<%BANNER%>

Hybrid Computational Phantoms of the 1, 5, and 10 Year Old Male and Female Reference Individuals

Permanent Link: http://ufdc.ufl.edu/UFE0022173/00001

Material Information

Title: Hybrid Computational Phantoms of the 1, 5, and 10 Year Old Male and Female Reference Individuals
Physical Description: 1 online resource (45 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: dosimetry, nurbs, pediatric, phantom, radiation
Nuclear and Radiological Engineering -- Dissertations, Academic -- UF
Genre: Nuclear Engineering Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This research has and will continue to bring international recognition to the University of Florida as a new standard in the modeling of the human body for radiation dose calculations. The improved anatomical description of these models will yield more realistic calculation of the radiation dose received by individuals undergoing medical irradiation (general x-ray imaging, fluoroscopy, CT, radiation therapy, nuclear medicine). Furthermore, the flexibility of these phantoms will also allow for their use in other areas, such as radiation protection and homeland security applications.
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.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Bolch, Wesley E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022173:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022173/00001

Material Information

Title: Hybrid Computational Phantoms of the 1, 5, and 10 Year Old Male and Female Reference Individuals
Physical Description: 1 online resource (45 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: dosimetry, nurbs, pediatric, phantom, radiation
Nuclear and Radiological Engineering -- Dissertations, Academic -- UF
Genre: Nuclear Engineering Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This research has and will continue to bring international recognition to the University of Florida as a new standard in the modeling of the human body for radiation dose calculations. The improved anatomical description of these models will yield more realistic calculation of the radiation dose received by individuals undergoing medical irradiation (general x-ray imaging, fluoroscopy, CT, radiation therapy, nuclear medicine). Furthermore, the flexibility of these phantoms will also allow for their use in other areas, such as radiation protection and homeland security applications.
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.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Bolch, Wesley E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022173:00001


This item has the following downloads:


Full Text

PAGE 1

HYBRID COMPUTATIONAL PHANTOMS OF THE 1, 5, AND 10 YEAR OLD MALE AND FEMALE REFERENCE INDIVIDUALS By DANIEL LEE LODWICK A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2008 Daniel Lee Lodwick 1

PAGE 3

To my Lord and Savior, Jesus Christ, through whom I am able to accomplish great things. 2

PAGE 4

ACKNOWLEDGMENTS I thank my research advisor, Dr. Wesley E. Bolch, for his guidance through these last few years and his positive encouragement. I also than k Dr. Choonsik Lee for all of his hard work in helping me to complete this set of phantoms and for introducing me to much of the software that was required to complete this work. I would also like to ac knowledge Dr. Choonik Lee for his segmentation of the 5 year voxel phantom. Finall y, I thank the National Cancer Institute for supporting this research under grant RO1 CA116743 (subcontract by Rensselaer Polytechnic Institute). 3

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 3 LIST OF TABLES ...........................................................................................................................5 LIST OF FIGURES .........................................................................................................................6 LIST OF ABBREVIATIONS ......................................................................................................... .7 ABSTRACT ...................................................................................................................... ...............8 CHAPTER 1 HISTORY OF PHANT OM DEVELOPMENT ......................................................................10 Introduction .................................................................................................................. ...........10 Stylized Phantoms ..................................................................................................................10 Voxel Phantoms ......................................................................................................................11 Hybrid Phantoms ....................................................................................................................12 2 DESCRIPTION AND DEVELOPM ENT OF HYBRID PHANTOMS .................................14 Computed Tomography Data Collection ................................................................................14 Image Segmentation ............................................................................................................ ...15 Polygon Modeling ..................................................................................................................16 Non-Uniform Rational Basis Spline Modeling ......................................................................17 Skeletal Modeling ............................................................................................................17 Alimentary Tract Modeling .............................................................................................18 Standardization of Phantoms ..................................................................................................1 8 Anthropometric Data Matching .......................................................................................19 Organ Mass Matching .....................................................................................................20 Creation of Complimentary Gender ................................................................................22 Voxelization .................................................................................................................. ..........22 3 CONCLUSIONS ................................................................................................................. ...40 LIST OF REFERENCES ...............................................................................................................42 BIOGRAPHICAL SKETCH .........................................................................................................44 4

PAGE 6

LIST OF TABLES Table page 2-1. Descriptions of CT image data sources. ............................................................................312-2. Anthropometric reference values for th e alimentary tract of 1, 5, and 10 year individuals ..........................................................................................................................322-3. Anthropometric reference values of 1, 5, and 10 year old individuals ..............................332-4. For 1 year male and female: list of organ masses for hybrid and voxel phantoms. ...........342-5. For 5 year male and female: list of organ masses for hybrid and voxel phantoms ............362-6. For 10 year male and female: list of organ masses for hybrid and voxel phantoms ..........38 5

PAGE 7

LIST OF FIGURES Figure page 2-1. Original and resegmented cervi cal spine of 1 year old phantom .......................................242-2. Ribcage of 1 year old phantom ..........................................................................................252-3. Cranial resegmentation. ................................................................................................. ....262-4. Results of cranial resegmentation. .....................................................................................2 72-5. The 1 year male and female pha ntoms: anterior and lateral views ....................................282-6. The 5 year male and female pha ntoms: anterior and lateral views ....................................292-7. The 10 year male and female phant oms: anterior and lateral views. .................................30 6

PAGE 8

LIST OF ABBREVIATIONS CAP Chest/abdomen/pelvis CT Computed tomography GI Gastrointestinal HU Hounsfield Unit ICRP International Commi ssion on Radiation Protection ICRU International Commission on Ra diation Units and Measurements IRB Institutional research board MR Magnetic resonance NCAT NURBS-based cardiac torso NURBS Non-uniform rational basis spline UFH University of Florida hybrid 7

PAGE 9

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science HYBRID COMPUTATIONAL PHANTOMS OF THE 1, 5, AND 10 YEAR OLD MALE AND FEMALE REFERENCE INDIVIDUALS By Daniel Lee Lodwick May 2008 Chair: Wesley E. Bolch Major: Nuclear Engineering Sciences Traditionally, computation of radiation dose to individuals was calculated using a host of stylized phantoms. These phantoms were easily used and manipulated, but lacked anatomical realism of organ shape and position. More re cently, tomographic phantom construction was made possible through advances in computer technology. The expl icit segmentation of CT or MRI images led to this phantom series that ma intained anatomical realism, but lacked the versatility of stylized phantoms. A new class of phantom has been introduced that maintains anatomical realism and allows for non-uniform deformation of organs. This hybrid phantom utilizes both NURBS and polygon mesh surfaces to describe the complex geometries of internal organ shapes. This new technology was used in the effort to create hybrid phantoms of the 1, 5, and 10 year male and female phantoms. After select ion of the CT data, the hybrid phantoms were created through the four steps, which are (1) image segmenta tion, (2) polygon mesh modeling, (3) NURBS modeling, and (4) stan dardization. Standardizing of each phantom was performed by matching anthropometric data, alimentary tract length data, and ICRP 89 reference organ mass data. The targeted tolerance for anthropometric and alimentary data was 5%, whereas the targeted tolerance for matching of organ mass was only 1%. With the exception of the thymus 8

PAGE 10

9 for the 1 year old phantom, all organs were ma tched in the hybrid phantoms within 1%. All anthropometric data were matched within 5% as well as the alimentary tract lengths (except for the left and right colon of 1 year old phantom). It is important to note that in the voxelized phantom, skin mass was not matched w ithin 1% for any of the phantoms. Development of the 1, 5, and 10 year UFH phantoms has shown th e range of scalability, as well as the potential for use in phantom mode ling. These models serve as the standard for radiation dose assessment for reference individual s, and they will also be the starting point for expanding research into the difference in radiation dose to non-re ference individuals.

PAGE 11

CHAPTER 1 HISTORY OF PHANTOM DEVELOPMENT Introduction Analysis of radiation risk in the human body requires knowledge of the absorbed dose to organs. However, radiation dose is not direct ly measurable in th e body. For this reason, phantoms are used to estimate the doses in real individuals. Two methods exist including taking direct measurements in physical phantoms and using Monte Carlo radiation transport to calculate the absorbed dose to organs in an anthropom orphic computational phantom. A computational phantom is a 3D model of the all internal orga ns and the outer body contour. The goals of a good computational phantom are: anatomical realism, versatility, and having internal organ masses consistent with standards set by the Interna tional Commission on Radi ation Protection (ICRP) The two main types of computational phantoms th at are traditionally ut ilized are stylized (mathematical) phantoms and voxel (tomographic) phantoms. While both have distinct and positive aspects that highlight their usefulness in radiation dose calculation, each presents specific problems that are drawbacks to their us e. Therefore, neither of these two types of phantoms are able to satisfy all of the desire d characteristics of a computational phantom. The hybrid phantom was developed in order to combin e the best characteristics of each of these phantoms. Stylized Phantoms Stylized phantoms use mathematical shapes a nd surfaces (spheres, ellipses, toroids, etc.) to estimate all internal organs and the outer body contour. These phantoms provide an easy to use phantom that is non-uniformly deformable through simple parameter modifications. This type of phantom was first created by Oak Ridge National Laboratory as an adult male model (Snyder et al 1969). This phantom was later expanded to include the male and female adult 10

PAGE 12

phantoms (Kramer et al 1982) and a series of pediatric ph antoms (Cristy and Eckerman 1987). Stylized phantoms were also generated to represen t pregnant females at the end of each trimester of gestation (Stabin et al 1995). These phantoms have the flexibility of changing individual organ masses, which allows the phantoms to correlate to reference values. However, despite all of these organs having the correct anatomical mass, the orga ns lack the correct anatomical shape and position. Furthermore, organ pairs in close proximity (e.g., liver and stomach) should be nestled against each other, but the shapes of the stylized organs do not perm it for the organs to be placed in such close proximity. The more complex contours that are re quired to realistically model organ shape were too advanced for the computer technology when these phantoms were created in the 1980s. Voxel Phantoms As computer technology a dvanced, a new format of voxel phantoms was created. Voxel phantoms use image sets from CT or MRI imag ing of live or deceased patients and thereby maintain realistic organ depth and position. The or gans are visually identified and segmented on the image sets. The result is a large array of voxe ls, each with a specific ti ssue and material label. However, there are several drawbacks to th is methodology that must be considered. One issue is that each voxel phantom is modele d directly after an image set of a patient. Therefore, once the model is complete, the phantom is patient-specific to the particular patient from whom the image data were obtained. The only method for s caling of these phantoms is to uniformly scale the entire phantom by changi ng the voxel dimensions for every voxel in the matrix. There is no process by which to scale indi vidual organs in order to match reference organ masses. This process was previous ly used by several groups (Kramer et al 2004a, b; Nipper et al 2002; Pazik et al 2007). Additionally, voxel phantoms re quire significant time to create. 11

PAGE 13

Each phantom is manually segmented from a set of axial slice images and each pixel of each slice is tagged as a particular organ or tissue, which is a very labor intensive process. Another issue with the creation of voxel phantoms is that there are several organs in CT images that are not always identi fiable. This problem is usually found in soft tissue organs that cannot be identified due to c ontrast resolution issues. For example, delineating the exact boundaries of the pancreas in relation to the st omach and intestines can prove too challenging even for an experienced radiologist. This problem is further exacerbated in pediatric image sets where operating characteristics are usually tailor ed to decrease dose at the expense of image resolution. Furthermore, MR imaging, which woul d provide better soft tissue resolution, requires longer time for image acquisition and has higher risk of patient motion and image blur. Pediatric patients may not even be considered for imaging pr ocedures using this modality as it would often require anesthesia, which adds si gnificant risk to the patient. Hybrid Phantoms A new type of phantom that is coming into wide popularity is the hybrid phantom. The term hybrid was developed to describe the nature of the phantom as having the benefits of both the stylized and voxel phantoms. These models u tilize the advances in computer graphics and modeling for the purposes of accurately describi ng human anatomy while maintaining flexibility of use. The hybrid models contain non-unifo rm rational basis splin e (NURBS) surfaces to describe most organs and polygon mesh surfaces to describe other organs. The NURBS technology was previously deve loped for modeling of a 4D phantom for image correction of cardiac motion in nuclear medicine imaging (Segars et al 1999). The versatility of use of these surfaces was highlighted in this NURBS-based cardiac torso (NCAT) phantom. The spline surfaces were created to mode l the internal organs of the torso and were manipulated easily through movement of control points to simulate the contraction and 12

PAGE 14

13 relaxation of the heart at several time steps (Segars et al 2002). Subsequently, this technology was employed to transform the UF voxel newborn female phantom (Nipper et al 2002) into the UF hybrid newborn male and female phantoms (Lee et al 2007). The NURBS technology allows for individu al organ mass matching through non-uniform deformation. Because these phantoms also utilize CT or MR images of patients, they maintain anatomical realism of organ shape and position. Thus the reference hybrid phantom is one that utilizes the realistic image data while maintain ing the versatility of the free deformation and matching the reference mass for eac h organ as defined by the ICRP.

PAGE 15

CHAPTER 2 DESCRIPTION AND DEVELOPM ENT OF HYBRID PHANTOMS Computed Tomography Data Collection The development of hybrid phantoms can be se parated into four distinct steps: image segmentation, polygon mesh modeling, NURBS m odeling, and voxelization. Before beginning image segmentation, sets of CT image data must be selected. The CT image data archive at the Department of Radiology at Sh ands Childrens Hospital at th e University of Florida was searched by Dr. Choonsik Lee under the approval of the Instituti onal Research Board (IRB) and through HIPAA-compliant practices. Several scans of pediatric pa tients were selected based upon the age of the patient at the time of examin ation, the anatomical regions of coverage and the axial slice resolution of the scans. After se lection of these scans, each examination was reviewed by a pediatric cardiologist, Dr. Jonathan Williams, to ensure that each patient exhibited normal anatomy and did not contain any disease or surgery that w ould otherwise alter the size or position of the internal organs. Finalized data were a collection of 6 CT ex amination studies: 3 chest/abdomen/pelvis (CAP) and 3 head examinations. Two of the he ad examinations also contained ultra high resolution images of the cervical spine at 0.75 mm slice thickness. The data were chosen to yield a CAP scan and a head scan as near as possible to each of the target ages (1, 5, and 10 years). Each head and torso pair were segmented separate ly and were not combined until the later stages of phantom development. Also, high resolution images were obtained of an 18 year old male cadaver at 1.0 mm slick thickness. The scans us ed were of the left arm and left leg, which contained 1098 slices and 819 slices, respectively. The imaging and patient parameters of the CT image sets are summarized in Table 2-1. 14

PAGE 16

Image Segmentation The segmentation was performed using 3D-DOCTOR (Able Software Corp., Lexington, MA, USA). The skeletal tissue of head and to rso regions were segmented semiautomatically using brightness thresholding. First, the Interactive Segmentation feature was employed and a Hounsfield Unit (HU) threshold was determined by visual inspection of image slices. Next, all slices were automatically segmented using the HU threshold previously determined and all pixels with a HU value larger than that threshold were included. This automatic segmentation was manually reviewed to ensure correctness. The most common problem that had to be manually corrected was in the spongiosa region of the bone. Many times in the phantom, certain regions of the spongiosa that contain highe r concentration of bone marrow and lower concentration of ossified trabecular bone are be low the HU threshold and must be manually added to the segmented bone tissue. This problem is particular ly evident in the vertebrae, where there is a very thin layer of cortical bone. Several of the vertebrae were segmented entirely by manual segmentation. After bone segmentation, regions are segmented in the spaces between the vertebral bodies to represent intervertebral disks. However, the axial resolution of each head scan was too large to allow for explicit segmentation of the cervical in tervertebral disks. Therefore, high resolution image sets of the cervical spine were segmente d from the 2.3 year old and 12.3 year old datasets (Fig. 2-1). The 2.3 year old c-spin e was scaled to fit the 1 year and 5 year phantoms, while the 12.3 year old c-spine was scaled to fit the 10 year phantom. At this juncture, the skeleton of each phantom contained a complete head and torso, but no arms and legs. An additional dataset was used of an 18 year old male cadaver. The bones of the left arm and leg were segmented at 1.0 mm slice thickness to be attached to the phantom at a later stage. 15

PAGE 17

As for construction of the 5 year old phantom the UF voxel phantom of a 4 year old male patient was utilized (Lee et al 2006). The voxel model was imported into the 3D-DOCTOR software and segmented in the same manner as the 1 and 10 year phantoms. The only change to the original segmentation was the inclusion of intervertebral disks. Unlike bone tissue segmentation, the internal or gans are almost exclusively segmented by manual segmentation. One exception to this is the lungs. For the lungs, a threshold technique is again used, except that the region included has a HU value less than the threshold value. All other organs are segmented manually using the nodal segmentation method of 3D-DOCTOR. Traditionally image segmentation is performe d by tagging each pixel of each image as belonging to a particular organ. However, the nodal segmentation method utilizes nodes that are placed in sequence on the outer boundary of the or gan and each node is connected to the next by a line. This line is ultimately closed once th e entire boundary of the organ has been defined, which yields a contour of the organ in that image plane. Polygon Modeling After every organ has been segmented in 3D-DOCTOR there is a set of contours for each organ defining the outer boundaries in each image. The 3D Rendering function is used to render three dimensional structure that fits the contours of each organ. The structur e that is created is a polygon mesh. There are two options for this rendering, Complex Surface and Simplified Surface The Simplified Surface is dramatically smoothed, which does not faithfully represent the original segmentation and alters the resulting organ shapes. The Complex Surface option was used because it is a genuine rendering of the segm entation. After rendering, the polygon mesh model was exported into a Wavefront Object file form at. This file format contains information on location of each polygon and the corresponding organ tag. The polygon mesh model was 16

PAGE 18

subsequently imported into Rhinoceros (McNeel, Seattle, WA, US A) for construction of NURBS surfaces. Non-Uniform Rational Basis Spline Modeling Much of the internal organs in polygon mesh format are directly converted to a NURBS surface. This was achieved by lofting a surface over the contours of the polygon mesh. However, not all organs could be represented by NURBS surfaces. For example, the brain was not converted to a NURBS surface because the brain is encased in the cranium. If a smoothed surface were generated for the br ain, then there would be overla pping regions between the brain and the cranium due to the roughness of the cranium. Several organs were not clearl y identifiable on the CT images Therefore, stylized shapes were generated to model these organs. The position of each organ was determined using anatomical landmarks. For example, the thymus was positioned superior to the heart and was bounded by the lungs, clavicles, trachea and thyroid. Skeletal Modeling The skeleton is another orga n that cannot be represented by NURBS surfaces. The spline surfaces are not able to model highly complex structures as it would require too many input parameters to create a smooth spline surface. For example, the vertebrae are exceedingly complex in shape and, as such, are left in polygon me sh format. In fact, the only bone site that is modeled by NURBS surfaces is the ribcage. In order to model the ribcage, the center track of each rib was specified with a curve and several cross sections were take n along the length of the rib in order to create a pipe along the central track with the same cross section. It wa s necessary to use more than one cross section through the rib as the rota tion of the curve from the vertebrae to the costal cartilage caused a rotation of the cross section about the central track. This resulte d in several segmented surfaces 17

PAGE 19

representing each rib. The segments were blended using the function Merge Surface and the resulting tube-like surface was closed on the ends using the Cap Planar Holes function. Additionally, the costal cartilage was generated using the skel etal structure and a priori anatomical knowledge because it is not visually distinguishable from the muscle surrounding the ribcage in the CT data. The central track for each rib was extended to reach from the end of each rib towards the sternum and the contour of each rib was used for cross sectional shape to generate the volume of the cartilaginou s portion of the ribcage (Fig. 2-2). Alimentary Tract Modeling The tortuosity and close proxim ity of the alimentary tract, coupled with the low contrast between much of the soft tissu e in the abdomen, make direct conversion of segmented organs into NURBS surfaces a poor method of modeli ng. For the stomach, the NURBS surface was created and the control points were manipulated such that the shape and position of the organ were matched. Because of the variability in stomach shape and position from person to person, in addition to the variation due to food intake, both the original segmentation and supplemental anatomical literature (Zhang 1999) were utilized in creation of the NURBS stomach surface. As for the esophagus, small intestine, and colon, two pipes (one for th e outer wall, one for the inner wall) were fit to the central track of the lumen. Standardization of Phantoms A significant benefit to hybrid phantom deve lopment is the versatility to match the phantom to a host of reference values. This al lows not only for standardization of dosimetry calculations to that of a reference individual, but also allows for modifying the phantom to match individual variations in body morphometry for reconstructive radiation dose calculations. 18

PAGE 20

19 Anthropometric Data Matching The next step of development was to ma tch the following reference anthropomorphic values: Sitting height Standing height Cranial circumference Arm length Buttock circumference Biacromial breadth Neck circumference Waist circumference It is important to note that the values of biacromial breadth, and buttock, neck, and waist circumference were not available for the 1 year old individual. Therefore, only 5 and 10 year phantoms are matched to these parameters. The reference values for these measuremen ts were reported by several sources. The standing height was provided by ICRP Pub lication 89 (ICRP 2002). The sitting height, biacromial breadth and buttock circumference wa s provided from the third National Health and Nutrition Examination Survey (NHANES III) that was conducted by the Centers for Disease Control (CDC) ( http://www.cdc.gov/nchs/nhanes.htm ). The fourth NHANES yielded reference values for waist circumference for 5 and 10 year phantoms. The reference values for arm length, head circumference, and neck circumference were obtained from the Anthrokids project by the US Consumer Product Safety Co mmission (CPSC) in the late 1970s ( http://www.itl.nist.gov/div894/ovrt/projects/anthrokids ). The reference values for these measuremen ts were reported by several sources. The standing height was provided by ICRP Pub lication 89 (ICRP 2002). The sitting height, biacromial breadth and buttock circumference wa s provided from the third National Health and Nutrition Examination Survey (NHANES III) that was conducted by the Centers for Disease Control (CDC) ( http://www.cdc.gov/nchs/nhanes.htm ). The fourth NHANES yielded reference values for waist circumference for 5 and 10 year phantoms. The reference values for arm length,

PAGE 21

head circumference, and neck circumference were obtained from the Anthrokids project by the US Consumer Product Safety Co mmission (CPSC) in the late 1970s ( http://www.itl.nist.gov/div894/ovrt/projects/anthrokids ). Furthermore, several anthropometric values for the alimentary tract were matched to reference values given in ICRP Publication 100 (ICRP 2006). These values included reference lengths for esophagus, small intestine, right colon, left colon, and rectosigmoid colon (Table 22). It was assumed that right colon was defined as the ascending colon plus the first half of the transverse colon and the left col on was defined as the second half of the transverse colon and the descending colon (up to the rectosigmoid). Organ Mass Matching After matching anthropometric reference values, all organs were scaled to match the reference mass as defined by ICRP reference (ICRP 2002) within 1% (Tables 2-4, 2-5, 2-6). In order to determine the mass of each organ in the phantom, reference densities were obtained from the International Commission on Radiatio n Units and Measurements (ICRU) Publication 42 (ICRU 1992). A major challenge in organ mass matching wa s found with the scaling of the skeleton. After matching anthropometric values, the change in skeletal volume was realized through use of the Smooth and the Offset Mesh functions. However, the cranium could not be modified without intensifying the holes found on the in ferior portion of the cranium, as well as the socket of the eye and in the temporal bone. These holes originate in the segmentation due to the large in-plane shift in location relative to the axial resoluti on. This was corrected through image resegmentation in 3D-DOCTOR. Each slice of the craniu m was thoroughly reviewed using the feature Overlay Neighboring Boundaries in order to visualize and corre ct the locations where neighboring 20

PAGE 22

boundaries do not overlap (Fig. 2-3). Results of th is resegmentation on the cranium are shown in Fig. 2-4. However, not all organs were able to be matched with 1% precision. For each phantom, the thymus was created to fill the space that was bounded by the thyroid, trachea, lungs, clavicles, and heart. Therefore, the largest volume that th e thymus can occupy is found when its surface is abutting each of the neighboring organs. For the 1 year old phantom, the thymus was abutting the neighboring organs and could not be expanded any further, but wa s still below reference mass by 46.5%. Furthermore, after the refe rence lengths for the alimentary tract (ICRP 2006) were matched, the pipe diameter was adjusted to match the reference mass (ICRP 2002). It is important to note that for the 1 year old phantom, the reference mass for left colon was equal to that of the right colon, but the reference lengths were 18 and 21 cm, respectively. Therefore, it was not possible to divide the la rge intestine such that the lengt h was matched within 5% and the mass matched within 1%. Therefore, the colon was split such that the mass was correctly matched, and the resulting errors in the length of the left and right co lon were -7.8% and 8.4%, respectively (Table 2-2). Additionally, the reference gastrointestinal (GI) content from ICRP 89 was not matched using the densities from ICRU 46. The volumes of GI content that would be unrealistically large using these values and it proved too difficult to be modeled. Therefore, an effective content density was utilized so that the re ference content mass would be matched. After matching the individual organs of each phantom, the total body mass was matched through manipulations of the outer body contour. The tissue within th e outer body contour and outside of all of the organs was termed residual soft tissues (R ST). This region includes fat, 21

PAGE 23

muscle, lymphatic tissue, blood vessels, some bone-associated cartilage, and other connective tissues. The density and composition of this re gion was determined as a volume-weighted average of these component tissues in this region. Creation of Complimentary Gender The reference values given for the 1, 5, and 10 year individual specified by the ICRP are not delineated into specific male and female va lues. The gender specific separation of anatomy was assumed to be insignificant for prepubescen t individuals, which is supported by other studies (e.g. Slyper 1998). Therefore, the creation of the gender compliment of each phantom was created through the removal of se x specific organs from the orig inally segmented phantom and insertion of sex specific organs of the opposite gender. This not only allowed for creation of phantoms for which no CT data was readily availa ble, but also maintained uniformity for each gender-pair of phantoms. By utilizing this, a ny radiation dose differences calculated may be directly attributable to differences in gender. Voxelization The ultimate goal of these phantoms is their input into a Monte Carlo radiation transport code for calculation of radiation dose. However, no Monte Carlo code ex ists that can perform this calculation using NURBS or polygon mesh surfaces. Therefore, the NURBS and polygon mesh surfaces of the hybrid phantom must be voxelized. The voxelization process effectively converts all of the surfaces of the phantom into an array of vo xels, each with a specific organ and tissue tag. The algorithms by which the in-house MATLAB code Voxelizer operates is described in a previous study (Lee et al 2007). The first step of voxelization was to export the hybrid model from Rhinoceros using the Raw Triangles format at a meshing tolerance of 10 de grees. This was selected according to a study on the effect of volumetric discrepancies due to the density of polygons used to describe 22

PAGE 24

23 surfaces (Lee et al 2007). The voxel resolution of each phantom was selected such that it was small enough to model the skin because it is the th innest structure in the phantom. The resulting isometric voxel resolution was 0.663 x 0.663 x 0.663 mm3, 0.697 x 0.697 x 0.697 mm3, and 0.824 x 0.824 x 0.824 mm3 for the 1, 5, and 10 year phantom s, respectively. After voxelization, the outermost voxels of the phantom were reassi gned to skin tissue. Th e resulting voxelized organ masses were calculated through counti ng of the voxel volumes and compared to the reference values from ICRP (Tables 2-4, 2-5, 2-6).

PAGE 25

Figure 2-1. Left anterior oblique view of originally segmente d cervical spine (left) and high resolution cervical spine (right) for 1 year old phantom. Ossified tissue is represented in orange and nonossified tissue (car tilage) is repres ented in yellow. 24

PAGE 26

25 Figure 2-2. Example of costal car tilage modeling in hybrid phantoms. This shows the ribcage for the 1 year old phantom. Ossified tissue is represented in orange and nonossified tissue (cartilage) is represented in yellow.

PAGE 27

26 Figure 2-3. Resegmentation of cran ium using 3D-DOCTOR. This i ndividual node points are moved in order to maintain overlap between the current slice boundaries and the neighboring slice boundaries The white arrow in the left window indicates the movement of the current slice boundary and the dotted bl ue arrow indicates the movement of the next slice boundary.

PAGE 28

Figure 2-4. Originally segmented (top) and resegmented (bottom) models of the 1 year old cranium. 27

PAGE 29

28Figure 2-5. Anterior and lateral vi ews of the 1 year male (left) and female (right) phantoms.

PAGE 30

29Figure 2-6. Anterior and lateral vi ews of the 5 year male (left) and female (right) phantoms.

PAGE 31

30 Figure 2-7. Anterior and lateral vi ews of the 10 year male (left) and female (right) phantoms.

PAGE 32

Table 2-1. Summary of CT examinations used for phantom development including CT imaging parameters and patient description. Patient age (y)Target age (y)Gender Examination type Axial resolution (mm) No. of axial images 1.7 1 FC/A/P 3.00 116 2.3 1 F Head 4.50 40 2.3 1 FC-Spine0.75 230 6.7 5 MC/A/P 5.00 116 6.7 5 MHead 4.50 36 11.2 10 FC/A/P 6.00 97 12.3 10 F Head 4.50 46 12.3 10 FC-Spine0.75 220 31

PAGE 33

Table 2-2. Anthropometric reference values fo r the alimentary trac t of 1, 5, and 10 year individuals and the resulting error in hybrid phantoms. Reference data 1 MF5 MF10 MF Length (cm)Esophagus 13 18 23 SI 120170220 R i g h t c o l o n1 82 32 L e f t c o l o n 2 12 63 Rectosigmoid 21 26 31 UF hybrid phantoms Length (cm)Esophagus 13.6418.2022.55 SI 119.01168.30219.76 Right Colon 19.5123.4427.04 Left Colon 19.3625.2929.87 Rectosigmoid 19.9927.2129.80 Percent error (%) 5% tolerance Esophagus 4.951.11-1.96 SI -0.83-1.00-0.11 Right colon 1.91-3.44 Left colon -2.73-3.65 Rectosigmoid -4.824.65-3.86 8 1 8.37 -7.80 32

PAGE 34

Table 2-3. Anthropometric reference values of 1, 5, and 10 year old individuals and resulting error in hybrid phantoms. 1 MF5 MF10 MF1 MF5 MF10 MF1 MF5 MF10 MF Height Standing 76.0109.0138.075.8109.6139.4-0.30.61.0 Sitting 48.860.473.447.960.875.0-1.90.72.3 Length Total arm 32.647.161.032.047.662.0-1.81.01.7 CircumferenceHead 47.351.152.848.150.154.61.6-2.03.4 Neck 24.927.9 25.027.8 0.2-0.3 Waist 55.066.7 57.670.0 4.94.9 Buttock 57.975.2 57.171.9 -1.3-4.3 Breadth Biacromial 25.031.2 24.129.7 -3.5-4.5 Targeted values UF hybrid phantoms Percent difference Anthropometric parameters 33

PAGE 35

Table 2-4. Organ masses for hybrid and voxel phantoms for 1 year male and female. Although separate male and female phantoms exis t, values are listed collectively. Organ System DensityComment Target Volume ICRP 89 (g / cm3) (ICRU 46) (cm3) mass (g)% Diffmass (g)% Diffmass (g) Respiratory System ET1 (anterior nasal layer)1.03ave soft tissue (male) 0.14 0.14 ND ET2 (posterior nasal layer)1.03ave soft tissue (male) 2.04 2.03 ND ET2 (oral cavity layer)1.03ave soft tissue (male) 0.64 0.64 ND ET2 (larynx)1.0750:50 soft tissue/cartilage 3.764.000.0%3.97-0.7%4.00 ET2 (pharynx)1.03ave soft tissue (male) 0.84 0.82 ND Trachea1.0750:50 soft tissue/cartilage 1.411.510.7%1.500.0%1.50 Bronchi extrapulmonary1.0750:50 soft tissue/cartilage 2.17 2.15 ND Lungs (inclusive of blood)0.40calculated 150.000.0%149.91-0.1%150.00 Left Lung0.40calculated 69.36-0.6%69.32-0.6%69.77 Right Lung0.40calculated 80.640.5%80.590.4%80.23 Alimentary System Tongue1.05muscle (newborn) 9.5210.060.6%10.030.3%10.00 Salivary glands1.03ave soft tissue (male) 23.3023.98-0.1%23.93-0.3%24.00 Parotid1.03ICRU-46 ave soft tissue 13.5914.0-0.2%13.9-0.6%14.00 Submaxillary1.03ICRU-46 ave soft tissue 6.807.00.0%7.00.0%7.00 Sublingual1.03ICRU-46 ave soft tissue 2.913.00.5%3.00.4%3.00 Tonsils1.03ave soft tissue (male) 0.490.500.0%0.50-0.1%0.50 Esophagus wall1.03gastrointestine 4.854.98-0.4%4.97-0.5%5.00 Stomach wall1.03gastrointestine 19.4220.040.2%20.010.1%20.00 Stomach contents1.03ave soft tissue (male) 65.0567.000%66.79-0.3%67.00 Small Intestine wall1.03gastrointestine 82.5284.94-0.1%84.75-0.3%85.00 Small Intestine contents1.03ave soft tissue (male) 90.2937.96-59%37.86-59.3%93.00 Colon Right wall1.03gastrointestine 19.4220.090.5%20.010.1%20.00 Right contents1.03ave soft tissue (male) 38.8322.34-44.1%22.26-44.3%40.00 Left wall1.03gastrointestine 19.4219.95-0.3%19.91-0.4%20.00 Left contents1.03ave soft tissue (male) 38.8322.18-44.6%22.13-44.7%40.00 Rectosigmoid wall1.03gastrointestine 9.7110.010.1%9.95-0.5%10.00 Rectosigmoid contents1.03ave soft tissue (male) 19.4222.9714.8%22.8614.3%20.00 Liver1.05liver (40wk fetus) 314.29329.52-0.1%329.27-0.2%330.00 Gall Bladder wall1.03ave soft tissue (male) 1.361.40-0.3%1.400.3%1.40 Gall Bladder contents1.03ave soft tissue (male) 7.778.010.1%8.010.1%8.00 Pancreas1.03ave soft tissue (male) 19.4220.000.0%19.91-0.5%20.00 Circulatory System Heart wall1.04heart (40wk fetus) 48.0850.180.4%50.080.2%50.00 Heart content1.06blood (newborn) 45.2847.990%47.96-0.1%48.00 Urogenital System Kidneys (all regions)1.04kidney (40wk fetus) 67.3169.89-0.2%69.81-0.3%70.00 Cortex (70%)1.04kidney (fetus/child/adult) 49.5951.60.0%51.5-0.1%51.58 Medulla (25%)1.04kidney (fetus/child/adult) 17.7118.3-0.5%18.3-0.6%18.42 Pelvis (5%)1.04bladder (adult-empty) 3.543.7-0.5%3.7-0.6%3.68 Urinary Bladder wall1.04bladder (adult-empty) 8.659.010.1%8.97-0.3%9.00 Urinary Bladder contentsa1.01urine of ave density 31.6810.11-68.4%11.00-65.6%32.00 Penis1.05muscle (newborn) 3.91 3.49 ND Scrotum1.03ave soft tissue (male) 2.91 1.88 ND Testes (2)1.04testes (adult) 1.441.500.0%1.49-0.9%1.50 Prostate Gland1.03ave soft tissue 0.971.000.0%1.00-0.2%1.00 Ovaries (2)1.05ovaries (adult) 0.760.800.0%0.800.1%0.80 Uterus1.05ovaries (adult) 1.431.500.1%1.500.2%1.50aNo reference value is given in ICRP 89 and thus an approximate value is used as defined in the ORNL stylized newborn phantom UFH NURBSUFH Voxel 34

PAGE 36

35 Table 2-4. Continued Organ System Density Comment Target Volume ICRP 89 (g / cm3) (ICRU 46) (cm3) mass (g)% Diffmass (g)% Diffmass (g) Skeletal System Coastal Cartilageb1.10volume-averaged 26.2 23.5 Intervertebral Discsb1.10cortical bone (ICRP89 Para 436) 10.8 10.8 Bone Tissues1.43effective ave density 561.66810.00.6%812.000.9%805.00 Bone (CB, TB)1.66cortical bone (infant) 590.00 Active Marrowc1.03red marrow (adult) 150.00 Inactive Marrow0.98ICRU-46 ave soft tissue 20.00 Miscellaneousd1.03ave soft tissue (male) 45.00 Integumentary System Skine1.10skin (newborn) 318.18ND 345.74-1.2%350.00 Additional Tissues Adrenal Glands (2)1.03ave soft tissue (male) 3.884.010.3%4.010.1%4.00 Brain1.03brain (newborn) 922.33941.40-0.9%941.40-0.9%950.00 Breasts (2)0.96adipose (newborn #2) ND0.44 0.44 ND Ears1.10cartilage (adult) 4.35 3.29 4.35 External nose1.0566:33 soft tiss / cartilage 2.83 2.02 2.83 Eyeballs (2)1.03ave soft tissue (male) 6.807.000.1%6.97-0.4%7.00 Lens (2)1.07eye lens (adult) 0.200.210.3%0.210.9%0.21 Pituitary Gland1.03ave soft tissue (male) 0.150.150.0%0.150.3%0.15 Spinal Cord1.04brain (newborn) ND23.81 23.77 ND Spleen1.06spleen (40wk fetus) 27.3628.85-0.5%28.81-0.7%29.00 Thymus1.03ICRP 89 Para 606 29.2716.07-46.4%16.04-46.5%30.00 Thyroid1.05thyroid (adult) 1.711.800.0%1.800.0%1.80 Residual Soft Tissue1.00effective ave density 2509.17ND 6646.290.3%6623.5 Bone-Associated Cartilage1.10240.9 Separable Fat0.96adipose (newborn #2) 3600.00 Skeletal Muscle1.05muscle (newborn) 1900.00 Separable Connective Tissues1.03ave soft tissue (male) 350.00 Fixed L y mphatic Tissuesf1.03ave soft tissue (male) 71.23 Blood (large vessels)g1.06blood (newborn) 137.38 Miscellaneous ROBh1.03ave soft tissue (male) 324.00 Totals by Organ System Respiratory System 161.34 161.17 155.5 Alimentary System tissues of organ walls 569.45 568.57 569.9 Alimentary System GI tract and gall bladder content 180.46 179.91 268.0 Circulatory System heart wall and content 98.18 98.04 98.0 Urogenital System kidneys and urinary bladder wall 78.90 78.78 79.0 Urogenital System urinary bladder content 10.11 11.00 32.0 Urogenital System internal sex organs (ovaries, uterus, prostate)i3.30 3.30 3.30 Urogenital System external sex organs (penis, scrotum, testes) 8.32 6.86 1.5 Skeletal System bone tissues 810.00 812.00 805.0 Integumentary System ND 345.74 350.0 Additional Tissues excluding rest of body 1030.93 1028.92 1029.3 Additional Tissues rest of body ND 6646.29 6623.5 Total Body Tissues (F) 97430.3%9714 Total Body Tissues (M) 97500.4%9715 Total Body Mass (F) 9934-0.8%10014 Total Body Mass (M) 9941-0.7%10015bSkeletal cartilage excludes the following non-bone associated regions of cartilage: external nose and ears, larynx, trachea, an d extrapulmonary bronchicAssumed to include the 7% of total blood volume in the newborn as per Section 7.7.2 of ICRP 89dAs per Section 9.2.15 of ICRP 89, miscellaneous skeletal tiss ues include periosteum and blood vessels, but exclude periarticula r tissue and bloodeSkin masses given here are for the female phantom, and are 0.15% higher in the male phantom due to the addition of the penis an d scrotumfEstimated from the reference adult values given in Section 7.8. 2 of ICRP Publication 89 and scaled by newborn to adult total bo dy massgTaken as 25.92% of total blood pool as per Section 7.7.2 of ICRP 89 (other tissues, arota, large arteries, large veins)hMiscellaneous rest-of-bod y is added to force the total bod y mass to its ICRP 89 reference value of 3500 giMale phantom masses additionally include soft tissuea occupied by the uterus and ovaries in the corresponding female phantom UFH NURBSUFH Voxel

PAGE 37

Table 2-5. Organ masses for hybrid and voxel phantoms for 5 year male and female. Although separate male and female phantoms exis t, values are listed collectively. Organ System Density Comment Target Volume ICRP 89 (g / cm3) (ICRU 46) (cm3) mass (g)% Diffmass (g)% Diffmass (g) Respiratory System ET1 (anterior nasal layer)1.03ave soft tissue (male) 0.75 0.75 ND ET2 (posterior nasal layer)1.03ave soft tissue (male) 17.25 17.25 ND ET2 (oral cavity layer)1.03ave soft tissue (male) 1.32 1.32 ND ET2 (larynx)1.0750:50 soft tissue/cartilage 6.576.600.4%6.600.4%7.00 ET2 (pharynx)1.03ave soft tissue (male) 1.06 1.06 ND Trachea1.0750:50 soft tissue/cartilage 2.354.510.5%4.5192.2%2.50 Bronchi extrapulmonary1.0750:50 soft tissue/cartilage 3.62 3.62 ND Lungs (inclusive of blood)0.18calculated 300.0300.000.0%300.000.0%300.00 Left Lung0.18calculated 148.26 148.26 139.53 Right Lung0.18calculated 151.74 151.74 160.47 Alimentary System Tongue1.05muscle (newborn) 18.1018.00-0.6%18.00-0.6%19.00 Salivary glands1.03ave soft tissue (male) 33.0133.080.2%33.080.2%34.00 Parotid1.03ICRU-46 ave soft tissue 19.4219.490.4%19.490.4%20.00 Submaxillary1.03ICRU-46 ave soft tissue 9.719.780.7%9.780.7%10.00 Sublingual1.03ICRU-46 ave soft tissue 3.883.880.0%3.880.0%4.00 Tonsils1.03ave soft tissue (male) 1.941.940.0%1.940.0%2.00 Esophagus wall1.03gastrointestine 9.719.61-1.0%9.61-1.0%10.00 Stomach wall1.03gastrointestine 48.5448.44-0.2%48.44-0.2%50.00 Stomach contents1.03ave soft tissue (male) 80.5881.270.9%81.270.9%83.00 Small Intestine wall1.03gastrointestine 213.59214.280.3%214.280.3%220.00 Small Intestine contents0.59ave soft tissue (male) 199.45200.140.3%200.140.3%117.00 Colon Right wall1.03gastrointestine 47.5747.600.1%47.600.1%49.00 Right contents0.82ave soft tissue (male) 60.9461.050.2%61.050.2%50.00 Left wall1.03gastrointestine 47.5748.000.9%48.000.9%49.00 Left contents0.33ave soft tissue (male) 76.0076.100.1%76.100.1%25.00 Rectosigmoid wall1.03gastrointestine 21.3621.22-0.7%21.22-0.7%22.00 Rectosigmoid contents0.37ave soft tissue (male) 67.2267.220.0%67.220.0%25.00 Liver1.05liver (40wk fetus) 542.86544.050.2%544.050.2%570.00 Gall Bladder wall1.03ave soft tissue (male) 2.522.52-0.2%2.52-0.2%2.60 Gall Bladder contents1.03ave soft tissue (male) 14.5614.48-0.6%14.48-0.6%15.00 Pancreas1.03ave soft tissue (male) 33.9834.120.4%34.120.4%35.00 Circulatory System Heart wall1.04heart (40wk fetus) 81.7382.420.8%82.420.8%85.00 Heart content1.06blood (newborn) 127.36128.050.5%128.050.5%135.00 Urogenital System Kidneys (all regions)1.04kidney (40wk fetus) 105.77106.460.7%106.460.7%110.00 Cortex (70%)1.04kidney (fetus/child/adult) 77.9378.620.9%78.620.9%81.05 Medulla (25%)1.04kidney (fetus/child/adult) 27.8327.68-0.5%27.68-0.5%28.95 Pelvis (5%)1.04bladder (adult-empty) 5.575.580.2%5.580.2%5.79 Urinary Bladder wall1.04bladder (adult-empty) 15.3815.380.0%15.380.0%16.00 Urinary Bladder contentsa1.01urine of ave density 61.3962.081.1%62.081.1%62.00 Penis1.05muscle (newborn) 1.58 1.58 ND Scrotum1.03ave soft tissue (male) 0.69 0.69 ND Testes (2)1.04testes (adult) 1.631.630.0%1.630.0%1.70 Prostate Gland1.03ave soft tissue 1.171.170.0%1.170.0%1.20 Ovaries (2)1.05ovaries (adult) 1.901.89-0.7%1.89-0.7%2.00 Uterus1.05ovaries (adult) 2.862.870.5%2.870.5%3.00aNo reference value is given in ICRP 89 and thus an approximate value is used as defined in the ORNL stylized newborn phantom UFH NURBSUFH Voxel 36

PAGE 38

37 Table 2-5. Continued Organ System Density Comment Target Volume ICRP 89 (g / cm3) (ICRU 46) (cm3) mass (g)% Diffmass (g)% Diffmass (g) Skeletal System Coastal Cartilageb1.10volume-averaged 57.5 57.5 Intervertebral Discsb1.10cortical bone (ICRP89 Para 436) 73.1 73.1 Bone Tissues1.41effective ave density 1297.031304.60.6%1304.60.6%1830.00 Bone (CB, TB)1.70cortical bone (infant) 1260.00 Active Marrowc1.03red marrow (adult) 340.00 Inactive Marrow0.98ICRU-46 ave soft tissue 160.00 Teeth1.6515.00 Miscellaneousd1.03ave soft tissue (male) 55.00 Integumentary System Skine1.10skin (newborn) 518.18ND 520.00.4%570.00 Additional Tissues Adrenal Glands (2)1.03ave soft tissue (male) 4.854.860.2%4.90.2%5.00 Brain1.04brain (newborn) 1197.121197.000.0%1197.00.0%1245.00 Breasts (2)0.96adipose (newborn #2) ND4.20 4.2 ND Ears1.10cartilage (adult) 1.32 1.3 5.58 External nose1.0566:33 soft tiss / cartilage 2.38 2.4 7.32 Eyeballs (2)1.03ave soft tissue (male) 10.6810.64-0.4%10.6-0.4%11.00 Lens (2)1.07eye lens (adult) 0.310.310.0%0.30.0%0.33 Pituitary Gland1.03ave soft tissue (male) 0.240.240.0%0.20.0%0.25 Spinal Cord1.04brain (newborn) ND18.68 18.7 ND Spleen1.06spleen (40wk fetus) 47.1747.180.0%47.20.0%50.00 Thymus1.03ICRP 89 Para 606 29.2729.15-0.4%29.1-0.4%30.00 Thyroid1.05thyroid (adult) 3.243.250.3%3.20.3%3.40 Residual Soft Tissue1.01effective ave density 2509.17ND 2500.0-0.4%11878.80 Bone-Associated Cartilage1.10528.3 Separable Fat0.96adipose (newborn #2) 5000.00 Skeletal Muscle1.05muscle (newborn) 5600.00 Separable Connective Tissues1.03ave soft tissue (male) 700.00 Fixed Lymphatic Tissuesf1.03ave soft tissue (male) 190.00 Blood (large vessels)g1.06blood (newborn) 388.80 Miscellaneous ROBh1.03ave soft tissue (male) 0.00 Totals by Organ System Respiratory System 340.52 340.52 309.5 Alimentary System tissues of organ walls 1700.54 1700.54 1096.6 Alimentary System GI tract and gall bladder content 360.00 360.00 315.0 Circulatory System heart wall and content 369.84 369.84 220.0 Urogenital System kidneys and urinary bladder wall 204.76 204.76 126.0 Urogenital System urinary bladder content 98.94 98.94 62.0 Urogenital System internal sex organs (ovaries, uterus, prostate)i1.60 1.60 6.20 Urogenital System external sex organs (penis, scrotum, testes) 12.20 12.20 1.7 Skeletal System bone tissues 3773.74 3773.74 1830.0 Integumentary System ND ND 570.0 Additional Tissues excluding rest of body 1548.19 1548.19 1357.9 Additional Tissues rest of body ND ND 11878.8 Total Body Tissues (F) 16726 17395 Total Body Tissues (M) 16728 17397 Total Body Mass (F) 17088 17772 Total Body Mass (M) 17090 17774bSkeletal cartilage excludes the following non-bone associated r egions of cartilage: external nose and ears, larynx, trachea, an d extrapulmonary bronchicAssumed to include the 7% of total blood volume in the newborn as per Section 7.7.2 of ICRP 89dAs per Section 9.2.15 of ICRP 89, miscellaneous skeletal tiss ues include periosteum and blood vessels, but exclude periarticula r tissue and bloodeSkin masses given here are for the female phantom, and are 0.15% higher in the male phantom due to the addition of the penis an d scrotumfEstimated from the reference adult values given in Section 7.8.2 of ICRP Publication 89 and scaled by newborn to adult total bo dy massgTaken as 25.92% of total blood pool as per Section 7.7.2 of ICRP 89 (other tissues, arota, large arteries, large veins)hMiscellaneous rest-of-body is added to force the total body mass to its ICRP 89 reference value of 3500 giMale phantom masses additionally include soft tissue occupied by the uterus and ovaries in the corresponding female phantom UFH NURBSUFH Voxel

PAGE 39

Table 2-6. Organ masses for hybrid and voxel phantoms for 10 year male and female. Although separate male and female phantoms exis t, values are listed collectively. Organ System Density Comment Target Volume ICRP 89 (g / cm3) (ICRU 46) (cm3) mass (g)% Diffmass (g)% Diffmass (g) Respiratory System ET1 (anterior nasal layer)1.03ave soft tissue (male) 0.75 0.67 ND ET2 (posterior nasal layer)1.03ave soft tissue (male) 17.25 17.28 ND ET2 (oral cavity layer)1.03ave soft tissue (male) 1.32 1.39 ND ET2 (larynx)1.0750:50 soft tissue/cartilage 11.2712.030.2%12.010.1%12.00 ET2 (pharynx)1.03ave soft tissue (male) 1.06 1.05 ND Trachea1.0750:50 soft tissue/cartilage 4.234.520.4%4.48-0.3%4.50 Bronchi extrapulmonary1.0750:50 soft tissue/cartilage 3.62 3.54 ND Lungs (inclusive of blood)0.32calculated 500.000.0%498.40-0.3%500.00 Left Lung0.32calculated 250.767.8%249.877.4%232.56 Right Lung0.32calculated 249.24-6.8%248.53-7.1%267.44 Alimentary System Tongue1.05muscle (newborn) 30.4832.020.1%31.83-0.5%32.00 Salivary glands1.03ave soft tissue (male) 42.7243.990.0%43.88-0.3%44.00 Parotid1.03ICRU-46 ave soft tissue 25.2426.00.0%25.9-0.3%26.00 Submaxillary1.03ICRU-46 ave soft tissue 12.6213.00.0%13.0-0.3%13.00 Sublingual1.03ICRU-46 ave soft tissue 4.85 5.00.0% 5.0-0.4%5.00 Tonsils1.03ave soft tissue (male) 2.913.000.0%2.99-0.4%3.00 Esophagus wall1.03gastrointestine 17.4818.000.0%17.94-0.3%18.00 Stomach wall1.03gastrointestine 82.5284.85-0.2%84.68-0.4%85.00 Stomach contents1.03ave soft tissue (male) 113.59116.990%116.02-0.8%117.00 Small Intestine wall1.03gastrointestine 359.22370.530.1%368.69-0.4%370.00 Small Intestine contents0.82ave soft tissue (male) 199.45163.000%162.54-0.3%163.00 Colon Right wall1.03gastrointestine 82.5285.090.1%84.85-0.2%85.00 Right contents1.15ave soft tissue (male) 60.9470.000.0%69.90-0.1%70.00 Left wall1.03gastrointestine 82.5285.050.1%84.70-0.3%85.00 Left contents0.46ave soft tissue (male) 76.0035.000.0%34.86-0.4%35.00 Rectosigmoid wall1.03gastrointestine 38.8340.000.0%39.75-0.6%40.00 Rectosigmoid contents0.52ave soft tissue (male) 67.2235.000.0%34.83-0.5%35.00 Liver1.05liver (40wk fetus) 790.48829.630.0%828.71-0.2%830.00 Gall Bladder wall1.03ave soft tissue (male) 4.274.40-0.1%4.410.2%4.40 Gall Bladder contents1.03ave soft tissue (male) 25.2426.010.0%25.97-0.1%26.00 Pancreas1.03ave soft tissue (male) 58.2559.990.0%59.89-0.2%60.00 Circulatory System Heart wall1.04heart (40wk fetus) 134.62140.450.3%140.050.0%140.00 Heart content1.06blood (newborn) 216.98229.380%229.20-0.3%230.00 Urogenital System Kidneys (all regions)1.04kidney (40wk fetus) 173.08179.72-0.2%179.44-0.3%180.00 Cortex (70%)1.04kidney (fetus/child/adult) 127.53132.4-0.2%132.2-0.3%132.63 Medulla (25%)1.04kidney (fetus/child/adult) 45.5547.3-0.1%47.3-0.2%47.37 Pelvis (5%)1.04bladder (adult-empty) 9.11 9.50.1% 9.5-0.1%9.47 Urinary Bladder wall1.04bladder (adult-empty) 24.0425.030.1%24.91-0.4%25.00 Urinary Bladder contentsa1.01urine of ave density 98.0298.94-0.1%98.61-0.4%99.00 Penis1.05muscle (newborn) 8.23 8.18 ND Scrotum1.03ave soft tissue (male) 1.98 2.82 ND Testes (2)1.04testes (adult) 1.922.00-0.1%1.99-0.5%2.00 Prostate Gland1.03ave soft tissue 1.551.600.0%1.59-0.4%1.60 Ovaries (2)1.05ovaries (adult) 3.333.49-0.1%3.47-0.8%3.50 Uterus1.05ovaries (adult) 3.814.000.1%3.99-0.3%4.00aNo reference value is given in ICRP 89 and thus an approxim ate value is used as defined in the ORNL stylized newborn phantom UFH NURBS UFH Voxel 38

PAGE 40

39 Table 2-6. Continued Organ System Density Comment Target Volume ICRP 89 (g / cm3) (ICRU 46) (cm3) mass (g)% Diffmass (g)% Diffmass (g) Skeletal System Coastal Cartilageb1.10volume-averaged 57.5 55.2 Intervertebral Discsb1.10cortical bone (ICRP89 Para 436) 73.1 67.9 Bone Tissues1.38effective ave density 2674.353680.40.0%3665.76-0.4%3680.00 Bone (CB, TB)1.75cortical bone (infant) 2300.00 Active Marrowc1.03red marrow (adult) 630.00 Inactive Marrow0.98ICRU-46 ave soft tissue 630.00 Teeth1.6530.00 Miscellaneousd1.03ave soft tissue (male) 90.00 Integumentary System Skine1.10skin (newborn) 745.45 ND 0.00-100.0%820.00 Additional Tissues Adrenal Glands (2)1.03ave soft tissue (male) 6.807.010.2%7.000.0%7.00 Brain1.04brain (newborn) 1259.621309.990%1309.610.0%1310.00 Breasts (2)0.96adipose (newborn #2) ND7.65 7.63 ND Ears1.10cartilage (adult) 5.58 5.54 5.58 External nose1.0566:33 soft tiss / cartilage 7.32 6.89 7.32 Eyeballs (2)1.03ave soft tissue (male) 11.6511.99-0.1%11.92-0.7%12.00 Lens (2)1.07eye lens (adult) 0.340.360.5%0.360.1%0.36 Pituitary Gland1.03ave soft tissue (male) 0.340.350.2%0.350.1%0.35 Spinal Cord1.04brain (newborn) ND72.49 69.95 ND Spleen1.06spleen (40wk fetus) 75.4780.010.0%79.81-0.2%80.00 Thymus1.03ICRP 89 Para 606 36.5937.530.1%37.490.0%37.50 Thyroid1.05thyroid (adult) 7.527.900.0%7.89-0.2%7.90 Residual Soft Tissue1.02effective ave density 2509.17 ND 23736.034.6%22684.33 Bone-Associated Cartilage1.10427.3 Separable Fat0.96adipose (newborn #2) 7500.00 Skeletal Muscle1.05muscle (newborn) 11000.00 Separable Connective Tissues1.03ave soft tissue (male) 1100.00 Fixed Lymphatic Tissuesf1.03ave soft tissue (male) 320.00 Blood (large vessels)g1.06blood (newborn) 648.00 Miscellaneous ROBh1.03ave soft tissue (male) 1689.00 Totals by Organ System Respiratory System 540.56 538.82 516.5 Alimentary System tissues of organ walls 1700.54 1696.20 1700.4 Alimentary System GI tract and gall bladder content 446.00 444.13 446.0 Circulatory System heart wall and content 369.84 369.25 370.0 Urogenital System kidneys and urinary bladder wall 204.76 204.34 205.0 Urogenital System urinary bladder content 98.94 98.61 99.0 Urogenital System internal sex organs (ovaries, uterus, prostate)i9.10 9.05 9.10 Urogenital System external sex organs (penis, scrotum, testes) 12.20 12.99 2.0 Skeletal System bone tissues 3680.43 3665.76 3680.0 Integumentary System ND 0.00 820.0 Additional Tissues excluding rest of body 1548.19 1544.44 1468.0 Additional Tissues rest of body ND 23736.03 22684.3 Total Body Tissues (F) 317641.0%31453 Total Body Tissues (M) 317771.0%31455 Total Body Mass (F) 323071.0%31998 Total Body Mass (M) 323201.0%32000bSkeletal cartilage excludes the following non-bone associated regions of cartilage: external nose and ears, larynx, trachea, an d extrapulmonary bronchicAssumed to include the 7% of total blood volume in the newborn as per Section 7.7.2 of ICRP 89dAs per Section 9.2.15 of ICRP 89, miscellaneous skeletal tissues include periosteum and blood vessels, but exclude periarticula r tissue and bloodeSkin masses given here are for the female phantom, and are 0.15% higher in the male phantom due to the addition of the penis an d scrotumfEstimated from the reference adult values given in Section 7.8.2 of ICRP Publication 89 and scaled by newborn to adult total bo dy massgTaken as 25.92% of total blood pool as per Section 7.7.2 of ICRP 89 (other tissues, arota, large arteries, large veins)hMiscellaneous rest-of-body is added to force the total body mass to its ICRP 89 reference value of 3500 giMale phantom masses additionally include soft tissuea occupied by the uterus and ovaries in the corresponding female phantom UFH NURBS UFH Voxel

PAGE 41

CHAPTER 3 CONCLUSIONS Both stylized (equation-based) and to mographic (image-based) phantoms have traditionally been used for assessment of radiation dose. Each of these has a particular forte. Stylized phantoms remain more flexible and ea sier to use, while voxel phantoms are more realistic of human beings. The UFH phantoms de veloped combine both of these features into one. The direct segmentation of CT or MRI images results in a phantom that realistically depicts complex organ shapes. Furthermore, the use of NURBS surfaces in the UFH phantoms allows for a wide range of orga n scalability. This technology allo wed for creation of the 1, 5, and 10 year phantoms that are at the reference valu es for 8 different anthropometric measures, several different alimentary tract lengths, and at the ICRP masses for nearly all organs. In particular, the resourcefulness of this modeling technique was high lighted in the modeling of the alimentary tract. Previous segmentation methods required exact recognition of the soft tissue boundaries of the wall of the intestine in order to corre ctly model the GI tract. However, by using the central track of the lumen and fitting NURBS surfaces around it, the intestines and colon were able to be modeled in each of the phantoms using a pipe surface and the reference mass for each of these organs was matched within 1%. It is important to note that the GI contents were not matched to within 1% of ICRP reference valu es. However, these values are susceptible to the food intake and digestive rates of each individual, which makes the masses highly variable. Moreover, this method of surface modeling lend s the phantom to further uses. The freely deformable surfaces allow for scaling of the pha ntom to non reference dimensions. This would allow the user to utilize these phantom s in modeling of cases such as non-50th percentile individuals by height (ove rly tall or short) or weight (underweight or overweight) for radiation 40

PAGE 42

41 protection or medical dose assessment studies. Th e ease of NURBS surfaces could also allow for the modeling of disease states of an organ or modeling a solid tumor in the body. Further studies should be implemented to analyze these issues to expand the knowledge of radiation protection to a larger portion of the total population.

PAGE 43

LIST OF REFERENCES Cristy M, and Eckerman KF 1987 Specific absorbed fractions of energy at various ages from internal photon sources. Oak Ridge, TN: Oak Ridge National Laboratory. ORNL/TM8381/Volumes I-VII ICRP 2002 Basic anatomical and phys iological data for use in radi ological protection: reference values. New York, New York: Internati onal Commission on Radiological Protection. Publication 89 ICRP 2006 Human alimentary tract model for ra diological protection. Oxford; Pergamon Press: International Commission on Radiologi cal Protection. ICRP Publication 100 ICRU 1992 Photon, electron, proton and neutron interaction data for body tissues. Bethesda, MD: International Commission on Radiation Units and Measurements. Report 46 Kramer R, Khoury HJ, Vieira JW, Loureiro EC, Lima VJ, Lima FR, and Hoff G 2004a All about FAX: a Female Adult voXel phantom for Mont e Carlo calculation in radiation protection dosimetry. Phys Med Biol 49(23) 5203-16 Kramer R, Vieira JW, Khoury HJ, and Li ma FD 2004b MAX meets ADAM: a dosimetric comparison between a voxel-based and a math ematical model for external exposure to photons. Physics in Me dicine and Biology 49(6) 887-910 Kramer R, Zankl M, Williams G, and Drexler G 1982 The calculation of dose from external photon exposures using reference human phantoms and Monte-Carlo methods, Part 1: The male (ADAM) and female (EVA) adult mathematical phantoms. Neuherberg, Germany: GSF-National Research Center for Health and Environment. GSF Bericht S885 Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Str awn LM, Flegal KM, Mei Z, Wei R, Curtin LR, Roche AF, and Johnson CL 2002 2000 Growth Charts for the Unites States: improvements to the 1977 National Center fo r Health Statistics version. Pediatrics 109 45-60 Lee C, Lee C, Williams JL, and Bolch WE 2006 Whole-body voxel phantoms of paediatric patients UF Series B. Phys Med Biol 51(17) 4649-4661 Lee C, Lodwick D, Hasenauer D, Williams JL, Lee C, and Bolch WE 2007 Hybrid computational phantoms of the male and female newborn patient: NURBS-based wholebody models. Phys Med Biol 52(12) 3309-3333 Nipper JC, Williams JL, and Bolch WE 2002 Creation of two tomographic voxel models of pediatric patients in the first year of life. Phys Med Biol 47(11) 3143-3164 42

PAGE 44

43 Pazik FD, Staton RJ, Hintenlang DE, Arreola MM, Williams JL, and Bolch WE 2007 Organ and effective doses in newborns and infants undergoing voiding cystourethrograms (VCUG): A comparison of stylized and tomographic phantoms. Med Phys 34(1) 294-306 Segars WP, Lalush DS, and Tsui BMW 2001 Mode ling respiratory mechan ics in the MCAT and spline-based MCAT phantoms. Ieee Transactions on Nuclear Science 48 (1) 89-97 Segars WP, and Tsui BM 2002 Study of the efficacy of respiratory gating in myocardial SPECT using the new 4D NCAT phantom. IEEE Trans Nucl Sci 49(3) 675-679 Slyper AH 1998 Childhood obesity, adipose tissue di stribution, and the pediatric practitioner. Pediatrics 102 e4 (electronic version) Snyder WS, Ford MR, Warner GG, and Fisher HL 1969 Estimates of absorbed fractions for monoenergetic photon sources uni formly distributed in various organs of a heterogeneous phantom. New York: Society of Nuclear Medicine. MIRD Pamphlet No. 5 Stabin M, Watson E, Cristy M, Ryman J, Eckerm an K, Davis J, Marshall D, and Gehlen M 1995 Mathematical models and specific abso rbed fractions of photon energy in the nonpregnant adult female and at the end of each trimester of pregnancy. Oak Ridge, TN: Oak Ridge National Laboratory. ORNL/TM-12907 Zhang S-x 1999 An atlas of histology. New York: Springer

PAGE 45

BIOGRAPHICAL SKETCH Daniel Lee Lodwick was born in Columbus, Oh io, in 1984. Daniel is the son of David and Kathy Lodwick. Daniel graduated from Royal Palm Beach Community High School in 2002 and attended the University of Fl orida thereafter. In fall 2006, he earned his B.S. in nuclear engineering and graduated summa cum laude. Daniel is currently enrolled in the College of Engineering and pursuing his Master of Scien ce degree in nuclear e ngineering sciences.