• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Background and related researc...
 System design and construction
 System operation, calibration,...
 Conclusions and recommendation...
 Bibliography
 Biographical sketch














Group Title: automated dosimetry system for computed tomography x-ray scanners using silicon p-i-n diodes /
Title: An automated dosimetry system for computed tomography x-ray scanners using silicon p-i-n diodes /
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097457/00001
 Material Information
Title: An automated dosimetry system for computed tomography x-ray scanners using silicon p-i-n diodes /
Physical Description: ix, 108 leaves : ill. ; 28 cm.
Language: English
Creator: Lanza, John Joseph, 1953-
Publication Date: 1979
Copyright Date: 1979
 Subjects
Subject: Scanning systems   ( lcsh )
Diodes   ( lcsh )
Tomography   ( lcsh )
Diagnosis, Radioscopic   ( lcsh )
Radiology, Panoramic   ( lcsh )
Nuclear Engineering Sciences thesis Ph. D   ( lcsh )
Dissertations, Academic -- Nuclear Engineering Sciences -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 101-106.
Additional Physical Form: Also available on World Wide Web
Statement of Responsibility: by John Joseph Lanza.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097457
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000087524
oclc - 05530176
notis - AAK2892

Downloads

This item has the following downloads:

PDF ( 4 MBs ) ( PDF )


Table of Contents
    Title Page
        Page i
        Page ii
    Acknowledgement
        Page iii
    Table of Contents
        Page iv
    List of Tables
        Page v
    List of Figures
        Page vi
        Page vii
    Abstract
        Page viii
        Page ix
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
    Background and related research
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
    System design and construction
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
    System operation, calibration, and experimental results
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
    Conclusions and recommendations
        Page 100
    Bibliography
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
    Biographical sketch
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
Full Text











AN AUTOMATED DOSIMETRY SYSTEM
FOR COMPUTED TOMOGRAPHY X-RAY SCANNERS
USING SILICON P-I-N DIODES








BY



JOHN JOSEPH LANZA


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE
UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA


1979














ACKNOWLEDGMENTS


I would like to thank the members of my supervisory committee for

their invaluable assistance in completing my degree requirements. Dr.

Walter Mauderli, my chairman, has successfully guided me through many

trying times while working on the project. Dr. Genevieve Roessler, my

co-chairman, deserves special recognition because she was responsible

for my entering the field of radiation physics and for sponsoring a

research assistantship during the first three years of my graduate work.

Dr. Lawrence Fitzgerald was always willing to answer my questions any

time. Dr. Frank Agee allowed me to work with the CT scanner in Neuro-

radiology. Dr. Eugene Chenette was my first contact at the University

six years ago and quickly agreed to substitute for another member of the

original committee.

Special thanks go to Howard Brown, Ken Fawcett, Joe Mueller, and

Charles Rabbit for providing technical assistance during the course of

the project.

The Bureau of Radiological Health was responsible for the funding

that allowed me to complete these studies. I appreciate the guidance of

Bill Properzio, Tommy Morgan, and Tom Lee at the Bureau.

I would like to thank my friend, Dr. John Stampelos, for special

understanding during this work.

Finally, I would like to thank my typist for the patience she has

shown me in previous papers.


iii









TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS jii

LIST OF TABLES v

LIST OF FIGURES vi

ABSTRACT viii

CHAPTER

I INTRODUCTION 1

II BACKGROUND AND RELATED RESEARCH 7

Developmental Review 7

Dose Considerations 12

Present Dose Measurement Techniques 17

Semiconductor Radiation Detectors 25

III SYSTEM DESIGN AND CONSTRUCTION 44

Diode Probe/Amplifier Module 46

Control/Readout Module 67

IV SYSTEM OPERATION, CALIBRATION, AND EXPERIMENTAL
RESULTS 95

System Operation 95

Calibration and Results 96

V CONCLUSIONS AND RECOMMENDATIONS 100

BIBLIOGRAPHY 101

BIOGRAPHICAL SKETCH 107










LIST OF TABLES


Page


Table 1 Selected Whole Body Scanner Exposure Specifi-
cations as Reported by Manufacturers 19

Table 2 Resistor Selection for Exposure Range 75

Table 3 Variations in VFC Output Frequency 78

Table 4 Power Supply Requirements for the Dosimeter 94

Table 5 Reproducibility Studies of the Diode Probe 97

Table 6 Calibration of Diode Probe at 100 and 120 kVp 99












LIST OF FIGURES


Page


FIGURE

FIGURE

FIGURE

FIGURE

FIGURE



FIGURE

FIGURE

FIGURE

FIGURE

FIGURE


FIGURE 11.



FIGURE 12.


FIGURE

FIGURE

FIGURE



FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE


13.

14.

15.



16.

17.

18.

19.

20.

21.


BLOCK DIAGRAM OF DOSIMETRY SYSTEM

FIRST GENERATION CT SCANNER

SECOND GENERATION CT SCANNER

THIRD GENERATION CT SCANNER

PIXEL GRAY SCALE U + U2 + U3 REPRESENTED

AS A SINGLE SHADE OF GRAY ON THE IMAGE MONITOR

TYPICAL EXPERIMENTAL FILM EXPOSURE PROFILES

TYPICAL TLD EXPOSURE PROFILES

TYPICAL PENCIL IONIZATION CHAMBER PROFILE

BASIC SKETCH OF A SILICON P-I-N DIODE

ELECTRON-HOLE PRODUCTION FROM PHOTON INTER-
ACTIONS IN PHOTODIODE MODE OF OPERATION

ELECTRON-HOLE MOVEMENT IN PHOTOVOLTAIC MODE
OF OPERATION

DIODE PHOTOCURRENT VS EXPOSURE RATE FOR AN
RCA C30822 SILICON P-I-N PHOTODIODE

MODEL OF DIODE CURRENT MEASURING CIRCUIT

RADIATION INCIDENT ON THE DIODE

DIRECTIONAL EXPOSURE MEASUREMENTS ON THE RCA
C30822 PHOTODIODE

ENERGY RESPONSE OF RCA C30822 PHOTODIODE

DOSIMETER SYSTEM OPERATIONAL DIAGRAM

DIODE PROBE/AMPLIFIER MODULE

CROSS-SECTIONAL VIEW OF PLEXIGLAS DIODE PROBE

DIODE PROBE ASSEMBLY

DIODE SPACING IN THE PROBE FOR ONE-HALF OF THE
SYMMETRICAL ARRAY











FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE

FIGURE


22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.


FIGURE 40.



FIGURE 41.


AMPLIFIER BOARD CONFIGURATION

NONINVERTING OP AMP CONFIGURATION

INVERTING OP AMP CONFIGURATION

CURRENT-TO-VOLTAGE CONVERTER CIRCUIT

COMPONENT SIDE OF AMPLIFIER BOARD

WIRING SIDE OF AMPLIFIER BOARD

DIODE DETECTOR AMPLIFIER RESPONSE

PACKAGE LAYOUT BOARDS 1-4

PACKAGE LAYOUT BOARD 5

ABSOLUTE VALUE CIRCUIT

VOLTAGE-TO-FREQUENCY CONVERTER CIRCUITRY

EXPOSURE RANGE OVERSCALE CIRCUITRY

DECADE COUNTER AND ASSOCIATED CIRCUITRY

PIN CONNECTION FOR ICM 7217 IJI

CIRCUITRY FOR COUNTER PINS 2,9,10,14

COMPARATOR TRIP CIRCUITRY

CLOCK UP/DOWN COUNTER CIRCUITRY

COUNTER LOAD REGISTER AND PRINTER HANDSHAKE
CIRCUITRY

TIMING CHART FOR MUX SCAN OSCILLATOR OVERRIDE AND
LOAD REGISTER FUNCTION FOR ONE PRINT CYCLE

DISPLAY BLANKING CONTROL CIRCUIT


Page














Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy


AN AUTOMATED DOSIMETRY SYSTEM
FOR COMPUTED TOMOGRAPHY X-RAY SCANNERS
USING SILICON P-I-N DIODES

By

John Joseph Lanza

March, 1979


Chairman: Walter Mauderli, D. Sc.
Co-Chairman: Genevieve S. Roessler, Ph.D
Major Department: Nuclear Engineering Sciences

A dosimetry system for computed tomography (CT) x-ray scanners has

been developed featuring X-Y as well as Z-axis directional exposure

measurement capabilities. The device is unique because it uses an ar-

ray of closely spaced silicon p-i-n diodes as radiation detectors. This

arrangement allows detailed mapping of dose levels along the length of

the detector module.

Computed tomography scanners are a new generation of diagnostic

radiology devices using highly collimated x-ray beams that pass through

a body section axially. The photons not absorbed are detected (depend-

ing on the manufacturer's model) by scintillating crystals, ionization

chambers, or semiconductor materials, and their number is a function of

the density of the material they must traverse. The x-ray tube and

detectors generally rotate in unison at opposite sides around the

patient while collecting absorption data for later reconstruction by

computer and display.


viii






Prior to the development of this dosimeter, no simple and con-

venient method was available for performing exposure distribution

measurements on CT scanners. Previous measurement methods have in-

cluded the use of photographic film, pencil-type ionization chambers,

or an array of as many as 300 to 500 thermoluminescent dosimeter chips.

Silicon diodes are superior to these methods because of their small

size, higher photon stopping power, linear energy response, and minimal

x-ray beam directional dependence.

The system described features 25 evenly spaced silicon diode de-

tectors each with its own current-to-voltage converter. The amplified

output from each diode is transferred to the data handling module that

integrates the signal and provides permanent storage via a thermal

printer. Calibration factors can be applied allowing the readout to

correspond to the radiation exposure in Roentgens received by the

diodes during the course of a CT scan. The dynamic range of the in-

strument enables it to measure exposures in fields as high as 100 R/sec

and to present the integrated exposure information in three ranges

of 0-1, 10, and 100 R.














CHAPTER I
INTRODUCTION



The field of diagnostic radiology began in 1895 shortly after the

discovery of x-rays by Roentgen (Jo74). During the past three quarters

of a century improvements have been made in both techniques and equip-

ment but, until a few years ago, there were really no major advances in

diagnostic x-ray procedures.

A revolution was started in 1967 when Hounsfield of EMI Limited

developed a new x-ray transmission system--the computed axial tomography

scanner or CT scanner (Ho73). The CT system possessed features novel to

radiology, including:

1. X-ray beams that are highly collimated to reduce scatter con-
tributions;

2. The use of solid-state scintillation detectors possessing high
signal-to-noise ratios;

3. The use of mathematical reconstruction techniques to solve for
the attenuation coefficients of the tissues of interest, there-
by revealing adjacent areas of slightly varying density; and

4. The presentation of the radiologic image as an axial slice, thus,
eliminating the obscuration of detail due to overlapping tissues
as is currently found in conventional radiography.

The introduction of the CT scanner has greatly affected health care

delivery because of these unique characteristics. In many clinical situ-

ations, because of its ease of use, non-invasive technique, and lack of

need for inpatient care, the CT scan has replaced many common radiologi-

cal procedures such as arteriograms, pneumoencephalograms, and some

nuclear medicine examinations (Tu77;Ba7S;Ev78;Li75;Fi77).







The basic technique of CT calls for a highly collimated x-ray beam to

pass through a body section axially. The first units could scan only the

head but now scanners are available that can view all areas of the body.

The photons not absorbed in body tissues are detected, depending on the

model, by scintillating crystals, ionization chambers, or by semiconduc-

tor materials and their number is a function of the density of the

material they must traverse (CT77;Va77;Se76;As76). In most current units,

the x-ray tube and the detectors rotate in unison at opposite sides around

the patient while collecting absorption data for later reconstruction and

display. The earliest CT system took as long as five minutes to complete

one scan. Now, however, units are being produced with scan times as low

as one or two seconds (Le76;Pi78). Sub-second scanners are presently on

the drawing boards.

In 1973, the first EMI head units were installed in hospitals in the

United States (Co76). Since then, over 760 units of both types have been

sold in this country (Com77). Presently, more than ten companies are

either manufacturing or designing CT scanning systems.

Extensive research has been undertaken in hopes of improving resolu-

tion, scan speed, and data handling in CT systems (Go77a). Patient radia-

tion dosimetry is one area, however, that has not, as yet, been defini-

tively studied (Co76;C176;Ta77;Bur77;Co77). In the past history of

diagnostic radiology, misapplication of x-rays often resulted in dramatic

biological effects including leukemia and various neoplasms. The full

impact of the deleterious effects of the radiation exposure, in many

cases, was not realized until years later. Preliminary studies on CT

scanners indicate patient dosage varies as to (3a76;Mc75;Ba77;Mc74;Ph75a;

Ph75b;We77;Mc76;Pe73):







1. Scan time;

2. Number of slices taken;

3. Picture element size in the reconstructed matrix;

4. Slice thickness; and

5. The voltage and current settings of the x-ray tube on a parti-
cular unit.

A normal CT procedure usually consists of between five and ten con-

tiguous scans of a particular region of the body. Most units have the

feature of viewing two slightly overlapping sections or slices per scan

(Le74;Oh76a;CT76). Scan widths in many units can be varied between 3mm

and 15mm per slice (Va76;Se76). Skin doses in excess of five rads per

scan have been measured from some CT models (Oh76b;Mc76;ShoT-1. It must

be pointed out, however, that the exposure per scan is to a small area

of the body. This is in contrast to conventional radiological studies

where the dose is larger due to a higher scatter component. From these

considerations, integrated doses over the entire area of the scan may

add up to skin doses as high as 25 rads or more (We77). Internal doses

are smaller but still appreciable (Mc76).

The "Radiation Control for Health and Safety Act" (PL 90-602) was

issued to set standards that would act to reduce human exposure from

x-ray equipment (including CT scanners) (Re76). Numerous states have

established boards to determine standards and criteria for CT scanner

acquisition which includes the requirement that patient radiation expo-

sures be minimized while realizing the risk-benefit considerations of

diagnostic radiology (C176;St77). Adequate health care planning and

delivery must include a quantitative assessment of radiation exposure

patients can be expected to receive during normal x-ray procedures. If

adequate dosimetric studies can be performed during the infancy of CT






scanning, questions and criticisms as to the usefulness of this type of

procedure can be accurately evaluated.

Presently the Bureau of Radiological Health (BRH) is developing

regulations to modify or replace the performance standard for "Diagnostic

X-ray Systems and Their Major Components" (21CFR1020.30). At the moment,

CT scanners must uphold the minimum design and operational requirements

set for conventional diagnostic x-ray machines. Because of the unique

technology presented by the CT device, standards must be designed to

permit maximum patient medical benefit with a minimum obligatory radia-

tion exposure risk (Sc78;Co78;Pr78).

Dosimetric studies performed on CT scanners up to this time have

used photographic film, thermoluminescent dosimeters(TLD), or ion chambers

as the radiation detectors (Ca77;Mo77;De78). In a continuing study, the

BRH is conducting in-depth dosimetric research using as many as 500 TLD

dosimeters placed inside a Plexiglas phantom. This technique proves to

be both tedious and expensive. At the moment, however, it is the most

expeditious means of obtaining dose profile information.

The project described involved the design and construction of a

portable radiation dosimetry system for CT scanner operations. Present

phantoms for CT systems were intended mainly to measure only the image

quality of the various units available (Co77a;Al77;HT76). The device

constructed consists of a Plexiglas module containing 25 serially spaced

silicon diodes that are used as the radiation detectors (Figure 1). This

arrangement allows for the measurement of exposures at selected points

along the length of the diode module. The output of each diode detector

feeds into an amplifier that converts the photocurrent into a voltage.

This voltage signal is then sent to the control/readout module that con-

tains the circuits necessary for data integration and display. The final





5


























-


z w






< U
z
















I I II
i -
r0
= t:

















t0
-;-
























=I-
- -







Lfl


0
















z..
o
Q: V; vj

t-1 < V;c



5 >



3 3




a: !-:M
^-1 -I (




a:
i -----------------
II I1-1
^ ^
a a; K




0;n
&Jo
[_1 _ ^ -- -
1-1 T





oW z
3r ____





6

readout is in a form that, after applying appropriate calibration factors,

indicates the radiation exposure received by the detectors during the CT

scan. The dynamic range of the instrument enables it to measure exposures

in x-ray fields as intense as 100 R/sec and to present the integrated ex-

posure information in three ranges, namely, 0-1, 10, and 100 R.

The term "exposure" rather than dose will be used in describing the

actual experimental results obtained from the dosimeter because the de-

vice is calibrated so that the output of any detector is proportional to

the ionization in air for photons in the energy range produced by CT

scanners.














CHAPTER II
BACKGROUND AND RELATED RESEARCH

Developmental Review


Before a discussion of CT dosimetry can be undertaken, a brief re-

view of the principles behind the development of computerized tomography

should be presented. The major difference between computed tomography

and conventional diagnostic radiological techniques is that CT uses

mathematical recombination methods to produce the desired radiological

image. The mathematical basis for these techniques was essentially

developed by Radon in 1917 (Ka77). Bracewell, in 1956, improved these

ideas and applied them to solar radioastronomy (Br76). The first medical

application of image reconstruction was done by Oldendorf in 1961 and by

Kuhl and Edwards in 1963 (In76). It was not until 1967, however, that

the first clinically useful instrument was produced by Hounsfield of

EMI Limited in England (Ho73).

Although a number of variations have been developed, there are really

only three basic types or generations of CT scanners. The first genera-

tion, of course, is exemplified by the original EMI head scanner. It

consisted of an x-ray source and a pair of sodium iodide detectors (Ho73).

The x-ray source emitted a pencil beam of radiation that passed through the

patient as the source translated laterally (Figure 2). The detectors

moved in synchrony with the source on the opposite side of the patient.

At the end of the lateral pass, the whole assembly rotated one degree.

This rotation and lateral translation process repeated until a 1800 arc

had been completed. Nominally, one scan took about five minutes to

complete.


















I-,

cs
('2


E-


w
5
. E




9

The second generation scanner featured one or more sources with mul-

tiple detectors (Figure 3). The lateral translation and rotation (1800)

were still present except that the angular motion was larger. This im-

provement brought scan times down to about one minute.

Computed tomography's third generation of machines consisted of a

continuously rotating (3600),pulsed, fan beam source with an array of

detectors on the far side of the patient (Figure 4) (Ge77;CT78). Manu-

facturers have used xenon-filled ionization chambers as well as silicon

diodes as the detectors in these devices. A variation on this method

involves placing stationary detectors on the full 3600 scan circumfer-

ence. Scan times have been reduced to as low as one or two seconds by

some manufacturers (Pi78;0h76c).

As the x-ray beam sweeps through body tissue, those photons not

absorbed are detected and a record of their magnitude is stored on mag-

netic discs. Image reconstruction involves the mathematical recombina-

tion of all the sweeps from every angle made through the subject. In raw

form, a computer prints out numbers (Hounsfield units) that are propor-

tional to the average linear attenuation coefficient for a small volume

of tissue relative to water. These numbers are usually converted to a

shade of gray and then displayed on a cathode ray tube as a picture ele-

ment or pixel. The resulting image is an axial view of a body section

as if an individual were viewing a cross section of the patient from

below. The size of the picture matrix may vary from 80 x 80 to 512 x 512

pixels depending on predetermined objectives and the type of machine used.

It should be emphasized that the pixel is really a three-dimensional

concept represented in two-dimensions (X-Y) and its magnitude or shade of

gray is obtained from the summation of the relative attenuation
























M
u




w o

4-
&4
!5


X VU





11













\ \ \ l l I /I
\ \ I I I I I I /
\(\r l ll i l

l ll ll / /





o I








o /




X w








coefficients for a particular tissue volume (Figure 5). The thickness

of the scan (beam width) in the Z-direction is finite and indicates the

geometrical width of a single slice at the center of the patient (Th78).

In current machines, this dimension per slice can be varied from 3 to

about 15mm by collimation at the source and the detectors. The X and

Y-dimensions in the reconstructed image are a function of the matrix size

and of the reconstruction (lateral) distance scanned (Pa76).


Dose Considerations


A discussion of dose considerations in CT scanning procedures must

review a variety of factors that tend to be variable depending on the

manufacturer of the particular CT unit in question and the method of

dose measurement. An overview of some of these considerations would in-

clude paragraphs on x-ray tube characteristics, signal-to-noise considera-

tions in the detecting systems, and a description of present CT dose

measurement techniques.

Radiation Sources

Currently, almost all commercial CT scanners use x-ray tubes as

their radiation source since these devices exhibit high contrast/resolu-

tion scanning with a minimum of shielding problems (Mc77). Radionuclide

scanners are actively being developed (No77;De77a;De77b;Ge77;Bu77;Cho77)

and at least one manufacturer (Un78) has marketed a workable unit. The

two types of tubes that are generally used in present scanners are off-

the-shelf varieties with few, special, CT scan-added features.

Tubes that are used in lateral translate-rotation systems are oil-

cooled types with a fixed anode and line focus (2 x 16 mm focal spot).

They are operated at between 100 and 160 kVp at a tube current so as not

to exceed a 4000 kVp-mA heat loading. These tubes are of the three phase



































0
C 0

z g
+ z







+ 17
E =
Z-1















mrf
ur






C


z1

<
<1


=-







constant potential-type. The output of these tubes is polychromatic with

average energies between 40-50 keV prior to entering the patient. The

effective energy of the beam is generally higher than the mean depending

on the inherent and added filtration incorporated by each manufacturer.

For example, in the EMI head scanner operated at 120 kVp, 32 mA with

4.4 mm Al total filtration, the effective photon energy is 73 keV (Mc77).

Almost all of the manufacturers of sub-ten second rotational scan-

ners (fan beam) use a pulsed, rotating anode x-ray tube (Mc77). These

tubes are air-cooled and have a small focal spot of about 0.6 mm. Pulsed

tubes are operated at kVp's similar to fixed anode sources but use cur-

rents up to 500 mA producing pulses of 2-3 msec in length. Because of

their 3600 angular motion,fan beam devices produce a more homogeneous

dose distribution since every point in the subject is exposed to the

primary beam for similar periods of time.

Radiation Energy

A number of factors modify the x-ray beam energies utilized during

a CT scan (Pe77;Za77). Higher energy beams, in general, are better

transmitted through the subject. Their use leads to better photon sta-

tistics and, usually, lower surface dose. However, lower energy radiation

has characteristically higher attenuation coefficients and, therefore,

slight differences in tissue density can be more readily observed. The

polychromaticity of the x-ray source tends to integrate these effects.

This range of energies in itself, though, means that due to filtration

of the beam by the subject, the x-ray spectrum hardens as the depth of

transmission increases.

In general, radiation dose distributions vary considerably depend-

ing on subject characteristics such as size, shape and density (Jo74).

The clinician would normally scan infants using lower energies such as







100 kVp radiation because of their small cross section (Ag78). This

would produce higher skin doses but lower internal dose levels. Large

or obese adults would receive radiations of maximum energy usually be-

tween 130 and 140 kVp. In this case, the skin dose will be lowered but

internal dose values would increase. Scattered radiation also contri-

butes to the total dose. Forward scatter is relatively independent of

energy, whereas side scatter is greater for lower energies (Jo74). Thus,

even with adequate collimation, the scatter component could contribute a

sizeable fraction of the total dose.

It must be emphasized, that the clinician can greatly determine

total patient dose not only by limiting the number of scans but also by

using the most advantageous energy in a particular situation.

System Noise

The CT image is displayed in different shades of gray on an image

monitor or as printed digital data representing the relative attenuation

coefficients of the material through which the x-ray beam passed. The

image consists of a matrix of pixels corresponding to the density of

volume elements of the scanned material. The ability to recognize two

closely spaced volume elements of similar density depends on the resolu-

tion of the system and on the minimum pixel size. System noise due to

statistical fluctuations in the photon flux delineates the lower level

of resolution and the upper limits of the matrix size allowable (Ba77;

Go77b;Mct78;Ch77;Pa76;Ju77;Ba76).

Generally, to decrease the photon-induced noise significance, the

magnitude of the photon flux must be increased. This can be done by

operating the x-ray tube at a higher kVp, however, that also increases

the internal dose. Brooks and Di Chiro (Br76) have developed an expres-

sion describing the relationship between noise, resolution, and patient






2
dose. The variance, c of the reconstructed attenuation coefficients

due to noise is given by:


2 Y
a (i)c (i)
w h DO


where, S = the attenuation of the scanned material which is a function
of subject thickness and composition, and the quality of
the beam;

w = the pixel width;

h = the slice thickness;

D = the maximum skin dose at a point (generally, at the position
where the scan begins);

y = the depth dose conversion to a point of interest inside the
subject; and

P = the attenuation coefficient of the volume element.

From this relationship, a number of important observations can be made.

Since o (1) is proportional to 1/D reducing the noise by one-half means

increasing the dose by a factor of four (Mc77). Because 2 is propor-

tional to 1/03, decreasing the pixel width by one-half increases the

system noise by 2.7 unless the dose is increased eight times. It can

also be seen that increasing the pixel width causes not only a reduction

in noise but also a decrease in patient dosage. In order to obtain low

noise and high resolution pictures with CT machines of scan times less

than 10 seconds, large photon fluxes must be used, therefore, signifi-

cantly increasing patient exposure.

To keep patient exposures at a minimum, a number of factors must be

weighed by the manufacturer and the clinician. The determination of

patient dose for CT scanners is difficult to obtain especially when cne

considers the variations in manufacturer and use of CT equipment. Brooks

and Di Chiro (Br76) suggest three ways to reduce patient dose:









1. Filter out all photons less than 50 keV to reduce the skin dose;

2. Use only 3600 rotating scanners to distribute more evenly the
skin dose; and

3. Control more accurately the collimation at the sources) and
detector(s).

Methods of dose reduction utilizing improved reconstruction algorithms to

reduce noise contributions and the effects of beam polychromaticity are

currently being investigated (Pa76;Co77;Ch77;McD75;Spi77).


Present Dose Measurement Techniques


As eloquently phrased by Schneider, "for a CT system it is quite

difficult to estimate the surface flux or exposure at each point because

of the complex motion and geometry of the source" (Sc78, p. 98). Because

of collimation and filtration factors the

. beam incident on the surface of the patient is varying in
both flux and quality from point-to-point in a single traverse.
This and other features make the estimation of the flux inci-
dent on the patient difficult to calculate and extremely
tedious to measure. This makes detailed dose calculations
difficult, either by conventional attenuation methods or by
the more sophisticated Monte Carlo techniques. . Even if
the extensive and tedious measurements required to measure the
incident flux on one machine were undertaken no assumption of
generality for other machines, even of the same model, could
probably be made. (Sc78, p. 98)

Although some research into developing computer programs for predicting

dose distributions has been undertaken (Co77;Mc75), "it seems that the

only viable means of obtaining information on internal doses from CT

systems is by the experimental measurement of the doses themselves"

(Sc78, p. 99). In this regard, some dosimetric quantities must be defined.

The characteristics of the x-ray beam and collimator used in a

particular scanner are important in assessing the exposures expected.

Most machines use a single x-ray source that is collimated (at the source

and detector) to produce one or more beams depending on the generation








(i.e., translate-rotate or only rotate) of the machine in question.

Whatever the type of scanner, during an actual clinical examination, a

series of overlapping or nonoverlapping scans are made by moving the

patient couch into or out of the central beam. Thus, some terminology,

as presented by Jucius and Kambic, must be understood:

For a single CT scan, the exposure at any location in the
scanner area is defined as: The peak exposure at any point
in the cross section being scanned. For a series of CT
scans, the exposure is defined as: The average exposure to
the central scan of a series of scans at any point in the
volume being scanned. (Ju77, p. 2)

In reality, to effectively determine the exposure a patient is likely to

receive, a comparison of measured exposure versus image quality must be

undertaken since, if the scan is of poor quality, the clinician will order

a retake.

Dosimetric studies have been reported in the literature by both CT

manufacturers and independent researchers. The manufacturers of CT

scanner systems have provided preliminary information on the skin expo-

sures expected to be delivered to patients by their machines (Sp77;CT76;

Oh76a;Ya78;Ju77). Table 1 shows exposure characteristics of scme whole

body scanners as reported by the manufacturers. Independent investiga-

tors have provided the basis (using phantoms) for dosimetric studies

correlating exposure times, pixel sizes, collimation and filtration usage,

gantry motion, and other factors with the radiation exposure that a

patient would receive (Pe73;Mc74;Mc76;Mc77;Pa76;Ba77;We77;Th78). Three

methods have been used to measure this exposure:

1. Photographic film;

2. Thermoluminescent dosimeters; and

3. Pencil ionization chambers.







19


















U N 0
z 0 ) T
Z N 4- CO I
I (a r. (-
0 4 1








U)
0 CO



0


o
*<-,


0)


0 0 a) n.0
0 4.O 1 r-- 0
Z 4J co mn o
S CO






r- -
a4 4 0a

Q3 Ch >


S 0 0

(CO 0
Xa o
0o Q
0 0\ )
00






00
-4 a
0 n3 H *n 41 r

0 ro 0
'0 C 0

O 0 n
(U1








( C 0

U -4 o -
^i 0 -0 ,i c0)
C 0) 0) (-4
41 LO 1 Q 4 :5
0 0 0 O
(a m E-4 (U 0*0
4-4 -1 (U Q) 04
3 0 C C O X
c '0 0) 11 -4 -H C H
1 0 0 U- -- -4
a a c o e c a








Photographic Film

For many decades photographic film has been used to measure radia-

tion exposures. This is done using calibration techniques that relate

the density (blackness) of the film to the incident exposure (Pr64).

Some advantages of film techniques include:

1. Low cost;

2. Development of an exposure profile;

3. Fast operational response;

4. Relatively large range of exposures measures; and

5. Ready availability.

Some disadvantages in using this method of dosimetry include:

1. Qualitative measurements requiring the use of densitometry for
quantification;

2. Energy dependence at low energies unless filters are used;

3. Limitations to only surface measurements in CT use; and

4. Processing difficulties.

Figure 6 shows an exposure profile of a single and a series of

scans that would be typically produced by a double slice scanner with

contiguous incrementation between scans (Ju77). For multiple scans, as

can be expected, maxima and minima will be observed whose magnitude is

dependent on the separation between increments. Because of the problem

of calibration for selected exposure ranges and its other disadvantages,

film (except for its application as a beam localizer), is not generally

considered useful in CT dosimetry.

Thermoluminescent Dosimeters

Thermoluminescent dosimeters have been used in radiation measurement

applications for over two decades (Fo63). Lithium fluoride is the most

commonly used thermoluminescent crystal. The advantages of TLD include:
































































L'LI SNa








1. Wide range of exposure from a few mR to 105R;

2. Relative energy independence from 20 keV to several MeV;

3. Dose rate independence;

4. Precision as low as 5%; and

5. Geometry approaching that of a point detector.

Some disadvantages of TLD are (TL73;Ce69;Ju77):

1. The destructive readout;

2. The need for an external means of reading the exposure;

3. The time lag between exposure and obtaining the results; and

4. The many dosimeters that are required per measurement (up to
400).

The use of TLD is the most popular method of determining exposure

levels from CT scanners. Most exposure values in the literature obtained

by manufacturers or independent researchers have utilized TLD singularly

or in large numbers. In 1977, the BRH began using large arrays of TLD

to determine exposures delivered to phantoms from CT scanners (Mor77).

This investigation involved the use of a Plexiglas phantom containing

25 holes. Into each hole can be placed a Plexiglas dowel that has been

hollowed-out such that 45, 3 x 3 x 1 mm chips of LiF (TLD 100), can be

stacked side-by-side (Sc78). Thus, after a scan is completed, an expo-

sure profile in the Z-direction is obtained at many locations within the

phantom.

Figure 7 shows the exposure profiles of single and multiple CT scans

using TLD (Ju77;Ya78). A double slice scanner with contiguous incremen-

tation in the Z-direction is illustrated. For a single scan, the expo-

sure to be reported is defined as the peak exposure in the scan cross

section. For multiple scans, an average exposure in the central scan is

desired.














































































In O In


I nSOd In II









Instead of using large numbers of TLD to obtain a Z-axis multiple

scan measurement, it can be shown that all the exposure information re-

quired to report a series of scans is included in a single scan. The

multiple scan profile is merely the summation of single scans that have

been incremented in the normal patient couch movement distance (Z-

direction). The modelling of multiple scan exposures from a single scan

has proven to be advantageous (and accurate) and has shown in certain

instances that the average exposure for multiple scans can be either,

less than, equal to, or greater than peak single scan exposures.

The use of TLD for CT dosimetry provides high resolution dose

information. The tediousness of the technique will limit its practical

utilization as an "in-the-field" method of surveying CT units.

Ionization Chambers

Although ionization chambers themselves are not new, the idea of

using them to measure exposures from CT units is a recent development.

Ionization chambers can be designed to show (Su78;Ju77):

1. Relatively high sensitivity and large dynamic range;

2. Immediate quantitative readings; and

3. Surface or internal exposure readings.

A number of manufacturers have developed ionization chamber dosim-

eters that can be used for CT dosimetry (Ca77;Fa78;Vi78). Most dosimeters

now available are single pencil chambers with air equivalent walls that

have sensitive lengths between 5 and 10 cm. The major disadvantage of

this type of chamber is that an accurate representation of a beam pro-

file cannot be determined during a normal CT scan. However, a recent

paper by Moore, et al. (Mo78) reports the development of a segmented ion

chamber for CT dosimetry. This device would provide a means of obtaining








an exposure profile along the length of the chamber if minimal sensitivity

levels can be realized.

Figure 8 shows the typical response obtained from a pencil ioniza-

tion chamber due to a very narrow x-ray beam. Irradiating a small portion

of the detector is equivalent to exposing the whole chamber to a lower

intensity beam. The actual output reading of the chamber is in R-cm and

is equal to the total exposure due to primary and scatter radiation in

the Z-axis direction.

It appears that ionization chambers can be used to obtain integrated

exposure information from CT scanners. A segmented chamber would also be

able to provide an exposure profile along the length of the chamber.


Semiconductor Radiation Detectors


Most clinical dosimetry devices currently on the market use ioniza-

tion chambers, photographic film, or TLD as their x-ray detection method.

Few manufacturers have considered semiconductors for dosimetry in the

medical setting. However, with recent advances in semiconductor fabri-

cation techniques, their advantages of high speed, linear response, and

good sensitivity have increased the prospects of these devices as medical

radiation dosimeters (Si76;Th77).

Semiconductor Theory

Semiconductor radiation detectors behave like solid-state ionization

chambers (Fo63). Basically, during the process of radiation passage

through a semiconductor substance, charge carriers in the form of elec-

trons and holes are generated. When excess charges move across their

containing medium, they create an electric current due to the presence

of a modifying electric field. For a detector made of silicon, the













































































0
0 0
i-n


3aEWVH3 3AI,








minimum energy to produce an electron-hole pair (i.e., raise a valence

electron to the conducting band) called the band gap energy, Eg, is 1.1

eV. However, due to energy lost in the crystal lattice from vibrational

effects, the actual energy, W, required to form an electron-hole pair is

3.5 eV (Br61).

When compared to an air ionization chamber having an ion pair forma-

tion energy of about 34 eV, it can be seen that silicon has a greater

radiation sensitivity by a factor of about ten. Also, since silicon is

about 1850 times as dense as air, the ionization yield per unit volume

is 18,500 times greater than that found in an air-filled chamber (Fo63).

For a CT machine producing x-rays with a maximum energy of 120 keV,

the effective energy of the beam is about 73 keV. On either side of this

value, there is a distribution of x-ray energies. The primary types of

interactions occurring in the silicon crystal are due to Compton scatter-

ing and the photoelectric effect (Si76). Secondary electron production

(Delta rays) also result from the primary interactions. All of these

effects contribute the charge carriers that are collected to generate the

output current.

The first semiconductor devices produced from silicon were of the

p-n junction type. The n-type region was made of silicon in which im-

purity atoms with excess electrons (donor) were added. The p-type region

had impurities lacking electrons (acceptor), or, in other words, having

many hole sites. Due to the diffusion of charge carriers to the p- or

n-sides of the crystal, a region is formed at the junction of the p-n

materials in which no uncovered charges are found. This layer that is

absent of free charge carriers is called the depletion or intrinsic

region (Du69).









Silicon P-I-N Diodes

The size of the depletion region determines the amount of charge

carrier production that will take place in the device, since for most

efficient charge collection, the range of the ionizing particles should

be less than the smallest dimension of the region. Another factor depen-

dent on depletion region thickness is random noise generation. The junc-

tion capacitance is inversely proportional to the depletion width.

Therefore, to keep noise as low as possible, the intrinsic width should

be large (Jo62). For these reasons, silicon p-i-n diodes with thicker

depletion regions than p-n junctions have been developed and applied to

x-ray photon detection.

There are two major kinds of p-i-n diodes--the double diffused junc-

tion and the lithium (Li)-drifted types (as shown in Figure 9) (B162;

Zi62;Am63;So75). For the double junction type, boron (p') is first

diffused into a high resistivity (~2000 ohm-cm) silicon wafer. Then
+ +
phosphorous (n ) is diffused into the p substrate. The width of the

intrinsic region, i, is determined by the depth of the phosphorous diffu-

sion and, of course, by the width of the wafer. The Li p-i-n is formed

by the drifting of Li (n ) under an external bias through a boron (p )

substrate. This method produced much larger intrinsic regions, on the

order of 1 mm or more, than the diffusion process. The outer electrodes

of the device are made of a thin layer of either gold or aluminum.

Diode Operational Modes

There are two different modes of operation in which a silicon p-i-n

diode can be configured--photodiode or photovoltaic. Photodiode operation

involves the application of a reverse bias, V, to the device as shown in

Figure 10. An intense electric field develops across the intrinsic






29































E- z
c- C

0 0


c: -4



o E






Oya: C



SO 2




a c +





30

































z














.. ____ _^ ^s "








region such that when an ionizing particle passes through producing elec-

tron-hole pairs, the electrons are swept toward the positive (+) elec-

trode and the holes toward the negative (-) electrode (Sc66). The genera-

ted photocurrent, Ig, is modified by a leakage current present in the

device due to the applied bias. The width of the intrinsic region and,

therefore, the sensitivity of the device are also controlled by the bias

voltage. However, increasing V also produces larger leakage currents,

thus, limiting the lowest exposure level allowable for accurate measure-

ment (Jo63).

For use as x-ray dosimeters in the photodiode mode, the p-i-n de-

vice would have to be calibrated at every applied bias voltage. This

would pose a problem in a system requiring the use of a large number of

diodes. Because the current-voltage characteristics of each device

differ, the circuitry required to apply a bias to each device so that

their outputs under the same irradiation conditions were similar, would

be formidable.

The photovoltaic mode of operation involves no externally applied

bias on the diode; therefore, the width of the depletion region remains

constant. Under equilibrium conditions, due to diffusion of charge

carriers, an electric field is generated across the depletion region

(Sc67). The field is directed from the n- to the p-type layers. When

ionizing radiation enters the depletion region, the electrons are swept

towards the n-type side while the holes move toward the p-type layer.

(Figure l). Thus, a photocurrent, Ig, is produced ccnsiszing of the

photovoltaic output current, IE, in the reverse direction through the

external circuit, and a junction current, I., in the forward direction.

For an external load resistance R = 0, the so-called short circuit DC
E











































hv

ELECTRIC FIELD PRCDU(

IRRADIATICN PRCDU(


IRSADIATICN PRCDU(


:ZD 3Y ELECTRON-HCLI DIFFUSION

C:D ELECTRONS


C:D HOLES


FIGURE 1i. ELECTRON-HOLE MCVEMENT IN7
PHOTOVCLTAIC MODE OF OPERATION


r
E








current, IE, in the external circuit will attain its maximum value equal

to I (Sc64).
g
When operated in the short circuit mode, the current, IE, is linear-

ly proportional to the exposure rate as seen in Figure 12. This figure

shows the results of a preliminary test using an RCA C30822 silicon p-i-n

diode. Practically speaking, exposure rates as low as 1.0 mR/min can be

measured using silicon diodes (Ba64). Current levels in the nanoampere
-9
(10 Amp) range have been measured in diodes operated under CT exposure

conditions.

The relationship between exposure and the generated current in the

diode is not a simple function. According to Scharf (Sc67), the value

for I which under short circuit conditions becomes equal to IE, is
g
approximately given as:

q A g (w + L )
I = I (2)
g E (IL + 1)
n

where, q = an electron charge, coulomb;
2
A = the irradiated surface area, cm ;
-3 -1
g = the generation rate of charge carriers, cm sec

w = the width of the depletion region, cm;

L = the average diffusion length of electrons, cm;
n

p = the linear attenuation coefficient of the radiation in
silicon, cm.

The generation rate, g is related to the exposure rate as:

86.9 U (AX/At)
en (3)
o = W( /p)
en air


where, l = linear energy absorption coefficient of the radiation
in silicon, cm-1;

W = average energy required to produce an electron-hole
pair, 3.5 eV;












25



2o









F10








0 IcO 150 200 250

EXICStXRE RATZ (Z:.s)



F:GURZ 12.-~~'' VS E=SURZE R=- F -i

C30822 S:ZCZCC ?-:-'. __CTCOD1-CCE








2 -1
(I /p) = the mass-energy absorption coefficient in air, cm g ;
en air
Ax/At = the exposure rate, R/sec.

Rearranging equations 2 and 3 and solving for exposure rate yields:

I W(en /P) ( L + 1)
AX/At (R/sec) = i ((4)
qA 86.9 p en (w + L )
en n

The integrated exposure, X, measured by the diode is given by:

X = K t (Ax/At) (5)

where, K = an applied diode correction factor obtained experimentally
for each diode;

t = integration time, sec.

Diode Response Modifying Factors

Temperature effects

Temperature effects must be considered when describing the current

generated in a p-i-n diode. A current is produced in all junction de-

vices even when there is no radiation exposure (K173). Without an ex-

ternal bias applied, a dark current, I T, due to the thermal generation

of charge carriers, can be measured. The magnitude, and even the sign

of this dark current, is entirelydevice dependent, but, generally, in-

creases with rising temperature (K177a). Average values for I are in

the 10 12A range.

Variations in the ambient temperature will not affect the radiation

induced current unless the rate of electron-hole recombination is signi-

ficantly altered. Baily and Kramer (Ba64) have shown that for a p-i-n

diode operating in a temperature range from 0 52.5C there is a complete

independence of generated current and temperature.

Any variations in the short circuit current, IE, by temperature are

due directly to changes in the thermally generated current, I (Jo63;Sc64;

Sc71;Pet73;K173;Kl77a;K177b;K177c;K178;Ba64). This effect is related to








variations with temperature in the internal resistance of the p-i-n

junction, R.. Figure 13 shows a diagram of the diode model as it is

connected to an amplifier circuit (K173). For short circuit operation,

the load resistance RE equals zero and, therefore, ID = I. However,

to measure the detector current, an instrument with a finite input

resistance must be attached in the external circuit. Assuming a value

for RE = 0, IE may be calculated from (K173):


R.
I = I I = ID (6)
E D j DR + RE

The current in the external circuit, IE, is always less than ID because

of the presence of R.. This means that the circuit is really not operat-

ing completely in the short circuit mode. To make the effect of R. as

small as possible, RE should have a low value. For example, if R /R =

1,000, the difference between I and I is 0.1%. Unfortunately, R. does
E D J

not remain constant for varying temperatures. As the temperature changes,

R. varies according to the relationship:


-3/2
R. T exp(E /2kT) (7)
3 g

where, T = the ambient temperature;

E = the band gap energy for silicon;
g
k = the Boltzman constant.

The differences in R. observed between detectors is due to the resistiv-
3
ity of the silicon used in their fabrication (K173).

Appropriate values of RE can be used under certain conditions to

compensate for the temperature effects on IE (Sc71). In junction devices,

the sign of the temperature coefficient for the photocurrent will limit

their ability to be compensated. If a positive temperature coefficient





































LL.







UU


U CU
z

U)I U)









U 0~
U) -4
0






U)



0 E3
-4 -
+1 U

-43
li-4
13 CII







is found (experimentally), compensation can be accomplished simply by

the addition of an external load resistance, RE, of appropriate value.

The magnitude of R, can be found experimentally or mathematically, using

ratios of output currents (K178). Should the diode possess a negative

photocurrent temperature coefficient, a thermistor-type device may be

needed for compensation (Sc7l).

CT scanners must be operated at temperatures between 20-250C due

to the sensitive electronics used in their control systems. The device

to be discussed in Chapter 3 would be operated in the same temperature

environment as the scanner. Because of the characteristics of the diodes

used and the method of signal amplification employed, temperature varia-

tions will have a negligible effect on the performance of the CT dosinetry

system.

Directional dependence

A number of authors have researched the directional dependence of

silicon diodes (Wh63;Gu62;Jo62;Sh78). It should be clear that the

geometry characteristics of each diode will dictate its directional sensi-

tivity to radiation. If a diode is operated such that the majority of

the primary beam flux is perpendicular (axial orientation) to the intrin-

sic region, as shown in Figure 14, directional dependence will be minimal.

This is shown in Figure 15 in which an RCA C30822 diode was irradiated

during preliminary testing by the author. In these tests, the diode was

located in a Plexiglas rod that was inserted into a Plexiglas phantom.

Because of the Plexiglas, scattered primary and secondary electrons

generated in the medium (assuming electron equilibrium) will also impinge

upon the detector producing interactions that add to the total number of

charge carriers collected (Ba65). Thus, the actual current I_ found in















0!-
2Z



o5


<








2 -








O-- o
2 <













>--






'U
=>2 z
















m<
zw
S2Uj

cc w





40











0



S20




-10




2700 I-- 90
10 20

nAmp













1800



FIGURE 15. DIRECTIONAL EXPOSURE MEASUREMENTS
ON THE RCA C30822 PHOTODIODE








the external circuit is due to a combination of sources and mathematical

calculation for the exact photocurrent produced is difficult.

Energy dependence

Numerous investigators have demonstrated the energy dependence of

the response (compared to air) of silicon diodes to the incident radia-

tion (Sc64;Ra66;Ba64). Preliminary experimental work has indicated an

increase in generated photocurrent response, shown in Figure 16, for

an RCA C30822 as the energy of the x-ray beam was increased. A correc-

tion factor will probably be needed when different CT scanner energies

are used.

Another aspect of energy dependence involves the use of silicon with

an atomic number of 14 to measure the exposure in muscle tissue which has

an effective atomic number of 7.72 (Mc75). Fowler (Fo66) states that

there is increased response in silicon to the exposure rate as compared

with tissue. The use of Plexiglas, with an effective atomic number of

6.47 to model tissue in the phantom should not greatly affect the number

of Compton and photoelectric interactions since the electron densities of

tissue and Plexiglas are similar (Ph75a;Ph75b).

Radiation damage

Silicon p-i-n diodes are subject to damage from interactions with

x-rays. Damage to silicon is caused by the radiation displacing atoms

from equilibrium locations producing lattice defects which serve as

trapping centers for charge carriers (holes and electrons) (D.e64). Re-

ported effects of these imperfections called Frenkel defects include a

decrease in charge-carrier lifetime, an increase in detector resistivity,

and an increase in charge-carrier generation from a trapping center

(Pr64;Ro62). Coleman and Rodgers (Co64) suggest that because of their























-30





25





-2 20


NO PHANTOM ----
z

15
U



10

5 cm PLEXIGLAS PHANTOM



5







60 80 100 120 140 160


BEAM ENERGY (kVp)


ENERGY RESPONSE OF RCA C30822 PHOTODIODE


FIGURE 16.








larger intrinsic regions, p-i-n detectors should show a larger radiation

effect than p-n diodes. Loferski and Rappaport (Lo58) report no changes

in short circuit currents when p-n junctions were exposed to 140 keV

electrons. Raja (Ra66) indicates no observable radiation damage was

exhibited by a p-n device exposed to 50 kVp x-rays for an integrated

total of two million Roentgens. The probability of sustaining substan-

tive radiation damage to the diodes used in the CT dosimeter are negli-

gible because of the low photon energies used and since the integrated

exposures are not large over the expected lifetime of the diode detectors.

Electromagnetic interference

P-i-n diodes are high impedance devices that are subject to noise

generation because of electromagnetic disturbances. The diodes are sensi-

tive to visible light and radio frequencies prevalent in buildings from

power lines and other electrical and electronic equipment can cause the

generation of spurious signals. To circumvent this effect, a layer of

Teflon-backed adhesive aluminum foil surrounds the diode probe module.

The aluminum layer is connected to the common system ground.
















CHAPTER III
SYSTEM DESIGN AND CONSTRUCTION



The dosimetry system consists of two major subassemblies that form

a completely portable instrument. The block diagram in Figure 17 shows

the electronic interconnections of the system components. The diode

probe is comprised of 25 silicon semiconductor diodes which when placed

into a phantom and properly calibrated can provide accurate X-Y as well

as Z-axis directional exposure measurements. The diodes are operated

without applied bias in the short-circuit DC mode. Contiguous with the

diode probe is a suitably shielded enclosure containing the amplifying

and switching electronics for each detector.

Located at a distance of 20 feet from the Diode Probe/Amplifier

module is the Control/Readout subassembly. The control electronics will

convert the analog amplifier signal into a digital form for subsequent

analysis. The readout mechanism will consist of a thermal printer pro-

viding permanent exposure value storage from each detector as well as

positional identification information.

During a CT x-ray exposure, the photocurrent produced in each diode

is amplified and the signal is sent to the Control/Readout module. An

electrometer operational amplifier acting as a current-to-voltage conver-

ter is used for this purpose. In the Control/Readout subassembly, the

voltage signal is then converted to a square wave of a frequency dependent

on the absolute value of the input voltage. The output pulses are then

integrated in a four decade counter and their total is proportional to









45



















x -




















zz






0
aL











ul
go
I-L
z













C-31
o0





ao OU c.
















aa
o U,
c'4








the exposure received by the diode detector during the CT scan. Follow-

ing the exposure, an automatic dark current subtract is performed. When

readout is desired, a numerical output can be obtained using the thermal

printer.


Diode Probe/Amplifier Module


The Diode Probe/Amplifier module as shown in Figure 18 consists of

the diode probe containing 25 silicon p-i-n diodes and an enclosure hous-

ing the necessary amplifying and switching electronics for each diode.

Diode Probe

The diode probe houses the 25 diodes used as the radiation detectors.

A number of different diodes were tested in this capacity but the RCA

C30822 was chosen because of its small size (5 mm diameter active region),

large cross-sectional area (20mm2 active area), and high photocurrent

output. The C30822 has an n-type substrate that is diffused with boron

to form the intrinsic region. This depletion region is between 100 and

150 um in depth. Experimental values for the C30822 showing the direc-

tional dependence, photocurrent vs. exposure rate, and energy dependence

were presented in Chapter 2. Further tests with other diodes of the

same type showed these values to be similar for the group used in the

crobe.

A unique mounting arrangement has been developed to house the detec-

tors positioned in the probe. Figure 19 shows a cross-sectional view

through the diode probe and Figure 20 is a photograph of the completed

assembly with the diodes emplaced. The probe consists of a section of

12.7 mm diameter Plexiglass rod stock that has been milled as depicted

in Figure 19 by Metal Fab, Inc. (Gainesville, Fla.). The rod was






47




















w
a


0

0 r


0


o iC






a
OI










SC
Ct: / /
/ /

1-! / / '

/ / M
(1< I I C

i / 7 /-

















EACH SIGNAL
WIRE GROOVE
.3 X .3 mm


GROUND WIRE
GROOVE
1 X 1 mm


6---



65 mm




_ L L


I-127 mm




FIGURE 19. CROSS-SECTIONAL VIEW OF
PLEXIGLAS DIODE PROBE









49


















































I:
w

0


0
0
aE:

HEEEEEEEEEEEEEE
::i iiiiiiiiiiiiiiiiiiii




EEEEEEEEEEEEEE








designed to allow the portion extending from the amplifier enclosure to

be placed into a 125 mm thick plexiglass test phantom for radiation

measurement purposes. Inside the amplifier module, the probe extends to

the output socket connector.

As seen in Figure 19, a portion of the circumference of the rod was

milled off and then 25 equally spaced grooves running longitudinally

along the major axis were cut into the remaining circumference of the rod.

Into the center of the previously machined top portion, a deep groove

120 mm in length was milled. It is into the center of this groove that

the 25 diodes were placed with the separation between devices shown in

Figure 21 for one-half of the symmetrical array. Into each groove on the

outer circumference of the rod, a hole was drilled at a point 1 mm to one

side of the position of the diode placement. This was done to allow

signal wires to be brought from the n-side of each diode to the amplifier

module. An additional groove also has been milled parallel to the deep

center groove along the entire length of the rod to accommodate a wire

carrying the common ground return from the p-side of each diode.

Because its low atomic number would minimally perturb the x-ray

flux, aluminum wire 0.008 inch thick (EC-soft) was obtained from the

California Finewire Corp. (Grover City, Calif.) to carry the current sig-

nals from each diode. The aluminum wire was attached to each diode using

a carbon-based epoxy Eccobond 60-L obtained from Emerson & Cuming, Inc.

(Canton, Mass.). This carbon epoxy was chosen because it contributed

negligibly to the scatter radiation generated in the probe. Following

diode emplacement and wiring, a silicone potting compound Eccosil 2CN

(Emerson & Cuming, Inc.) was used to encapsulate the detectors to exclude

the presence of air in the probe.


















































































C




Um











Um


Ur)








To complete the probe, a Plexiglas top section was milled that

closely replaced the original portion that was previously removed. Two

layers of a Teflon-coated aluminum foil were then folded around the probe

and connected to the system ground. Finally, the entire portion of the

probe not placed into the amplifier module was inserted into a section

of 19 mm diameter Plexiglas tubing that provided structural strength and

environmental protection.

Amplifier Module

The amplifier module is directly connected to the diode probe. A

portion of the probe extends into the module and is used as a guide for

the output wires from each diode. The amplifier enclosure is a square

box 100 mm on a side and is made of .063 inch aluminum sheet. Figure 22

shows the general layout of the amplifier module. The three printed cir-

cuit boards containing the amplifier electronics are visible. The socket

at the rear of the enclosure allows output of the amplified signals from

each detector and input of power and switching control signals.

Operational amplifier characteristics

An operational amplifier (op amp) is a directly coupled amplifier

employing feedback to control signal gain potentials (St76). The device

is constructed of many transistor amplifiers and is available in mono-

lithic or discrete forms. The op amp can operate as an amplifier, con-

troller, or signal generator over a wide variety of frequencies including

DC operation. For its present purpose as a current-to-voltage converter,

the op amp is the logical choice because of its high stability over time,

temperature variations, and gain changes.

The ideal op amp possesses the following characteristics:

1. Differential voltage gain = c;

2. Common mode voltage gain = 0;

































I,
, mi -ii

" > .>
'V ,'
' NI


-. /








3. Bandwith = -;

4. Input resistance, R. = o;

5. Output resistance, R = 0;
o

6. Perfect balance, output voltage V = 0, when input voltage, V.
= 0;

7. No parameter drift with temperature changes; and

8. Input noise = 0.

None of these ideal parameters could or even need be obtained in actual

practice. For the purposes of the amplifier in this project, compromises

can be made so that only a few of these characteristics need to approach

the ideal.

Op amp operating configurations

There are two basic configurations in which an op amp, using feed-

back, can be described--the noninverting and inverting modes. A brief

description of the noninverting mode will be given followed by an analysis

of the inverting-type since it will be used for the amplification in the

dosimeter.

The noninverting op amp (Figure 23) uses the noninverting lead (+)

for signal input and the inverting connection (-) for feedback purposes.

The output voltage, V in this circuit is in phase with the input vol-

tage, V.. This configuration has many important applications including
1

its use as a voltage follower allowing for isolation of signal source and

load (Gr71).

The inverting configuration (Figure 24) requires the input signal to

be fed into the inverting (-) input while the noninverting input (+) is

connected to common signal ground. There are two basic operating rules

(ideally) for this setup that simplify circuit analysis. First, the two

op amp input terminals draw no bias current. Secondly, the voltage

between the input terminals, V equals zero (virtual ground). From this
x






55



























o


z
0
















t7'
/\ C
z
z



II,


2














iT








it should be realized that the current flowing through RI is II and is

numerically equal to (V. V )/R The current flowing through Rf is

equal to (V V )/R Since the inverting input has infinite input
x o f

resistance and ideally draws no bias current, all of Il must pass through

R Therefore, I = If, and since V = 0,


V. V
I = = I and, (8)
1 R R f


V = I R (9)
o I f

Op amp error factors

There are a number of factors that could influence the operation of

the amplifier in this circuit. Ideally, these parameters would not

exist, but in the real world they can produce errors and, consequently,

their presence must be known and then minimized. Op amp designs are con-

tinually being improved leading to more ideal devices.

Input offset voltage (V ). When the input voltage, V., to an op

amp is zero, ideally its output is also equal to zero. In most cases,

however, a finite output voltage can be measured even when the input is

zero. The input offset voltage is that potential applied to bring V

0. The value of V is temperature sensitive and is usually defined in
os

relationship to temperature (i.e., 4V/OC).

Input bias current (Ib). The input bias current, Ib, is the average

of the two op amp input currents. Ideally, Ib = 0, so that possible in-

terferences with the feedback circuitry are removed. The Ib of the op

amp should be minimal since this, together with noise properties intrin-

sic to the device, determines the lower limit of current measurements.








Input offset current (I. ). The input offset current, I. is the

numerical difference between the currents going into both inputs of the

op amp. This parameter is a function of temperature so circuit component

compromises must be made for operation in the temperature range of inter-

est (Mi72).

Slew rate. The slew rate of an op amp describes the amplifiers

ability to change its output voltage when a step input voltage is applied.

This rate of change is usually given in volts per microsecond (V/ys).

Circuit description of the amplifier module

Figure 25 shows the amplifier circuit that is used for each of 25

diodes in the dosimeter. A number of op amps were tested but the Analog

Devices, Inc. (Norwood, Mass.) AD 515LH was chosen for this application.

The AD 515LH exhibits a low input bias current of 75 femtoamp maximum, a

typical input offset voltage of 0.4 mV, an offset voltage drift of only

25 IV/oC, and a slew rate of 1.OV/Ps (An78a).

The amplifying circuit is designed to function as an exposure rate

measuring device. When used to monitor the radiation output of CT scan-

ners, this operational mode causes some problems. It is generally ob-

served that for CT units to perform very rapid scans, a large photon flux

must be emitted to obtain sufficient information in a short time period.

This means that the silicon diodes will be required to measure large,

almost instantaneous photon fluxes from fast scan (pulsed) machines as

accurately as they would measure smaller magnitude fluxes from slower

devices (continuous flux). To ensure a reasonable dynamic response in

the amplifier, two input voltage ranges have been incorporated into the

feedback loop. The equation that describes the operation of the current-

to voltage converter is:



















































1
L_


< z
3H
U n:







V = I R (10)
o D f

where, V = the output voltage;
o
ID = the diode produced current; and

Rf = the feedback resistance.

Because the amplifier is used in the inverting mode, the output voltage

is the opposite polarity of the current flowing into the negative inver-

ting input.

To facilitate these two input ranges, resistances Rfl and Rf2 are

located in the feedback loop. Feedback resistance Rfl is permanently lo-

cated in the circuit and defines the lower maximum input exposure rate

range which has been defined as 2 R/sec. Resistance R2 which has a
.L2

value fifty times less than Rfl is used to measure the maximum exposure

rate of 100 R/sec and can be connected into the circuit using a General

Electric (Syracuse, New York) HllF3 opto-isolator. The switch used to

perform this function is located on the front panel of the Control/

Readout module. The opto-isolator exhibits an "off" resistance measured

at about 5.6 x 10 11, and an "on" value of about 300Q thus, not affect-

ing the operation of the circuit. Resistances Rfl and Rf2 are in par-

allel so the effective feedback resistance when the opto-isolator is

activated equals:


fl f2
R = (11)
eff R R
fl + f2

This effective resistance is slightly less than the value of Rf2 but the

error is not important in circuit operations.

If it is assumed that the maximum voltage output of the amplifier

is 10 volts, then a calculation for the desired feedback resistance can

be made. Previous experimentation with diagnostic x-ray machines








has shown an average diode current/exposure rate of about .115 namp/mR-

sec. This measurement was performed using the C30822 diode located in

a Plexiglas phantom at 5 cm depth (assumed average tissue depth). Using

this value and rearranging equation (8), for the maximum exposure rate of

100 R/sec (105 mR/sec) gives the following value for Rf2:

10 V
R = V (12)
.115 namp/mR-sec x 10 amp/namp x 10 mR/sec


= 8.69 x 105

Since the lowest exposure rate range must have a resistor 50 times as

large as Rf2, then Rfl should equal 4.35 x 10 7. The actual values

chosen were: Rfl = 5 x 107 and Rf2 1 x 106 The resistors, obtained

from Eltec, Inc., (Daytona Beach, Fla.), were high precision devices

(2%) that exhibited low drift with temperature (E178).

A 5 pf polystryene capacitor, C is located between the inverting
s

input and the output terminals of the op amp. The capacitor is used to

stabilize the feedback loop because of the instabilities that may arise

due to the presence of a parasitic capacitance of 2-5 pf located at the

inverting input of the op amp (An78a;St76).

The output of the amplifier must pass through 20 feet of shielded

multi-conductor cable before the signal reaches the Control/Readout

module. Because the capacitance in the cable (Belden 8776) is about 40

pf/ft, resistor R is located at the output signal of each amplifier to

set up a time constant limiting the output current of the device.

Belden 8776 is a 15 pair individually shielded cable that is used

to interconnect the dosimeter's two modules. Twenty-five conductors

are used to carry the output from each amplifier. Two conductors supply

12 volts to the AD 515LH op amps and another line is used to turn the








opto-isolators on through current limiting resistors. The shields locat-

ed. on each pair of wiresare connected to the common system ground.

Thirty-seven pin Amphenol plugs and sockets are used between both modules

to permit precise and positive transfer of signals.

All of the electronic components comprising the amplifier module are

located on three identical double-sided epoxy-glass printed circuit

boards. The boards were designed by the investigator but fabricated by

Technetronics, Inc. (Casselberry, Fla.). Figures 26 and 27 show the com-

ponent and wiring sides of the boards, respectively. Two boards contain

nine separate amplifier sections while one board holds only seven ampli-

fier circuits. Nylon integrated circuit sockets are employed for op amp

emplacement. Because of the low currents to be handled by the amplifier,

shielding and guarding techniques must be emphasized. Guard rings sur-

rounding the inverting input (on both sides of the board) and tied to

ground are necessary to prevent leakage current errors. Since high impe-

dance circuits are conducive to picking up stray AC signal in the environ-

ment, the amplifier boards are completely enclosed in a grounded metal

box.

The output current from the n-side of each detector is directly

routed to its complementary amplifier input (inverting input). The loca-

tion in the op amp socket and the printed circuit board through which the

inverting lead passes has been drilled out, thus ensuring that this high-

ly sensitive input does not come in contact with any material that could

lead to the generation of noise signals in the amplifier. The feedback

resistors Rf and Rf2 and stabilizing capacitor Cs are directly connected

to the inverting lead. The opposite ends of Rfl and C are connected

to the socket pin that contacts the output pin 6f the amplifier. Resistor

































































FIGURE 26. COMPONENT SIDE OF AMPLIFIER BOARD

























I .^i ^Jimii
lwor
Blil10


7 1 -i ,3RwIJI
- a* a-


FIGURE 27. WIRING SIDE OF MPLIFIER BOARD








Rf2 is attached to an input of the opto-isolator for each amplifier.

The opto-isolator output is sent to the amplifier output pin. Resistor

R is mounted on the printed circuit board and the final output from each

amplifier is sent through it to a 37 pin socket which is located at the

rear of the amplifier module.

Power supply bypassing is used on the circuit boards to eliminate

high and low frequency noise. A 22 4f tantalum electrolytic capacitor

and a .01 pf ceramic capacitor combination is present for every three

amplifiers on the three boards.

Amplifier response

An amplifier does not instantaneously show a voltage output when an

input signal is applied. Likewise, the amplifier output does not imme-

diately drop to zero when the input signal is removed. If this were true

(an ideal amplifier), the equation describing the area under the curve in

Figure 28a would be simply:

TOTAL AREA = V t (13)
0 0

This represents the integral of the voltage signal V over time t and

is proportional to the x-ray exposure received by the silicon diode de-

tectors.

In the amplifier circuit described previously, a finite feedback

resistance and capacitance is present in all situations. Figure 28b

illustrates an approximation of the actual curves describing the rise and

decay exponentials of the amplifier. Modelling this curve produces two

integrals:


AREA = V [f o (1 e ) dt + et^dt] (14)
0 0
to

= V [ t T + 2Te-to/T (15)
o o



















vo 0



0 to



a.) IDEAL AMPLIFIER RESPONSE


















0 0

) L MPLFIR RESPONSE

b.) REAL AMPLIFIER RESPONSE


FIGURE 28. DIODE DETECTOR AMPLIFIER RESPONSE








In these equations, t is the scan time of the CT unit and T is equal to
o

the feedback resistance multiplied by the feedback capacitance C If
s

C = 5 pf, R = 50 x 106 and the CT scan time is one second, the area
s fl

under the curve from equation (14) equals .9998 V This value when com-
o

pared to an ideal amplifier response of 1 V shows only a .025% differ-
o

ence. If Rf2 = 1 x 106 a value of the total area of .9999 is obtained

which differs from the ideal condition by less than .001%. Thus, the

amplifier response will accurately follow the ideal case under the condi-

tions the dosimeter will most likely face during measurements.


Control/Readout Module


The second major component of the dosimeter is the Control/Readout

module. This assembly is connected to the Diode Probe/Amplifier module

via a 20 foot multiconductor cable.

The Control/Readout module contains the data acquisition circuitry

needed to process the signals produced by the diode detectors. This

module consists of 25 absolute value circuits (AVC), 25 voltage-to-

frequency converters (VFC), 27 four decade counters, a voltage comparator,

a clock up/down count section, and the circuitry necessary to interface

the counters with the thermal printer used as the final readout device.

Input to this module is via a 37 pin socket that accepts the multicon-

ductor cable carrying signals from the diode amplifiers. A 30 conductor

ribbon cable is attached to this socket allowing transfer of signals to

a specific AVC, VFC, and decade counter group.

All of the circuitry associated with the Control/Readout module is

located in a portable enclosure. Five Vector plugboards with 44 pin edge

connectors are used to hold the integrated circuits and other discrete









components in the data analysis system. Four of these boards are identi-

cal and each contains six AVC, VFC, and decade counter sections as seen

in Figure 29. The fifth board, seen in Figure 30, holds only one AVC,

VFC, and counter section but houses the circuitry necessary to interface

the data acquisition components with the thermal printer. Wire-wrap

techniques are used to interconnect all of the circuit components.

Unless otherwise noted, CMOS integrated devices are used in the dosin-

eter. since they exhibit low power dissipation levels, good noise immunity,

multiple levels of voltage operation, and high speed.

Two power supplies, one providing +5 V, 12 V and the other 15 V,

are available to supply the power needed by the dosimeter.

The Absolute Value Circuit

The addition of an absolute value circuit (AVC) in the dosimeter

circuitry was necessary because negative voltages could be produced by

the diode current amplifying electronics. Normally, the photovoltage

applied to the VFC is positive; however, a positive dark current from

each diode or a positive op amp input offset voltage could produce a nega-

tive voltage input to the VFC. This cannot be allowed so an AVC is

included in each of the 25 data analysis sections to ensure always posi-

tive voltage input to the VFC.

The operation of the AVC can be understood by referring to Figure 31

(Gr78). Each circuit is located between a corresponding diode detector

op amp and a VFC. The circuit uses a bipolar op amp (L1 1458) and an

n-p-n (2N3904) transistor as its active components. In addition to pro-

ducing a positive voltage output for either polarity input, a sign bit is

available at the collector of the transistor. A positive input voltage

to the op amp produces a high sign bit while a negative voltage produces

a low bit value.

















s 2 "















N a. a. 0 p K I
r. 8
S o















I-1 f 0







a. u
> 2
E.- > 0 0C a
zo M U v j t,
n: i. uP
U= 0 m 0 z
a o
w E Z
C 0



E- E- E- E-
C
- 0% M 0 C
c-4 V ' r- ' n -i ( (


E 8 8 0% 8 8 8
0

c m n *

C u
q.



' < o u a u a =
<~;EE
3

DEE-


wwl~
w II
w w]

wwwI


S 1







I I [ El
w w~ wl


N-**


7







































































S CO








During operation of the dosimeter when a positive photovoltage is

produced by the AD 515LH, the LM 1458 is disconnected from the circuit

because a negative output turns off D2 and the voltage is transferred

directly to the VFC via resistors R1 and R2. If a negative voltage is

developed by the AD 515LH, the output of the LM 1458 becomes positive

turning on D2 through Zener diode D Diode D1 is included to ensure

proper operation of the polarity bit transistor. Transistor Q1 is turned

on when the output of the LM 1458 is positive, thereby indicating a low

output at the collector.

Because the input resistance is unequal for positive and negative

voltage signals, a reversal voltage error of significant magnitude could

be observed if the AD 515LH source resistance is greater that 10 or is

not constant. This problem, however, is not faced because the 2kQ resis-

tor located at the output of the AD 515LH is several magnitudes larger in

resistance than the op amp output resistance over the voltage range of

interest. This 2kS resistance is simply subtracted from the value that

1 would normally show if unity amplification is desired. The value of

R then becomes approximately 8kQ. Individual matching of input resis-

tance is required for each amplifier circuit.

The Voltage-to-Frequency Converter

A VFC is used to transform the analog signal from each diode ampli-

fier into a digital form that can be processed and provide an equivalent

exposure reading. A VFC and counter combination is used in the dosimeter

to perform signal integration because of the expense and stability prob-

lems that were encountered when an amplifier and capacitor analog inte-

grating design was first contemplated. A VFC possesses an inherently

digital readout and integrating this digital signal produces a count

total that is equivalent to the area under the voltage curve generated by

the diode amplifiers (Te73).








The VFC used in the dosimeter is an Analog Device AD 537JD. This

monolithic bipolar device consists of an input amplifier (buffer), a pre-

cision oscillator system (current-to-frequency converter), an internal

reference generator, and a high current output stage (driver) (An78h).

Figure 32 shows the circuit configuration of the AD 537JD used in the do-

simeter. A full-scale input of 10 volts is determined by Rscale The
scale

full-scale frequency is set by external components Rscale and C (.001 uf

polystyrene capacitor) and can be calculated from the relationship:

F = Vm /10 C1 Rscale
max 1 scale
Figure 32 shows that Rscale has three ranges that are determined by

three fixed (R ) and three variable (R ) resistors. These three ranges

are utilized to set the frequency output for a specified input voltage

and can be switched into the circuit using a three position, 25 pole

rotary switch located on the Control/Readout module front panel. The

three exposure ranges of 0-1, 10, and 100 R are defined by the proper re-

sistance combination as indicated in Table 2.

Because of the input voltage characteristics of the VFC used in the

dosimeter, two overscale ranges have been incorporated into the device

as shown in Figure 33. Since the 0-1 R range is defined only to 1500 mV,

an LED located on the front panel is turned on when an LM 339 comparator

is tripped. The same LED is operated by another comparator that is acti-

vated when a 10 V level is exceeded on the 10 and 100 R ranges. Control

of the particular ccmparator used for an exposure setting is determined by

the front panel rotary switch.

The output of the VFC, taken from pin 14, had to be individually

calibrated using R for each full-scale range. The resistor RT was also

individually selected to trim the input offset voltage of the input op


















c
c


m Ln
(N



I'






















U2



0


04

x







0


o
O 0
0 0
I I Ir

0 0 0











*r1


























O o O




I-

















o o a


Ul
0




0

-4
M00

4J
O


-4
u
1)
r-
U,

































2








Z

t:
a


Y






C:








amp for optimum frequency calibration in all ranges. The final output

signal is a square wave of 5 V magnitude with a frequency proportional

to the input voltage. Table 3 shows the variations in the output fre-

quency over the input voltage averaged from all 25 VFC. From these re-

sults, it can be seen that there is a considerable overlapping of ranges.

The correct integral exposure range and exposure rate range for a parti-

cular CT unit and exposure time will have to be experimentally determined

at the time of measurement. This question is more fully analyzed in

Chapter 4, SYSTEM OPERATION, CALIBRATION, AND EXPERIMENTAL RESULTS.

The Decade Counters and Associated Circuitry

The decade counters that are used to integrate the digital signals

from the VFC are the central controlling components in the dosimeter.

Each counter and its associated circuitry is responsible for data acqui-

sition, storage, and display. Figure 34 shows a block diagram of the sub-

systems comprising the decade counter and associated control and readout

circuitry. The dosimeter uses a Datel (Canton, Mass.) DPP-7A1 thermal

printer for automatic sequential data printout and permanent exposure

value storage.

The operation of the dosimeter is controlled by diode 13 (the center

diode in the detector array) in that when the voltage from amplifier 13

exceeds a preset level (depending on dark current and offset voltage

considerations), a voltage comparator turns on allowing the decade counter

to accumulate pulses. The comparator also switches on the count up/down

clock that is used to indicate scan time length and to control background

subtract following completion of the scan.

The counters used in the dosimeter are Intersil (Cupertino, Calif.)

ICM 72171JI CMOS up/down four decade units (In78). A pin connection


























C:












U












0-

= 0

00








H




0;





a


C C 0 0
C 0 r 0
_ i -
n oo o
0 0 0 C


















O 0 0 1


N Li 0 0














LI O
c 0 0
0 L 0
c o c
r^; in o






in i-







c c c


Li C C N LI


c + i















-i C C 0 C
- O. 0 0 i

L 0

















S0 0 l 0
















N L 0 0
SC


0
0

















~ci








C:


0 0 0









0 0 0 0








0 C 0
c O
O C 0 C









0 00 0







C C C C
0 0 0 0
















0
0





79














a A


W2 4

















UOU




U C
E-,












0
U
.-- --.
I-----* "









diagram for the counter can be seen in Figure 35. The counter provides

-ultiplexed 7-sec-ent ccton anode cutpu=s that can directly crive LD's

or the prinzhnead of the thermal printer used as the cutput s-orage device.

The internal multiplexing (. ux) oscillator can be overridden by signals

orcduced in the herm-al printer, thereby controlling digi display se-

quencing. Display blanzing and unblank-ng is accomplished using digital

logic circuitry in coordination with control signals from the printer.

Twenty-five counters are used to integrate signals from the V-EC

specific to each for readout by the prin er. Counter 26 is connected in

zar=11el with counter 13 and is used -o acnizcr -he outzut of diode 13.

A National Semiconductors (San-a Clara, Calif.) NS3 7882 four decade

=u-l-iplxed LETD display is found on the front p-anel of -his mcdule to ex-

hiit -he integratedC cutput of counter 26. This is srcvided to allcw a

"real tine" display of the scan data at the center diode (13) so that

exposure rae an d ata overrange i'ormation is i-' ediately attainable.

A pin-by-pin description of t-e operation of the counters n associ-

ation with the control circuit-y will now be presented.

The decade ccunter-oin 8. Pin S is used to input the square waves

from the VTC specific to each counter. The count inpt pin is provided

with a Schai-=: trigger to allow operaticn in noisy envirc ments and to

prevent aul-iple -riggering.

The decade counter-pins 2, 9, 10, 14. The use of these pins is con-

trolled by the ccmarazor, -he absolute value circuitry, and the count

up/dcwn clock. A diagram of the operation cf nhese pins can be seen in

Figure 36.

Figure 37 shows the comparator circuitry that- is used 'o sense the

switchinc c-.nt which turns on the clock counter (27) and the -in-erating
















ZERO


COUNT INPUT


STORE


UP/DOWN


LOAD REGISTER





SCAN OSCILLATOR


RESET


4 MSD

30 MSD


20 MSD


MSD


+5V


DISPLAY BLANK


G


B


GROUND





F


D


A


C


FIGURE 35. PIN CONNECTION FOR ICM 7217 IJI









83






<
< z

Q(r)


&I
<


























CN-
CN



o0










S-n >
cILn











C
0 00























0N
Lnn
Lnn







W LO;





< 0~

rn u
mi,



+ m








counters (1-26). The ccmparator uses hysteresis to assure proper switch-

ing in a potentially noisy circuit. Positive feedback is utilized to set

up a finite input voltage difference between the on and off states of the

comparator. A Datel aDM-4100L digital voltmeter measures in mV, either

the voltage from diode amplifier 13 (the CT scanner may be on or off), or

the reference voltage, Vr, that is used to set the trip point of the

comparator depending on the position of switch S1 (located on the front

panel of the Control/Readout module). The reference voltage is taken from

an output of the DVM4 and can be varied by a potentiometer located on the

front panel. When the comparator is on, a -5 V signal level is produced

for control purposes.

Because pulsed beam CT scanners do not produce a continuous photon

flux, a 4098 dual monostable multivibrator delay circuit has been incor-

porated into the circuit to ensure that the comparator stays at a high

level between x-ray tube pulses. An OR gate enables the circuit to func-

tion in a continuous source beam as well as in the pulsed mode.

In Figure 38, the circuitry for the up/down clock counter logic is

illustrated. An NE 555 oscillator used as an stable multivibrator pro-

ducing a continuous 10 Hz square wave operates as the system clock. The

clock pulses are directed into pin 8 of counter 27. The 7-segment counter

output is displayed on a four decade LED display (NSB 7882) and indicates

the length of tine of the CT scan. The LED display can measure a scan

time up to 999.9 seconds.

Figure 36 shows the circuitry configured around counter input pins

2, 9, 10 and 14 for counters 1-27. Pin 2 cn counter 27 and pin 9 on

counters 1-26 operate in conjunction in order to store data in the output

latches of each 7217 for later retrieval. When the comparator is tripped






85















oz
uI t




>
E-4












CC






00
E--


>
ow III I I I
lu uE






























Cv








at the beginning of a scan, latches B and C use the comparator signal an

the high level from the Zero (pin 2) output from counter 27 to allow con-

stant updating of the output latch in counters 1-26. Pin 2 produces a

low level when the counter content is at zero. When the scan is com-

pleted, and automatic background subtract is performed after which the

counter output latches are set (i.e., data stored) in readiness for read-

out. Resetting of each counter is performed through Latch A using a man-

ual switch located in the front panel. The reset switch is operated

through a 4043 latch to make it bounceless. Counter 26 requires direct

data output into a four decade LED so the addition of a 4077 and a 4049

to its pin 9 store input allows the display to be erased (by the manual

reset) before another scan is started. Pin 10 sets the up or down count

mode. When a high polarity AVC sign bit is present and the comparator

is on (corresponding to a positive photocurrent), the up count mode is

chosen. The comparator goes low when the CT scan is completed. If the

voltage from diode amplifier 13 is now negative (indicating a positive

dark current or offset voltage), the background subtract control will

cause counters 1-26 to count Vp, thereby compensating for the constant

negative signal error. The reverse occurs if the quiescent diode ampli-

fier voltage is positive. The output of Latch C is connected through a

4049 to an LED indicator located on the front panel that actuates when

the background subtract function is completed.

The decade counter-pins 11, 13. Pins 11 and 13 of counters 1-25

are used along with some associated circuitry as the control interface

between the data analysis system previously described and the thermal

printer. Figure 39 is a diagram of the counter/thermal printer inter-

face components. Pin 13 allows outside overriding of the onboard Mux






87








z
-4




-4
z








>4

E-


zz
t-44
-4 -41'























CN.
U.,
RI z
-4
-43










Z QA
tom o: 4

w cz
Y I 1 ~a
-ro z lT IC.)

uC


n~C;
a Z~o I -
U.,








scan oscillator of each counter. The thermal printer used in the do-

simeter produces a number of internal signals that are used to provide

controls for the operation of the data analysis circuitry. A Shift

Buffer signal taken from the printer is inverted and used to trigger the

Mux sequencing input (pin 13) that selects what decade digit is to be

accessed next. The Shift Buffer signal is high for 2 msec and low for

25 msec. This provides adequate time for data readout from a particular

digit which occurs during the low state between Shift Buffer pulses.

Figure 40 shows the timing chart for this operation. The entire sequence

of print startup is controlled by a manual Print signal caused by the

actuation of a switch located on the thermal printer front panel. The

operation of this print switch begins the automatic printout of exposure

data from each counter. The Print pulse is generated during each succeed-

ing print cycle using a "handshake" arrangement with a Busy signal from

the printer that remains high during printout and paper advance activi-

ties. A delay circuit using a 4098 dual monostable multivibrator facili-

tates this operation. At the end of the first cycle, the Busy signal

goes low causing a pulse delayed by about 220 msec to be sent to the re-

mote Print control in the printer, thus allowing automatic sequencing to

counter 2 and subsequent printout of the remaining counters. When the

twenty-fifth cycle has been completed, an inhibit signal from the display

multiplexing system (to be discussed) prevents further Print signals

reaching the printer unless the manual print switch first is engaged. A

switch is present in the Busy line to prevent premature operation of the

printout circuitry.

The main feature of the ICM 7217 that ensured its choice as the

counter for the system is that provided by pin 11--the load register (LR)











































U




ci
















>4
t12


I

















a


^L4
3--
-


0


















z
O









0




















cn
-


0
!-
c:

















___
vl

3&


e1

o
m








control. When the LR input is forced to ground, the 7-segment display

driver outputs (pins 15-19, 21, 22) are disabled, the Mux oscillator is

inhibited (including any overriding signals), and the internal digit

select Mux counter is reset to the most significant digit (MSD). This is

a crucial control because the thermal printer begins printing from left

to right (i.e., starting with the MSD). Therefore, when the printer be-

gins the print cycle, the data value for the first character printed

(the MSD) must be present at the printhead input. This synchronizing

operation is accomplished using a 4017 decade decoder, a 4066 bilateral

switch, and the Shift Buffer input signals.

After initial system reset, activation of the manual print switch

causes the Shift Buffer signal to be sent by the printer. At this point

pin 3 of the 4017 is high inhibiting the I'-ux oscillator through the 4066.

When the first pulse from the shift buffer inputs into the 4017, pin 3

goes low, turning off the 4066 (releasing the LR pin) and allowing se-

quencing through the digit select counter starting with the MSD. The out-

put drivers of the particular digit chosen are only active when the Shift

Buffer pulse is low. The second MSD is selected by the second Shift

Buffer signal. The sequencing continues until the fifth Shift Buffer

pulse is input-d into the 4017. Although the printer has the capability

of displaying six digits, the last two digits have been externally blanked

since the counter output coversonly four decades. Pulses 5 and 6 from

the Shift Buffer perform no function. Pulse 7 resets the 4017 to zero

and holds the LR input of all ccunters low, -Cherebv resetting the Mux

counter to the MSD position. This sequence reoccurs during each cycle

to ensure proper prin- coordination between the counters and the zhermal

printer.








The decade counter-pins 23, 15-19, 21, 22. The control that enables

the 7-segment output from each counter to be transferred to the printer

is pin 23--display blanking. Figure 41 shows the circuitry of the Mux

system enabling or disabling the blanking control of all 25 counters.

Normally, all outputs of the three 4017's are low. A 4049 is provided

to invert this level causing blanking of all counter 7-segment outputs.

During the initiation of the print cycle by the manual switch, the printer

Busy output goes high clocking a pulse the length of the print cycle into

the 4017 decoder. This signal turns on the first channel output which

after passing through the 4049, unblanks the display drivers in readiness

for printout in conjunction with the Shift Buffer (LR control) signals

previously described. After completion of the first print cycle, the

Busy signal goes low initiating a 4098 delayed pulse (see Figure 38) to

go high 220 msec later for 1 msec providing a Print command that now

automatically reoccurs during each of the remaining 24 print cycles. At

the completion of the 25 print cycles, a 4098 delay resets all 4017 coun-

ters allowing another printout of all decade counters should that option

be desired. Print inhibition is obtained after all counters are examined

(see Figure 39) using a 4043 and a 4077 to prevent the delayed Print

pulse from reaching the printer after the Busy signal goes low at the end

of the twenty-fifth cycle.

Because of the blanking control available on the"7217, like 7-segment

outputs (segments a-g) from each counter can be tied together so that only

7 lines containing the segment information are directed into the printer.

Each line feeds into a 7405 hex inverter which, in turn, drives a tran-

sistor triggering the corresponding segment in the printhead.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs