.
TOF
ORNL P
1858
EEEEEEE
11:25 .4ILLE
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
:os
ORNA P-1858
Conf- 6512/0-1
MASILA
.
.
JAN 10 1999

KELDASID FOR ANNOUNCMENT
QUANTIFICATION OF SCAN RECORDS
C. C. Harris
IN NUCLEAR SCIENCE ABSTRACTS
Oak Ridge National Laboratory
Oak Ridge, Tennessee
Clinical scanning 18 sufficiently valuable that many persons are ex-
pressing a desire for scan information in excess of that provided by a
qualitative portrayal of the distribution of a radioactive tracer. It 18
becoming quite comuon for a physician interpreting a clinical scan record
to ask, "How much less was the counting rate, in the region of this area
of depressed activity, compared to its surroundings? I can see that it
was less, but how much less was it? Moreover, how did the total activity
of the pool compare with a reasonable uptake?" This is a Good Thing.
A decade ago, there arose an idea that a scan of a patient's neck,
24 hours after he had been given some radioactive iodine by mouth, could
be related quantitatively to the "24-hour thyroid uptake". The concept
ran into many objections immediately; it was criticized by many who used
words such as "attenuation", "collimator isoresponse plots", "self-absorption",
etc. Now, in retrospect, these objections don't make a lot of sense, but
at the time they were sufficient to cause many persons to lose interest
:. in the idea that scanning with a moving detector could result in a competent
quantitative measurement. As a result of this, and of the growing desire
for a better "picture", new developments in commercial scanning equipment
.. "Research sponsored by the U. 8. Atomic Energy Commission under contract
LEGAL NOTICE
with Union Carbide Corporation.
This report we prepared a an account of Government sponsored work. Malther the United
state, sor the Commission, nor may pornou acting on behall at the Commission:
A. Yakes may warranty or representation, axproud or implied, will respect to the acou- . i
racy, completescue, or wofalnous of the information contained in this report, or that the une
of any information, appurata, method, or prooos disclowed in the report may not interno
privately owned righto; or
D. Agmumu nay llabuhay with respect to the une of, or for damage roouldnt from the
un da information, appunts, Bethod, or produse diolandia the report.
Ao wood in the abort, "person acting a behall of the Commiaston" Amolados y « į
ployee or contractor of the Commission, or employw at much contractor, to the adent that
toch employs or contractor of the Commission, or employu o much contractor propers,
decontantes, or provides soconto, any taformation parmant to Wo employment or contrast
with the commission, or in employment with much contractor.
ws
DETROA,
U
..
.." 2
7
.11
.
V
AL.
.
.
'-
-
.
.
.
-
.
2
.
-
..
..
:
tended to stress qualitative records, with no more than the minimum necessary
regard for quantitative results. A scanning "systom" no longer included a
scaler that could be used to record total counts, though if it were desired
one could count the "dots on the record".
Fortunately, the idea did not die out entirely; many persons kept the
quantitative aspect alive in their development work in scanning, though ad-
mittedly many worried about the competency of the measurement in an absolute
sense and tended to regard their data as quantitative only in a relative
sense. There is much more to quantitative scanning than use of counting-rate
information; the use of total counts and spatial integration shows up remark-
ably often in the development work reported at the IAEA's Athens scanning
symposium in 1964 (1). The result of this obstinacy on the part of the
"scaming fraternity" 18 that now, in 1965, the clinical usefulness of
scanning is being greatly enhanced by the return to the concept of scanning
as a quantitative measurement.
I would like to discuss the relationship between scanning and a static
measurement, e.g. a "thyroid uptake", to show that they are equally quanti-
tative. Then, it would be useful to consider some of the things, both
simple and complicated, that can be done in storage and manipulation of
scan data.
"Static" Counting of Radioactivity in vivo
When we count, with a stationary detector, a pool of radioactivity
in vivo, and consider the result an accurate measure of the radioactivity
in the pool, we assume that the measurement 18 a proper summation of the
couts from each olomeat of volume in the detector's field of view. I
the detector 18 a right circular cylinder, the counting rate for each 12-
finitesimal volume 18 given with little error, by: :">
11
.
.
.
.
.
When
..
..
.
---• •
.
..
-
-
- 3
-
*+ ()(av)(€)()
* - (pave) 4(COSO)
Eq. 1
16
In this expression, n is the fractional solid angle subtended by the detector
as seen from the source; p 18 the emitted photon density in photons per second »
per/ unit volume (unit volume chosen small enough that self-absorption 18 not
significant); av 18 the elemental volume 3 :
18 the peak
efficiency of the detector at the energy of the given photons (we assume
here that no scattered photons are recorded); D. 18 the diameter of the crys-
tal, x 18 the distance from source to detector; u 18 the linear attoniation
coefficient of the source medium (assumed to be essentially homogenous); z
18 the path length, in the attenuating medium, of the photons in escaping
outward to the detector; 8 18 the angle between the detector axis and a
line from the center of the face of the crystal to the source.
To make an accurate measurement of the total activity in a body pool,
we must contrive to make the detector sum up the contribution from each
elemental source, each having its own X, its own 2, and its ow cos ,
thus :
Total coints = sum of all rates x time - Ert.
Die2i cos
Eq. 2

I av
16
where n is the number of the elemental sources. In the situation of the
usual thyroid uptake measurement, conditions are such that a reasonably
proper summation 18 made. Of course, since e le varies considerably, and"...
since we really cannot avoid recording some scattered radiation, we make

1
.
.
.
.
.
.
.
E
.
.
4
.
this a comparative measurement, and use additional stops to determine the
"uptake". (Where the pool 18 not much larger, if any, than the detestor
area, and the detector 18 reasonably distant, the quantity cos e can be con-
sidered as 1.0. The "B flltor method" compensates for cos 6 offects from
extra-thyroidal lodine.) In spite of the impossibility or confirming this
by measuring each X, 2, and cos , 1t 18 easy to believe that a properly
conducted "thyroid uptake measurement" 18 a quite acceptable quantitative
a88essment of radioactive iodine in the thyroid gland.
Now, consider an upcake-type measurement done with a collimated detec-
tor such as the gamma camera of Anger or the Autofluoroscope of Bender, 10-
stead of an "open" crystal with only outside collimation. This again re..
sults in a proper summation of responses to activities within the field
of view of the detector. It 18 true, of course, that a given region of
the detector may be "seeing" only a part of the group of source elements,
but counts from all parts of the detector go into the total. It is just
the same as using a multiplicity of counting systems, recording la parallel.
Thus, the only difference between using conventional thyroid uptake equipe:
ment and using a stationary scanner is that in the latter situation we re-
quire that the detected photons enter through a sieve (and from this we
obtain spatial information).
If, instead of a parallel-channel collimated gamma camera or the Auto-
fluoroscope, a detector with a focused collimator for moving-detector scanning
were to be used, it would also yield a proper summation of responses for the
source elements that it could "see". Of course, while stationary it would
thus normally see only a small part of a thyroid gland and would thus pro..
vide only part of the desired measurement. The point 18 that it, too, con
be considered a multiplicity of detection systems, recording in parallel.
.
.
-.-
ancor
-
- 5
It 16 interesting to compare the expresrions for canting rates from
simple sources for the thron systems we have considered. These are given
for a point source in air.
Open crystal, point source on crystal axis at distance x:
.
:
i
. 3
Gamma camera, paralle).-channel collimator, source at distance F from
crystal:*
p av
161
(T 18 collimator transmission, R 18 resolution radius as defined by
Anger (2).)
Focused, point source on axis at focus, distance † from crystal:
pa
Eq. 5
16
.1
(T 18 collimator transmission, assuming that the response of the de-
tector 18 the same for photons entering through any hole.)
Nomenclature 18 as used in Equations 1 and 2. The transmission" for
. the open crystal 18 considered to be one, or 100%. In Eq. 4, R 18 effectively
the radius of the area of the crystal illuminated by the source. Thus, these
expressions have not only the same form, but the same content as well.
* This is a rearrangement of Anger's expression for sensitivity (2), and
the conversion 18 shown in reference 3. Crystal diameter does not appear
since the derivation required that it be large enough to cover the region seen through
the collimator from a point at distance F.
* Weronica.
din..
..........
.............
..
..
..
..
.
....
..
...........
me moment
-
w
..
...
.
..
.6.
Expressions for counting rates from uniform shoet sources and homogenous
volume sources also bear the same resemblance to each other.
The Quantitative Nature of Movlag-Detector Scan Data
The equivalence of the three systems doscribed, for stationary counting,
should now be apparent. If we can believe that a "thyroid uptake" measure-
ment with an open crystal can be quantitative, we can also believe that the
same measurement with a gamma camera or Autor luoroscope is equally quantita.,.
: tive. The thing that might not be as obvious 18 that the moving-detector
scanner does the same job; it moves over a pool of radioactivity at
.
..
i
c.
...- .....
-
.
.
in
:
N
...
...
constant linear speed, counts what it sees, and moves on to another
region of the pool. .
It is true that distance, angles, and attenuation lengths undergo at
least moderate changes while a given element of a pool source 18 vibible to
the detector, but these can be considered to be changes in detection efficiency
Moreover these changes are well-behaved, in that as a moving detector be-
gins to see a given source element the efficiency increases, passes through
some maximum, then decreases as the source element moves out of the field
of view, doing 80 by retracing in reverse the increase that occurred 28
the source came into view. So, as the detector moves over a collection of
source elements, the total counts obtained are proportional tu the activity
present in exactly the same manner as are the total counts from a stationary
scanner, or from a stationary "thyroid counter". .
The quantitative value of cowts from a scanner, moving or stationary,
18 reinforced by the fact that, as long as the scan completely covers the
target, the total counts recorded are independent of source distance (3).
This fact 18 virtually independent of collinator type, mchanged by such
things as 1soresponse plots and can be confirmed easily by the reader. This
.
.
.
.
.
2
1
.
-
-
-9-
is very important and one of the things it tells us is that, in a thyroid
scan, cach tiny element of the gland will put a certain number of
counts on the record, regardless of its distance from the focal plane of
the collinator! It an element happens to lie in the path of the focal plane,
the counts due to that elemeut will be distributed in an area on the record
the size of the resolution circle at focus. If the focal plane 16 pussed
above or below the same element, the same number of counts will be recorded,
but they will be spread over a larger portion of the record.
It would appear that the idea of a decade ago that the total counts
in a thyroid could be related to the thyroid uptake" real].y had some merit.
If a scanner had some means -- other than dots on the record -- of record-
ing total counts, more use could be made of the inherently quantitative
nature of scan data. Though we chose here to talk about a thyroid scan
(because the thyroid 18 a convenient example of a body pool), the concept
18 extendable to any other type of pool or collection of radioactive sources.
.........
.
....
Storage of Quantitative Scan Data
.
dan
-
So far we have been reassuring ourselves that scan data 1s really
quantitative. Once that is done, the next step in quantification of scan
records 18 to store the data so that the quantitative nature, which necessar-
1ly includes position data, 18 preserved. The Digital Autofluoroscope
comes with this provision built in, and we won't worry about it. At
present the gamma camera uses total counts and 4 picture record, but this
1. provides quantitative data (4); in addition some of the methods, for extract-
ing quantitative data from scan records from moving-detector scanners can
be used with the camera. Therefore the discussion wil1 be confined to
the moving-detector scanner
-
ini t
--
erdiri
..
.
.
..
.
- 8.
Certainly the most alégant (and proper) way to record scan data 16 to
record the counts, as they come in, on a coordinate system without any arti-
ficial constraint (5). Ideally, each count should have, in this coordinate
system, a mique address; furthermore, this address should be identical with
the location of the axis of the collimator at acquisition, for niaime pre-
prejudice of data. (It is true that some false information 1. recorded by :
making the address the same as the location of the collimator axis, but any
other way of recording adds even more false information. In this situation ,
the elegant way 18 also the expensive way, but it will come some day. In
the meantime, much clinical scanning can be made more useful with simpler
forms of quantitative data storage.
Some of the recording or storage means that might be used are:
1. One or more scalers for recording counts from an entire pool...
. 2. Recording of discrete counts on the scan "picture".
3. Storage of the number of counts in picture elements, one at.a
time.
4. Recording the address of the detector at count acquisition. (The
existence of a stored address indicates the presence of a count.)
The first of these is easy to do, if there 18 a scaler in the scanning
system, and can be clinically very useful. For example, it has been shown
that macro-aggregated albumin I-131 scan data can be analyzed to provide
a reliable alternative to the more difficult differential oxygen uptake
method for estimation of relative pulmonary artery blood flow to each limg
...: (6.7). Recording of such data with scaling equipment can be as simple
or as complicated as one chooses. Some persons scan one side of the chest
and record total counts, and then do the same for the other side. The
scalers, one storing counts from each side of the chest can be programmed
..
::..........
1
.
..
1...
'
-
21
:
ir
..
.
.
1
.
- 9 -
by the scan motion itself (6). A scaler and line printor can be used
to record counts on each scan sweep, from each lung (8). Actually, the
recording method in this clinical situation could be any of the four methods
listed above, depending on total laformation desired and deponding on
equipment availability.
The important point is that here to a situation (and there seem to
be many others) where analysis of quantitative scan data, in addition to
the usual qualitative picture, has been clinically useful. Another impor-
tant point 18 that data should be stored in such a manner that the retrieved
Information retains enough integrity for the clinical task at hand. In .
view of the increasing usefulness of selective integration of regional data
(7), it would appear that storage and retrieval methods that retain the
maximm information possible will be the methods of choice, in spite of
their complexity and expense.
The storage of counts in a picture element by punching paper tape (7)
or by numerical means on an x-y coordinate system (9) 18 coming in for a
certain amount of attention. This method has the disadvantage that it
necessarily prejudices the stored data, but if the picture element 18
sufficiently small it is a reasonable compromise between cost and useful
retrieval of clinically meaningful data. Observed detail cannot be finer
than a picture element, so there is danger in making the element too large.
On the other hand, if the element is very small, a larger, composite ele-
ment can be constructed by integration as needed. Computers are very
adept at this job. Bopefully, as the use of computers increases, even
the small constraint of pre-choosing a picture element can be eliminated.
It will probably turn out that quite a bit can be done with scan data,
even when a scan picture 18 the only record, provided that the recording
i
- 10 -
Itselt 18 reasonably quantitative. The old dot pattern was not the castest
record in the world to interpret by eyeball, but it was reasonably quanti-
tative. One could always count the dots (spatial integration) wless' necessary
Information was missing because of Mackground erase". Of course, the more
discrete the appearance of a count, the less Information will be lost in
retrieval. To us, a discrete-dot can record (3), 16 interesting because
1t 18 inexpensive, can be made to look as if it were printed with a "gaussian":
spot" by viewing it through a $1.00 piece of non-glare glass (10), and it .
retains reasonably good quantitative integrity. In addition, from it can
be produced quantitative 18ocount contour maps by rescanning (11). Bender
and Blau (12) used a discrete-dot record to quantify the dynamics of the
passage of Hippuran 1-131 through renal cortex, medulla, and pelvis.
Even analog photorecording retains some quantitative properties, when
the trouble has been taken to relate density to counting rate, and can be
analyzed by densitometer or rescanner. :
A survey of the remarkable things that can be accomplished by opera- :
tions on scan data that has been retrieved from storage in good order iş
beyond the scope of this paper. Kuhl (7) has discussed some of these oper-
ations, including smoothing random fluctuations, data bounding, spatial
averaging, isocount line generation and differentiation processes :: A
subtraction process by Schepers and Winkler (Ref. 1, PP 321-329) 18 very
interesting.
Summary
The information obtained by use of moving-detector and stationary
scanners is considerably more quantitatiye than popularly believed over
the last decade. The clinical usefulness of scan information can be greatly
enhanced by proper storage of these data for subsequent retrieval and
Whendevido
a
.
-
o
...
- 11 -
manipulation. The use of computers provides a means of storage, retrieval,
and manipulation. Simpler methods are also possible and very useful. The
intelligent use of scan information requires the maximum possible use of
the quantitative aspect and no means, however simple, should be overlooked.
-
1.
4
.
Win
NT
• 12 -
.
.
$
,
.
.
.
.
..
::.
per
References
1. Medical Radioisotope Scanning, Proceedings of a Symposium held by the
International Atomic Energy Agency, Athens, April 1964, Vol. I, IAEA,
Vienna, 1964.
2. Anger, 1. O., Scintillation camera with multichannel collimators, J. Nuclear
Med. 5:515-531, July 1964.
3. Barris, C. C., Satterfield, M. M., Rose, D. A. and Bell, P. R., Recti.
linear vs. stationary scanners, (10) A Symposium on Frumdamental Prob-
lems in Scanning, Chicago, May 8-9, 1965, Charles C: Thomas, Springfield,
Ill., (in press).
4. Myers, W. G., Dynamic studies with a gamma-ray scintillation camera (10)
Reference 1, pp 377-390.
5. Bell, P. R., personal communication Dec. 1964.
Lopez-Majono, V., Chernick, V., Wagner, H.N. Jr., and Dutton, R. E.,
Comparison of radioisotope scanning and differential oxygen uptake of
the lungs, Radiology, 83:697-698, October 1964.
7. Kuhi, D. E., and Edwards, R. Q., Digital techniques for on-site scan
data processing, (in) A Symposium on Fundamental Problems in Scanning,
Chicago, May 8-9, 1965, Charles C. Thomas, Springfield, n., (in press).
8. Friedman, W., National Institutes of Health, pers. comm. Nov. 1965.
9. . Laughlin, J. 8., Kenney, P. J., Corey, K. R., Greenberg, E., and Weber,
D. A., Localization and total body high-energy gamma-ray scanning studies
in cancer patients, (10) Reference 1, pp 253-272..
20. Anger, 8. O., VanDyke, D. C., Gottschalk, A., Yano, Y., and Schaor, L.R.,
The scintillation camera in diagnosis and research, Nucleonics 23:57-62,***
January 1965.
-
-
--
...........
....................................................................
- 13 -
..: 11. Harris, C. C., Satterfield, M. M., Uchiyama, G., and Kimble, H.E.,
A "rescanner" with photographic color readout, to be published in
J. Nuclear Med..
12. Bender, M. A., and Blau, M., 'The Autofluoroscope, Nucleonics 21:52-56,
October 1963.
..


1
AI
(
L
'
1
1
.
.
AI
1
1
74
1
3
.
.
.
.
.
1.
.
.
43
.
TA
.
'
..
.
1..
.
DATE FILMED
2 / 18 /66