparison Pelvic Exam 50th% Toshiba 50th% Siemens 90th% Toshiba 90th% Siemens

PAGE 134

134 Table 4 1. Scanner image q uality comparison 50th % Phantom 90th % Phantom Scanner Image Quality Setting HU Std Dev. HU Std Dev. Siemens qual ref mAs 200 14.21 15.8 11.2% Siemens qual ref mAs 240 13.24 14.8 11.8% Toshiba SD 10 11.8 12.46 5.6% Table 4 2. Mass fraction of red bone marrow used in whole body marrow calculations Location Mass Fraction Clavicles 0.011 Scapula 0.091 Ribs 0.102 Sternum 0.026 Thoracic vertebrae 0.13 Lumbar vertebrae 0.13 Os coxae 0.265 Sacrum 0.081 Femur proximal 0.045

PAGE 135

135 Figure 4 30. Comparison of measured and predicted chest exam doses in 50 th percentile phantom 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Dose (mGy) Chest Exam Comparison to CTDI and ImPACT Doses 50th % 50th% Toshiba 50th% Siemens ImPACT Toshiba ImPACT Siemens CTDIvol Siemens CTDIvol Toshiba

PAGE 136

136 Figure 4 31. Comparison of measured and predicted chest exam doses in 90 th percentile phantom 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 Dose (mGy) Chest Exam Comparison to CTDI and ImPACT Doses 90th % 50th% Toshiba 50th% Siemens ImPACT Toshiba ImPACT Siemens CTDIvol Toshiba CTDIvol Siemens

PAGE 137

137 Figure 4 32. Comparison of measured and predicted abdominal exam doses in 50 th percentile phantom 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 Dose (mGy) Adominal Exam Comparison to CTDI and ImPACT Doses 50th % 50th% Toshiba 50th% Siemens ImPACT Toshiba ImPACT Siemens CTDIvol Toshiba CTDIvol Siemens

PAGE 138

138 Figure 4 33. Comparison of measured and predicted abdominal exam doses in 90 th percentile phantom 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 Dose (mGy) Adominal Exam Comparison to CTDI and ImPACT Doses 90th % 50th% Toshiba 50th% Siemens ImPACT Toshiba ImPACT Siemens CTDIvol Toshiba CTDIvol Siemens

PAGE 139

139 Figure 4 34. Comparison of measured and predicted pelvic exam doses in 50 th percentile phantom 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Dose (mGy) Pelvic Exam Comparison to CTDI and ImPACT Doses 50th % 50th% Toshiba 50th% Siemens ImPACT Toshiba ImPACT Siemens CTDIvol Toshiba CTDIvol Siemens

PAGE 140

140 Figure 4 35. Comparison of measured and predicted pelvic exam doses in 90 th percentile phantom 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 Dose (mGy) Pelvic Exam Comparison to CTDI and ImPACT Doses 90th % 90th% Toshiba 90th% Siemens ImPACT Toshiba ImPACT Siemens CTDIvol Toshiba CTDIvol Siemens

PAGE 141

141 Figure 4 36. Effects of beam width on organ doses in chest exam Figure 4 37. Effects of beam width on organ doses in abdominal exam 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 Dose (mGy) Beam Width Comparison Chest Exam 50th% 16 slice 50th% 32 slice 50th% 64 slice 90th% 16 slice 90th% 32 slice 90th% 64 slice 0.00 10.00 20.00 30.00 40.00 50.00 60.00 Liver Stomach Kidneys Colon Small Intestine Marrow Skin Dose (mGy) Beam Width Comparison Abdominal Exam 50th% 16 slice 50th% 32 slice 50th% 64 slice 90th% 16 slice 90th% 32 slice 90th% 64 slice

PAGE 142

142 Figure 4 38. Eff ects of beam width on organ doses in pelvic exam Figure 4 3 9 Plot of instantaneous dose rate versus time for various beam widths 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 Colon Small Intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) Beam Width Comparison Pelvic Exam 50th% 16 slice 50th% 32 slice 50th% 64 slice 90th% 16 slice 90th% 32 slice 90th% 64 slice 0.0 1.0 2.0 3.0 4.0 0 5 10 15 20 25 Dose (mGy) Time (sec) Beam Width vs Instantaneous Dose 64 slice 32 slice 16 slice 21.9 mGy 24.6 mGy 29.2 mGy

PAGE 143

143 Figure 4 40 Dose rates for various beam widths scaled according to scan length 0.00 0.05 0.10 0.15 0.20 0.25 0 50 100 150 200 Dose Rate (mGy/mm) Scan Length (mm) Instantaneous Dose Rate 64 slice 32 slice 16 slice 24.6 mGy 29.2 mGy 21.9 mGy

PAGE 144

144 Figure 4 4 1 Monte Carlo simulatio n of the average in plane dose distribution for an ATCM scan of an anthropomorphic phantom. [Reprinted with permission from modulations on the estimation of organ and effective doses in x ray computed 36, 4881 -4889 (2009) (page 4887, Figure 5c)]

PAGE 145

145 Figure 4 42. Point dose variability due to beam start angle Figure 4 4 3 Effects of z axis over ranging on out of field point doses 0 50000 100000 150000 200000 250000 0 2 4 6 8 Counts Time (s) Start Angle Variability scan 1 scan 2 763k 676k 11% variability 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 16 slice 32 slice 64 slice Dose (mGy) Z axis Over ranging in Helical CT Colon Small Intestine

PAGE 146

146 Figure 4 4 4 Tube current modu lation plot with fixed current lines for chest exam Figure 4 4 5 Comparison of fixed and modulated doses in the Siemens chest exam 0 50 100 150 200 250 300 350 400 0 10 20 30 40 50 60 70 Effective mAs per Slice Slice # Chest Exam Siemens 50th% 90th% 50th% fixed 90th% fixed 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Dose (mGy) In Field Fixed Organ Dose Comparison Chest Exam 50th% Fixed 50th% TCM 90th% Fixed 90th% TCM

PAGE 147

147 Figure 4 4 6 Plot of tube current response and fixed tube currents in Siemens abdominal exam Figure 4 4 7 Doses from fixed and modulated abdominal scan 0 50 100 150 200 250 300 350 0 10 20 30 40 50 Effective mAs per Slice Slice # Abdominal Exam Siemens 50th% 90th% 50th% fixed 90th% fixed 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Liver Stomach Kidneys Colon Small intestine Marrow Skin Dose (mGy) In Field Fixed Organ Dose Comparison Abdominal Exam 50th% Fixed 50th% TCM 90th% Fixed 90th% TCM

PAGE 148

148 Figure 4 4 8 Tube current plot of ATCM and FTC scans in a Siemens pelvic exam Figure 4 4 9 Comparison of pelvic organ doses in FTC and ATCM scans. 0 50 100 150 200 250 300 350 0 10 20 30 40 50 Effective mAs per Slice Slice # Pelvic Exam Siemens 50th% 90th% 50th% fixed 90th% fixed 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 Colon Small Intestine Bladder Prostate Gonads Skin Marrow Dose (mGy) In Field Fixed Organ Dose Comparison Pelvic Exam 50th% Fixed 50th% TCM 90th% Fixed 90th% TCM

PAGE 149

149 CHAPTER 5 CREATION OF PA T IENT SIZE DEPENDANT CT DOSE ESTIMATION CALC ULATOR Need For Patient Size Dependent Dose Estimation It was shown in Chapter 4 that there can be great variation in organ doses for ATCM CT exams. Doses can vary both between organs in a single exam, and for the same organ for patients of different sizes for exams with the same image quality setting. It was also shown that neither CTDI VOL nor the commercially available ImPACT CT dose spreadsheet do a particularly good job of estimating doses from these exams. It is because of this variation and lack of an effective tool to estimate doses that this work was undertaken. Methodology Utilizing a method similar to that of Angel et al. 31 each measured in field average organ dose was plotted against the patient perimeter for the particular phantom section. Patient perimeter was utilized in the Angel study because the exams were retrospective and based on acquired patient scan data for which other patient size metrics such as weight or BMI were not known. This methodology was continued for this study for the same purposes. If retrospectively trying to assess patient organ doses from a CT scan, often no add itional information on patient size is available outside of the actual scan images, from which patient perimeter can be easily obtained. Further, patient perimeter is most likely a more exact dimension describing patient morphology compared to patient weig ht, as patients having the same weight could have drastically differing outer body dimensions. These outer body dimensions can also differ greatly in different anatomical regions for patients of the same weight.

PAGE 150

150 For the chest exam, the perimeter was measu red at the level of the nipple in order to compare the results of this study to those published by Angel et al. This was slightly problematic in that the portion of the upper arm is still in the plane of the nipple. Measurements were taken around the entir e outer perimeter of the phantom (including arms) as well as just around the torso (sans arms) for both the 50 th and 90 th percentile phantoms at approximately slice 270 in the phantom. For the abdomen and pelvis exams, the perimeter measurements were taken at the level of the top of the iliac crest for both (approximately slice 205 in the phantom). Once the plots were created for the dose and patient perimeter data, a fit line was created between the two in order to create an equation in which one could in put patient perimeter for an exam and output an estimated patient organ dose from a n ATCM scan of that region. This process was repeated for both the Toshiba and Siemens scanners. Limitations and Validation of Method The primary limitation of this method f or estimating patient organ doses is that only two phantom sizes were available. A line fit between two points can only hint at the correlation and it is not enough for any sort of statistical analysis of the robustness of the method. Ideally average organ dose measurements could be taken for a variety of phantom sizes, but the time and monetary investment required to build an anthropomorphic phantom is great, along with the extensive time required to collect all the average organ dose measurements, which c an take up to 8 hours per exam, per phantom. In order to test the validity of the method, the average in field organ dose measurements from the lung and breast were plotted against the data from the Angel study, which to date is the only study in which the relationship between patient size and

PAGE 151

151 average organ dose has been investigated for ATCM CT scans. In the Angel paper, segmented computational phantoms were created from the CT scans of 32 women undergoing a chest CT on a Siemens Somatom Sensation 16 slice scanner with CARE Dose 4D tube current modulation technology. Average doses to the lung and breast were calculated using Monte Carlo methods, using the actual modulation curve saved from each exam to weight the tube current. The average doses of the lung and breast were then plotted against patient perimeter (measured at the level of the nipple) in order to create a scatter plot of data points. A linear function was then fit to this data in order to show the correlation between patient size and dose to the se two organs, the results of which are reproduced in Figure 5 1. The linear fit lines for both organs are shown in the lower right portion of F igure 5 1 along with their R 2 values, which were both found to show significant linear correlation (p <0.001 f or both). s dose points for both phantoms for the lung and breast were added to the Angel data to see if it fell in line with the larger range of patient data. As previously mentione d, the arms of the anthropomorphic phantom are in place at the level of the nipple, which differs from Measurement of patient perimeter was taken both including and exc luding the arms in order to see how the two methods affected the data fit. A graph of F igure 5 1 with the additional data points from this study added is shown in Figure 5 2 Both the lung and breast doses fit into the existing data well for perimeter mea surements taken both with and without arms. This despite the fact that phantom breast dose was measured as single point doses on the surface of the skin of a phantom with no actual breasts.

PAGE 152

152 Though both sets of phantom data fit into the existing data, for t he purpose of the chest exam it was decided to use the patient perimeter measurement with the arms, as that data made the lung results from this study (shown in Figure 5 2 as open circles) fit better with the Angel data. This choice was made because it is assumed that the lung measurements in this study are more comparable to what was done in the Monte Carlo Angel study. Additionally, the presence of the arms affects the tube current output during the scan. This discrepancy was only an issue for the chest exam, as there were no arms in the abdominal or pelvic studies. A linear fit line created between the two lung dose points produced an equation of organ dose (mGy) = 0. 1264 *(perimeter (cm))+ 5.34 which is very similar to that generated by the range of dose s and patient sizes in the Angel study Because the equation generated by this study was similar to that generated in the literature for more than 30 patients on the same CT scanner it was seen as validation of the method to create dose fit equations for the remainder of the organs in this study. It is anticipated that significant errors and discrepancies could arise from the use of only two data points to create these equations, but this work is seen as a first step towards the development of a more thoro ugh and robust system based on a wide range of patients. An additional limitation of this work is that the estimator is valid only for the image quality settings at which the input data was collected as higher image quality settings would lead to higher patient doses In an institution where these image quality settings remain fixed regardless of patient size, this would not be a problem, but these settings often vary between institutions depending on radiologist preference. The image quality settings can also vary at a single institution based on patient size, where exams

PAGE 153

153 performed on extremely large patients utilize a lower image quality setting in order to reduce patient doses, at the expense of noisier images. In order to provide a robust patient dose estimator, a large range of input data for a variety of image quality settings on different ATCM systems would be needed. CT Dose Estimation Calculator Using the collected data as inputs, linear fits were created in an excel spreadsheet for each organ and each exam on both the Siemens and Toshiba scanner. A clean interface was created which allows the user to input the patient perimeter for a particular exam, at which point dose values are generated for both scanners. An additional input was added for a CT DI VOL based scanner correction factor. A recent Medical Physics paper by Turner et al. 47 showed the utility of using CTDI VOL as a normalization factor to account for differences in output of various CT scanners. Fo r the purpose of this CT dose estimator, the user inputs the CTDI VOL per 100mAs for the appropriate pitch and collimation of their particular exam, and the spreadsheet scales the estimated doses to account for differences in the specific output of the use used to collect the organ dose input data. A screenshot of the created CT dose estimation calculator is shown in Figure 5 3. Future Work to Improve Accuracy and Robustness of Method As previously mentioned, this created dose calculator is fairly limited in its applicability as all calculations are based on only two dose measurements in two different sized patients. As such, it does a very good job of reproducing the doses measured in this work, but the accuracy of estimat ion s in other patients could vary drastically. In order to produce a more comprehensive dose calculator for ATCM CT scans, a large number of average organ doses would need to be collected from a range

PAGE 154

154 of patients. This type of work is not feasible for physica l phantom measurements due to the extensive time and cost involved in phantom fabrication and is better suited to computational simulations such as those done in the Angel paper. A wide range of computational phantoms covering a range of heights and weight s have been created by the ALRADS group at the University of Florida that would be well suited to this endeavor, though a method of simulating the tube current response would need to be developed as such a prospective method has yet to be created. Barring the use of computational models, an extension of the work of Angel et al. could be undertaken using segmented CT scan data from actual patient scans but investigating more organs over a variety of CT exams. Both of these methods tend to be very labor inte nsive but would result in a large amount of data from which to base the linear fit lines used to generate patient size specific organ doses. A large enough data set would also allow for a better idea of the range of doses to be expected from variable organ size and positioning, which is something that actual segmented patient scans would be better at addressing than computational models, in which organ locations are fixed for a specific sized patient. Overall, this project has demonstrated a methodology th at can be followed in order to collect a range of average organ doses for ATCM CT scans which can be used in order to better estimate patient doses from such procedures.

PAGE 155

155 Figure 5 1. Scatter plot of lung and breast doses for 32 women undergoing ATCM ches t Organs During Routine Chest CT: Effects of Tube current AJR: 193 1340 1345 (2009) (page 1343, Figure 3)] y = 0.1099 x + 5.9561 R = 0.3235 y = 0.1344x 0.3123 R = 0.4594 5 7 9 11 13 15 17 19 21 23 25 80 90 100 110 120 130 140 Organ Dose (mGy) Patient Perimeter (cm) Patient Perimeter vs Average Organ Dose lung breast Linear (lung) Linear (breast)

PAGE 156

156 Figure 5 2. Reproduction of F igure 5 1 with dose points from this work added y = 0.1099 x + 5.9561 R = 0.3235 y = 0.1344x 0.3123 R = 0.4594 5 7 9 11 13 15 17 19 21 23 25 80 90 100 110 120 130 140 Organ Dose (mGy) Patient Perimeter (cm) Patient Perimeter vs Average Organ Dose lung breast phantom lung w/o arms phantom lung w/ arms phantom breast w/o arms phantom breast w/ arms Linear (lung) Linear (breast)

PAGE 157

157 Figure 5 3. Screenshot of the created CT dose calculator

PAGE 158

158 CHAPTER 6 CONCLUSION Results of This Work The overall goal of this work was to develop a better method for estimating patient specific organ doses for tube current modulated MD CT exams This was necessary as current dose metrics and spreadsheet based methods for dose estimation are based on outdated CT technology and do not account for modern advances including wider beams, helical acquisition, and tube current modulated scans. Additionally, none of the current methods for dose estimation take patient size into account, which can drastically affect delivered doses from CT procedures. In order to accomplish this task, several tools were developed, including the previously develop ed FOCD system for dose measurements and the 50 th percentile anthropomorphic phantom In order to test the effects of patient size on organ doses, an adipose tissue equivalent substitute material was developed and used to fabricate a phantom add on that wo uld create a 90 th percentile by weight phantom. These tools were utilized to collect a library of specific average organ doses in both sized phantoms for a range of common CT exams on both a Siemens and a Toshiba CT scanner employing different tube current modulation schemes. The results of the organ dose modulation systems to variations in patient size. Additionally, the effects of beam width on patient doses were investigated an d average dose reductions on the order of 10% were found when using a 32 mm beam as compared to an 8 mm beam. The collected organ doses from all exams, both patients, and both scanners were compared to both CTDI VOL and a commercially available spreadsheet based CT dose

PAGE 159

159 estimator and neither were found to be a reliable indicator of organ doses in ATCM CT scans. Organ doses collected from this work were plotted against patient perimeter and were found to be in line with recently published literature on the t opic for breast and lung doses Using this correlation as validation, a patient size dependent CT dose estimator spreadsheet was created that takes both patient perimeter as well as scanner output factors, in the form of CTDI VOL in to account in order to a ttempt to more accurately estimate patient doses from modulated CT procedures. Potential Future Work As mentioned in Chapter 5, the current CT dose estimation spreadsheet is fairly rudimentary and based only on two patients. Additional data would make the estimator more robust and applicable to a wider range of patients, scans, and scanners. This data could be collected either through the development of additional physical phantom add ons to create alternate patient sizes, or through Monte Carlo simulation s of patient scans. These simulations could take the form of retrospective dose reconstructions done on segmented CT data from actual patient scans, as was done in the Angel study, or could take the form of prospective dose measurements in computational ph antoms of a variety of heights and weights. The only factor currently limiting prospective phantom dose measurements from being a reality is a computational model of specific tube current modulation systems. As seen from the comparison of the Sure Exposure and CARE Dose 4D systems in Chapter 4, such a computational model would have to be specific to a manufacturer, as even two similar approaches to the same problem can produce drastically different outcomes in terms of tube current response to anthropomorph ic phantoms.

PAGE 160

160 If additional physical phantom add ons were to be developed to collect physical dose measurements, the development of a FOCD system that could collect more dose points in a single scan would greatly speed up the data acquisition phase of the process, which is currently tedious and time intensive. A CCD based FOCD system is currently in development within the research group which would accomplish this and allow for more phantom dose measurements to be collected. Additionally, further research into the effects of CT tube start angles on point dose variability, which was briefly touched on in Chapter 4, could be investigated in order to better understand the variability and error associated with point doses in helical CT scanning. This informati on would also be helpful for understanding the errors inherent in the system that may not manifest in Monte Carlo simulations of average organ doses. Final Thoughts The research presented in this dissertation provides a method to quantif y the dose from a r ange of ATCM MDCT procedures in order to better predict patient size dependent doses for these procedures. Accurate dose prediction for CT imaging is becoming increasingly important, as several states are moving towards requiring dose information to be inc (California has recently passed legislation of this nature) as accurate and efficient as possible without being overly labor intensive. It is hoped that the tools developed in this study can be the beginning of a large collection of patient organ doses, leading to a robust dose estimation tool for clinical use

PAGE 161

161 APPENDIX A INDIVIDUAL POINT DOS E MEASUREMENTS Table A 1. Individual point dose measurements for chest exam s Chest Exam 50th Percentile Phantom 90th Percentile Phantom Siemens 16 Siemens Fixed Toshiba 64 Siemens 16 Siemens Fixed Toshiba 64 Organ Subsection Slice Dose (mGy) Dose (mGy) Dose (mGy) Dose (mGy) Dose (mGy) Dose (mGy) Thyroid Center 29 6.00 26.42 15.16 17.30 25.90 22.85 15.22 Lung Left 291.00 20.50 11.97 14.92 23.33 15.77 19.69 Right 291.00 20.53 11.04 13.60 26.30 17.63 18.55 Left 280.00 26.50 15.31 18.64 18.57 16.34 28.56 Right 280.00 19.76 12.59 17.02 24.11 20.62 24.91 Left 26 0.00 17.02 13.76 23.40 18.68 14.79 24.16 Right 260.00 17.96 15.91 27.30 19.60 16.30 29.04 Stomach Superior 251.00 26.06 11.19 24.08 19.34 13.49 30.44 Center Anterior 243.00 16.21 13.54 20.57 19.20 13.07 9.34 Center Posterior 243.00 18.37 14.14 1 8.86 20.42 13.35 10.34 Inferior 235.00 15.18 9.20 13.71 25.01 16.97 4.93 Liver Superior Anterior 251.00 17.16 14.01 22.36 17.36 13.48 30.63 Superior Posterior 251.00 18.69 11.82 22.06 22.11 14.49 29.39 Center 243.00 15.62 12.53 19.66 20.34 15.39 16.32 Inferior 235.00 17.40 9.45 16.48 21.85 15.80 5.46 Breast Left 270.00 15.71 13.32 29.08 13.73 15.83 40.39 Right 270.00 14.71 14.31 24.04 14.86 17.27 43.27 Esophagus Superior 300.00 19.01 13.22 15.69 17.50 15.28 16.94 Center 291.00 18.69 11.14 16.59 16.09 19.71 20.46 Center 280.00 20.47 11.25 21.13 17.49 15.52 25.24 Center 270.00 13.49 11.40 19.32 14.27 14.00 24.02 Center 260.00 15.24 10.90 21.68 15.57 12.86 19.02 Inferior 251.00 14.50 12.59 17.83 15.81 12.40 17.48 Marrow Clavicles 291 .00 17.33 9.85 21.83 27.10 17.33 27.25 291.00 15.77 7.16 16.40 22.67 14.27 28.64 Scapula 280.00 16.26 8.77 18.00 13.92 12.28 25.67 280.00 15.06 8.85 16.00 12.87 10.81 23.28 Ribs 270.00 12.75 9.61 25.70 12.56 11.93 24.47 270.00 14.03 11.91 22.9 4 15.78 16.44 28.57 Sternum 270.00 13.98 13.79 22.96 12.64 14.23 42.67 Thoracic vertebrae 270.00 10.85 8.75 16.15 11.58 11.59 19.02 Skin Anterior 251.00 25.19 16.51 25.55 22.46 17.80 41.29 Lateral 251.00 19.96 12.69 25.29 26.50 16.41 25.34

PAGE 162

162 Tabl e A 2. Individual point doses for abdominal exam s Abdominal Exam 50th Percentile Phantom 90th Percentile Phantom Siemens 16 Siemens Fixed Toshiba 64 Siemens 16 Siemens Fixed Toshiba 64 Organ Subsection Slice Dose (mGy) Dose (mGy) Dose (mGy) Dose (m Gy) Dose (mGy) Dose (mGy) Liver Superior Anterior 251.00 14.43 24.22 27.35 18.94 25.67 29.12 Superior Posterior 251.00 15.59 27.77 27.32 13.09 16.22 27.62 Center 243.00 14.48 24.92 24.28 17.00 19.11 34.49 Inferior 235.00 17.78 27.02 29.53 16.84 17.00 44.20 Stomach Superior 251.00 17.76 24.12 21.03 16.00 19.13 32.58 Center Anterior 243.00 14.55 24.41 31.95 16.40 17.90 37.08 Center Posterior 243.00 15.35 22.15 29.81 15.98 18.30 29.89 Inferior 235.00 14.65 25.01 25.34 17.84 18.18 34.26 K idneys Left 229.00 12.91 21.13 23.78 16.61 15.67 31.25 Right 229.00 13.88 22.46 22.15 17.29 15.72 25.83 Colon Left 229.00 15.99 23.86 31.23 21.45 17.18 40.17 Right 229.00 16.88 31.10 29.06 24.54 20.27 39.55 Left 221.00 12.27 22.41 30.50 19.54 16.58 41.25 Right 221.00 13.60 24.62 28.99 21.45 17.43 45.74 Transverse 221.00 16.55 29.06 27.17 17.62 18.28 44.08 Left 213.00 12.39 18.01 20.52 16.90 11.80 32.77 Right 213.00 15.80 23.37 26.84 22.53 15.46 35.18 Descending 187.00 0.54 0.87 0.69 0.25 0. 38 1.37 Small I ntestine Left 221.00 13.78 23.07 27.06 17.25 14.93 31.44 Right 221.00 11.63 19.21 18.31 14.20 11.16 21.09 Left 213.00 11.84 18.85 24.00 12.36 9.14 27.39 Right 213.00 11.50 18.36 16.99 12.41 9.74 20.90 Left 205.00 8.21 14.26 6.87 9.2 5 2.94 11.54 Right 205.00 5.60 8.75 6.15 6.60 3.04 7.58 Marrow Lumbar vertebrae 221.00 9.93 16.25 14.19 12.91 10.79 19.50 Os coxae 205.00 2.59 4.29 5.73 3.91 2.67 11.24 205.00 3.40 5.14 8.14 3.95 2.91 5.97 Sacrum 205.00 3.02 5.04 3.59 2.51 2.07 6 .54 Skin Anterior 229.00 21.31 31.26 37.08 26.02 22.56 54.18 Lateral 229.00 16.78 26.80 26.46 27.32 18.84 47.74

PAGE 163

163 Table A 3. Individual point doses for pelvic exams Pelvic Exam 50th Percentile Phantom 90th Percentile Phantom Siemens 16 Sie mens Fixed Toshiba 64 Siemens 16 Siemens Fixed Toshiba 64 Organ Subsection Slice Dose (mGy) Dose (mGy) Dose (mGy) Dose (mGy) Dose (mGy) Dose (mGy) Colon Left 229 2.07 2.05 3.25 2.43 2.60 2.66 Right 229 1.75 1.59 2.46 2.61 2.39 2.80 Left 221 4.56 3.51 10.79 5.74 5.02 7.45 Right 221 4.14 3.30 5.99 14.86 4.91 5.68 Transverse 221 3.92 3.64 7.79 14.87 11.70 17.16 Left 213 11.24 11.29 24.58 17.51 14.34 21.29 Right 213 15.17 12.16 25.11 22.19 16.86 34.14 Left 205 35.42 26.59 31.73 25.61 Right 2 05 22.82 25.88 27.54 25.60 Descending 199 28.02 23.01 33.59 27.00 Descending 199 20.58 20.98 22.43 21.95 Posterior 194 21.96 17.38 17.23 13.19 Descending 187 10.44 11.02 14.43 9.25 13.36 14.99 Small Intestine Left 221 3.89 3.71 7.71 14.23 13.85 6.54 Right 221 3.66 3.52 5.56 11.36 9.29 6.68 Left 213 11.24 10.40 21.05 18.75 15.79 15.67 Right 213 10.36 9.50 15.76 17.24 13.35 16.81 Left 205 13.78 12.19 27.55 18.55 16.42 22.58 Right 205 14.35 12.18 26.96 21.95 17.93 22.46 Center 199 24.36 19.86 30.22 25.87 Right 199 20.92 22.35 26.20 25.09 Left 194 29.79 22.47 18.03 15.66 Right 194 20.98 21.39 16.40 11.25 Bladder Center 187 10.19 11.28 14.92 12.26 19.30 16.50 Prostate Center 177 11.29 13.33 13.99 12.57 19.14 13.95 G onads Left 165 18.90 22.10 13.94 18.19 33.67 20.09 Right 165 18.80 22.35 12.24 20.63 36.68 17.93 Skin Anterior 187 34.72 21.42 30.95 44.41 26.49 45.79 Lateral 187 27.92 14.87 29.64 38.74 20.16 28.71 Marrow Lumbar vertebrae 221 2.69 2.29 3.50 9.13 8.8 1 4.20 Os coxae 205 12.41 11.69 25.68 23.56 18.95 32.84 205 12.99 13.01 23.77 24.87 19.02 25.45 Sacrum 205 6.81 6.64 12.27 16.03 15.91 17.28 Femur proximal 182 11.04 11.64 12.06 10.83 15.84 22.22 182 9.32 10.95 11.35 11.35 14.74 19.25

PAGE 164

164 LIST OF REFERENCES 1. M. A. Pozniak, D. J. Balison, F. T. Lee, Jr., R. H. Tambeaux, D. T. Uehling and T. D. Moon, "CT angiography of potential renal transplant donors," Radiographics 18 565 587 (1998). 2. J. Rydberg, K. A. Buckwalter, K. S. Caldemeyer, M. D. Phillips, D. J. Conces, A. M. Aisen, S. A. Persohn and K. K. Kopecky, "Mutlisection CT: Scanning Techniques and Clinical Applications," Radiographics 20 1787 1806 (2000). 3. D. J. Brenner and E. J. Hall, "Computed Tomography An Incre asing Source of Radiation Exposure," The New England Journal of Medicine, 2277 2284 (2007). 4. E. S. Amis, P. F. Butler, K. E. Applegate, S. B. Birnbaum, L. F. Bratemen and J. M. Hevezi, "American College of Radiology White Paper on Radiation Dose in Medi cine," (2007). 5. NCRP, "Ionizing Radiation Exposure of the Population of the United States," NCRP Report 160 (2009). 6. C. McCollogh, D. Cody, S. Edyvean, R. Geise, B. Gould, N. Keat, W. Huda, P. Judy, W. Kalender, M. McNitt Gray, R. Morin, T. Payne, S Stern and L. Rothenberg, Report No. 96, 2008. 7. D. J. Brenner, "Is it time to retire the CTDI for CT quality assurance and dose optimization?," Med Phys 32 3225 3226 (2005). 8. D. J. Brenner and C. McCollough, "Is it time to retire the computed tomog raphy dose index (CTDI) for quality assurance and dose optimization," Med Phys 33 1189 1191 (2006). 9. R. Dixon, "A new look at CT dose measurement: Beyond CTDI," Medical Physics 30 1272 1280 (2003). 10. R. Dixon, "Restructuring CT dosimetry A realis tic strategy for the future requiem for the pencil ion chamber," Med Phys 33 3973 3976 (2006). 11. R. Dixon and A. Ballard, "Experiemental validation of a versatile system of CT dosimetry using a conventional ion chamber: Beyond CTDI 100 ," Med Phys 34 33 99 3413 (2007). 12. R. Dixon, M. T. Munley and E. Bayram, "An improved analytical model for CT dose simulation with a new look at the theory of CT dose," Med Phys 32 3712 2738 (2005).

PAGE 165

165 13. K. Perisinakis, A. Tzedakis and J. Damilakis, "On the use of Monte Carlo derived dosimetric data in the estimation of patient dose from CT examinations," Med Phys 35 2018 2028 (2008). 14. F. D. Pazik, R. J. Staton, J. L. Williams, M. M. Arreola, D. E. Hintenlang and W. E. Bolch, "Organ and effective doses in newborns a nd infants undergoing voiding cystourethrograms (VCUG): a comparison of stylized and tomographic phantoms," Med Phys 34 294 306 (2007). 15. R. J. Staton, F. D. Pazik, J. C. Nipper, J. L. Williams and W. E. Bolch, "A comparison of newborn stylized and tom ographic models for dose assessment in paediatric radiology," Phys Med Biol 48 805 820 (2003). 16. R. J. Staton, J. L. Williams, M. M. Arreola, D. E. Hintenlang and W. E. Bolch, "Organ and effective doses in infants undergoing upper gastrointestinal (UGI ) fluoroscopic examination," Med Phys 34 703 710 (2007). 17. A. M. Groves, K. E. Owen, H. M. Courtney, S. J. Yates, K. E. Goldstone, G. M. Blake and A. K. Dixon, "16 detector multislice CT: dosimetry estimation by TLD measurement compared with Monte Carl o simulation," Br J Radiol 77 662 665 (2004). 18. P. Deak, M. van Straten, P. C. Shrimpton, M. Zankl and W. A. Kalender, "Validation of a Monte Carlo tool for patient specific dose simulations in multi slice computed tomography," Eur Radiol 18 759 772 ( 2008). 19. F. K. Korley, J. C. Pham and T. D. Kirsch, "Use of advanced radiology during visits to US emergency departments for injury related conditions, 1998 2007," JAMA 304 1465 1471. 20. H. Greess, J. Lutze, A. Nomayr, H. Wolf, T. Hothorn, W. A. Kale nder and W. Bautz, "Dose reduction in subsecond multislice spiral CT examination of children by online tube current modulation," Eur Radiol 14 995 999 (2004). 21. M. K. Kalra, M. M. Maher, T. L. Toth, B. Schmidt, B. L. Westerman, H. T. Morgan and S. Sain i, "Techniques and applications of automatic tube current modulation for CT," Radiology 233 649 657 (2004). 22. M. Gies, W. A. Kalender, H. Wolf and C. Suess, "Dose reduction in CT by anatomically adapted tube current modulation. I. Simulation studies," Med Phys 26 2235 2247 (1999). 23. N. Keat, 2005. 24. J. N. Althen, "Automatic tube current modulation in CT -a comparison between different solutions," Radiat Prot Dosimetry 114 308 312 (2005).

PAGE 166

166 25. M. K. Kalra, M. M. Maher, T. L. Toth, R. S. Kamath, E. F. Halpern and S. Saini, "Comparison of Z axis automatic tube current modulation technique with fixed tube current CT scanning of abdomen and pelvis," Radiology 232 347 353 (2004). 26. M. K. Kalra, M. M. Maher, R. V. D'Souza, S. Rizzo, E. F. Halpern, M. A. Blake and S. Saini, "Detection of urinary tract stones at low radiation dose CT with z axis automatic tube current modulation: phantom and clinical studies," Radiology 235 523 529 (2005). 27. M. K. Kalra, S. Rizzo, M. M. Maher, E. F. Halpern, T. L. T oth, J. A. Shepard and S. L. Aquino, "Chest CT performed with z axis modulation: scanning protocol and radiation dose," Radiology 237 303 308 (2005). 28. J. J. DeMarco, C. H. Cagnon, D. D. Cody, D. M. Stevens, C. H. McCollough, J. O'Daniel and M. F. McNi tt Gray, "A Monte Carlo based method to estimate radiation dose from multidetector CT (MDCT): cylindrical and anthropomorphic phantoms," Phys Med Biol 50 3989 4004 (2005). 29. J. J. DeMarco, C. H. Cagnon, D. D. Cody, D. M. Stevens, C. H. McCollough, M. Z ankl, E. Angel and M. F. McNitt Gray, "Estimating radiation doses from multidetector CT using Monte Carlo simulations: effects of different size voxelized patient models on magnitudes of organ and effective dose," Phys Med Biol 52 2583 2597 (2007). 30. E Angel, N. Yaghmai, C. M. Jude, J. J. Demarco, C. H. Cagnon, J. G. Goldin, A. N. Primak, D. M. Stevens, D. D. Cody, C. H. McCollough and M. F. McNitt Gray, "Monte Carlo simulations to assess the effects of tube current modulation on breast dose for multid etector CT," Phys Med Biol 54 497 512 (2009). 31. E. Angel, "Dose to Radiosensitive Organs During Routine Chest CT: Effects of Tube Current Modulation," AJR 193 1340 1345 (2009). 32. D. M. Bakalyar, "A critical look at the numerical coefficients in CTD I vol," Med Phys 33 2003 (2006). 33. R. Dixon, 2010. 34. D. E. Hyer, R. F. Fisher and D. E. Hintenlang, "Characterization of a water equivalent fiber optic coupled dosimeter for use in diagnostic radiology," Med Phys 36 1711 1716 (2009). 35. J. Winslo w, D. E. Hyer, R. F. Fisher, C. Tien and D. E. Hintenlang, "Construction of anthropomorphic phantoms for use in dosimetry studies" Journal of Applied Clinical Medical Physics 10 195 204 (2009).

PAGE 167

167 36. J. L. Hurtado, C. Lee, D. Lodwick, T. Goede, J. L. Willia ms and W. E. Bolch, "Hybrid computational phantoms representing the 50th percentile male and female adult: construction and applications for retrospective dosimetry," Health Phys (2009). 37. A. K. Jones, D. E. Hintenlang and W. E. Bolch, "Tissue equivalen t materials for construction of tomographic dosimetry phantoms in pediatric radiology," Med Phys 30 2072 2081 (2003). 38. International Commission on Radiological Protection, "2007 recommendations of the International Commission on Radiological Protectio n," ICRP Publication 103, Annals of the ICRP 37 1 332 (2007). 39. T. Yoshizumi, T. Nakamura, M. Yamane, A. H. Islam, M. Menju, K. Yamasaki, T. Arai, K. Kotani, T. Funahashi, S. Yamashita and Y. Matsuzawa, "Abdominal fat: standardized technique for measur ement at CT," Radiology 211 283 286 (1999). 40. S. W. P. Johnson, M. Wayson, B. Juneja, C. Lee, W. Bolch, "Hybrid Patient Dependent Phantoms Covering Statistical Distributions of Body Morphometry in the U.S. Adulst and Pediatric Population," P IEEE 97 2 060 2075 (2009). 41. K. M. Bushby, T. Cole, J. N. Matthews and J. A. Goodship, "Centiles for adult head circumference," Arch Dis Child 67 1286 1287 (1992). 42. U. S. Radiology, "On line radiology protocols," (UF & Shands Department of Radiology, Gaines ville Fl, 2008). 43. G. L. de la Grandmaison, I. Clairand and M. Durigon, "Organ weight in 684 adult autopsies: new tables for a Caucasoid population," Forensic Sci Int 119 149 154 (2001). 44. C. J. Martin, "Effective dose: how should it be applied to m edical exposures?," British Journal of Radiology 80 639 647 (2007). 45. T. C. Gerber, J. J. Carr, A. Andrew E and R. L. Dixon, "Ionizing Radiation in Cardiac Imaging.," Journal of the American Heart Association (2009). 46. M. van Straten, P. Deak, P. C. Shrimpton and W. A. Kalender, "The effect of angular and longitudinal tube current modulations on the estimation of organ and effective doses in x ray computed tomography," Med Phys 36 4881 4889 (2009). 47. A. C. Turner, M. Zankl, J. J. DeMarco, C. H. C agnon, D. Zhang, E. Angel, D. D. Cody, D. M. Stevens, C. H. McCollough and M. F. McNitt Gray, "The feasibility of a scanner independent technique to estimate organ dose from MDCT scans: using CTDIvol to account for differences between scanners," Med Phys 3 7 1816 1825.

PAGE 168

168 BIOGRAPHICAL SKETCH Ryan was born in Florence, South Carolina to Don and Casey Fisher. He is one of three children, along with older sister Megan, and younger brother Reid. He spent his formative years in the suburbs of Atlanta and attended Brookwood High School in Snellville, GA, where everybody somebody. Ryan attended college at the Georgia engineering in 2004. He then moved to Gainesville, FL to attend the University of nuclear engineering science s in 2006. His studies continued as he pursued a doctorate in nuclear engineering sciences in 2010. Ryan was heavily involved in collegiate cycling throughout his studies at both Ge orgia Tech and Florida.