Note: Descriptions are shown in the official language in which they were submitted.
;757
Background of the Invention
This invention relates to devices for obtaining infor-
mation about radiation sources.
Visible light can be both reflected and refracted.
The ordinary camera takes advantage of refraction by using an
optical lens to refract and focus the visible light coming from
an object, in order to produce an image of the object on film.
However, electromagnetic radiation of higher frequency than
vacuum ultraviolet (SUC]I as X rays and gamma rays, together
referred to hereinafter as "gamma radiation") cannot be efficiently
either reflected or refracted. Therefore the formation of an image
of a source of gamma radiation has been achieved with the use of a
collimator, which operates in somewhat similar fashion to the old
pinhole camera. The pinhole camera permits the light rom an
object to pass in a straight line through a pinhole in the camera
box, to produce an inverted image on the film.
Collimators ~ith an array o~ parallel channels, such as
the one depicted in Figure 1, have been used in the imaging of
gamma radiation sources. I~ith the axes of the channels pointed
toward a gamma radiation source, the channels are generally the
same size in both dimensions normal to the axis; usually the chan-
nels are circular~ triangular, or square in cross section. In
Figure 1, with the walls or septa of each channel made of lead
to absorb gamma radiation, and with a radiation detector (not
shown) placed on the side of the collimator opposite the source,
the radiation that can pass from a point source through a partic-
ular channel and reach the detector is defined by the solid angle
(A) subtended by the base of the collionating channel 2. Spatial
resolution of such a collimator is improved by reducing the solid
angle. However, the sensitivity of each channel, which
- 1~67S7
increases as does the amount of radiation passing throuyh the
channel, is improved by increasing the solid angle. Of course
it is desirab~e to improve both spatial resolution and sensitivity ,
the latter particularly so that the necessary time of observation
may be reduced.
Turning to the radiation detector, it is kno~m to use
the photoconductor as the basic element of suc~ a detector.
However, in known arrays of photoconductor detector elements
the photon absorption distance and the interelectrode distance
correspond to the same photoconductor dimension, and thus are
generally the same length. It would be desirable to make the
photon absorption distance large with xespect to the inter-
electrode dista'nce, for the larger the absorption dis~ance, the
better the sensitivity, because a larger fraction of the incident
lS photons are collected, and the smaller the interelectrode distance ,
the greater the efflciency, and also the rapidity, of collection
at the electrodes of electrical signals created in the ohoto-
conductor body by incident photons. .
,~
Summary of-the Invention
The invention provides a sensitive, fast, hig:~
. resolution, and convenient-to-use device for obtaining informa-
tion about the distribution of a gamma radiation source, and
further provides a radiation detector useful in such a device.
The information-obtaining device of the invention has
' high sensitivity without sacrifice of resolution. The device
includes a collimator that is easy to construct, ~ithout the
necessity for an intrica-te honeycomb of separate channels, and
the septa of the collimator can be conveniently made of tungsten
foill tungsten being a bett`er radiation-absorbing material than
' . _ '
~L~16757
lead, and t~us can ~e th~nner than lead septa, thereby improving the effective
collimator transparenc~. The output of the device can be readily and rapidly
transformed ~ convent~onal computer techniques to provide highly resolved
images of gamma radiation sources. The device has su~stantial present appli-
cation in the field of nuclear medicine, and also has industrial application.
T~e detector of the in~ention is easy to construct, and has an im-
proved resolution and signal-to-noise ratio. It has a high uniformity of
response to a given incident photon energy so that noise and spurious signals
caused by Compton effect scattered, lower energy photons originating at sites
remote from the primar~ radiation sources can be rejected. It also has a high
photon collection ef:Eiciency, for improved sensitivity, and a brief output
pulse, for improved temporal resolution.
The invention features in one aspect obtaining information about po-
sitional source of, for example, a gamma ray source, by slit collimating and
detecting aeam components from the source in a multiplicity of varying slit
locations, and using the resulting data to plot the source position.
Thus, in accordance wlth a broad aspect o the invention, there is
provided an instrument for obtaining positlonal source information which com-
prises: a slit collimator containing a multiplicity of slits for reception
therein of ~eam components moving in a straight line from a source, each slit
of said slits including an open end for orientation toward said source and
slit-defining walls extending inwardly from said open end, said walls extend-
ing in the same longitudinal direction, said walls including material of
character and thickness to absorb those of said beam components impinging
thereon, and said slits extending further in one transverse direction than in
the other transverse direction; a detector for separately detecting beam com-
ponents passing through said slits and for providing detection data output
signals, said detector being fixedly mounted relative to said collimator,
said detector comprising a plurality of detector elements equal in number to
3a sa~ slits, each of said detector elements being positioned to close the end
- 4 -
L6~157
of a respecti~e slit opposite each said open end; and a positioner,
said positioner heing interconnected with one of said collimator
and said source so as to simultaneously change the transverse posi- -
tion of said slits and detector elements relative to said source;
characterized in that each said detector element is positioned
within its respective slit between said slit-defining walls.
Preferred embodiments of the invention feature in this
aspect a collimator including a frame having an axis of rotation
and a plurality of flat sheets of gamma-radiation-absorbing
material maintained b~ the frame in parallel, spaced-apart re-
lation with respect to each other and parallel with the axis
of rotation, adjacent pairs of the sheets defining slits there-
between, each of the slits having an opening at one end thereof
and a base at the opposite end thereof and being unimpeded, for
permitting passage of gamma radiation therethrough, in a first
direction parallel to the axis of rotation, and being unimpeded,
~ithin the frame, in a second direction perpendicular to the axis
- 4a -
. ,.
1~i7~7
of rotation and parallel with the sheets, means for positioning
the collimator to maintain the axis of rotation pointed at a
gamma radiation source so that the slits are disposed to receive
gamma radiation therefrom while the collimator is rotated about
the axis, each of the slits subtending, in the plane defined ~y
the first and second directions, a much larger angle for receiving
radiation passing therethrough to the base thereof from the source
than it subtends in a second plane perpendicular to the second
direction, a detector effectively connected to the frame for
common rotation-with the collimator and positioned adjacent the
bases of the slits for detecting radiation passing through each
slit to the base thereof and providing an output representative
of intensity of detected radiation over the whole basè of each
slit as a function of the angle oi. rotation of the collimator,
and means for accumulating a matrix of such outputs, the matrix
being ordered according to the particular slit in which the
radiation causing the ou-tput was detected and according to the
particular angle of rotation of said collimator at the time the
radiation was detected, the matrix being suitable for transfor-
mation to a matrix corresponding to the image of the source.
` The invention features in another aspect a detectorfor detecting gamma radiation and for providing an output in
response to the radiation, the detector comprising a plurality
of detector elements of photoconductor material sensitive to
gamma radiation, the elements being strips arranged in parallel,
spaced-apart relation, to expose to incident gamma radiation
from a source of the same a generally planar surface made up of
one face .rom each of the elements, each of the elements having
a pair of electrodes affi~ed thereto, each one of the pair of
electrodes being positioned between adjacent elements r the planes
. _
. ~
1~.i757
of electrodes being perpendicular to the planar surface exposed
to incident radiation, and the thic~ness of each detector element
measured from the surface exposed to incident radiation along a
perpendicular line therefrom is large with respect to the dis-
. 5 tance between each pair of electrodes.
Certain preferred embodiments feature a detector com-
prising a plurality of detector elements equal in number to the
slits, the elements being effectively connected to the frame so
. that they remain positioned adjacent their respective bases while
the collimator is rotated about-the axis; processing circuitry
including an amplifier for amplifying detector outputs and an
amplitude discriminator for discarding from the outputs any com-
ponents having below a.minimum amplitude; a housing as the
ma;ntaining means, the housing illcluding a tube, a piate rotatabl~
mounted in the tube, and means for rotating the plate, the plate
. containing an aperture for recei~ving the frame so that when
the plate is rotated, the frame is rotated together therewith;
: an indexed motor adapted to rotate the plate in discrete angular
steps as the rotating means; collimator sheets made of tungsten
foil; a detector element including a scintillating sheet and a
photomultiplier connected to the sheet for providing an electrical
output; optical fibers interconnecting the sci~tillating sheet
with its photomultiplier; and detector elements positioned be-
tween the collimator she.ets at the base of the slits and main-
tained there by the rame. Other preferred embodiments featuredetector elements of photoconductor material; detector elements
composed of cadmium telluride; photoconductor detector elements
the planes of whose electrodes are parallel to the collimator
sheets and the distance through the photoconductor element in
the first direction is large wi~h respect to the distance between
. _ .
~ 7~7
each pair of electrodes in a third direction perpendicular to
- the first and second directions; a distance through ~he photo-
conductor element in the first direction no less than 5 mm and
the distance through the photoconductor element in the third
direction no greater than 0.75 mm; and photoconductor elements
in the form of s-trips.
Other advantages and features of the invention will
be apparent from the description and drawings herein of preferred
embodiments thereof.
~'16757
Brief Description ol the Drawings
Fig. 1 is a sectional view, along a vertical plane,
of a typical known channel collimator;
- Fig. 2 is a view in perspective of one embodiment
of the invention;
Fig. 3 is an enlarged plan view of a portion of the
embodiment of Fig. 2;
Fig. 4 is a view through 4-4 of Fig. 3, with portions
broken away;
Fig. 5 is a view through 5-~ of Fig. 3, with portions
broken away;
\ Fig. 6 is an exploded isometric view of a portion
of the embodiment of Fig. 3,
Fig. 7 is a diagrammatic view of the embodiment of
Figs. 2 through 6, with associated circuitry;
Fig. 8 is a greatly enlarged sectional view, along
a vertical plane, of a portion of a second embodiment of the
invention;
Fig. 9 is a sectional view, along a vertical plane
perpendicular to the plane of Fig. 8, of said second embodiment,
with portions broken away and with associated circuitry sho~n
diagrammatically; and
Fig. 10 is an exploded isometric view of a portion
of said second embodiment.
Description of the Preferred Embodiments
_ _
Camera 10, which embodies the invention, is shown
in Fig. 2, and comprises an integrated collimator 12 and de-
tector 14 mounted for rotation together in housing 16. Colli-
~167S7
mator 12, as better shown in Figs. 3 through 6, includes frame
18 made of steel and a series of fifty-one parallel sheets 20
of tungsten foil held in tension in frame 18. Frame 18 has two
opposite sides 22, which are 80 mm by S5 mm by 15 mm. A set
of three steel 5 mm diameter rods 24 connects sides 22 at the
ends thereof through holes 26 drilled through sides 22. Screws
28 when tightened in holes 20, which intersect holes 26 trans-
versely, hold rods 24 in place in sides 22, and form a s~uare
having an outer dimension of about 80 mm by 80 mm. Sheets 20
also have a set of three holes 32 (Fig. 6) at their ends, to re-
ceive rods 24, which thereby maintain sheets 20 in tension with
p a ~c~ lC77~ the assistance of lead spacers 34, which ~pearat~ adjacent sheets
20 at the sheet ends, and are held in position by rods 24 passing
throuyh holes in the spacers. Each sheet 20 is 0.15 mm thick,
30 mm wide, and 80 mm long. Lead spacers 34 are 0.85 mm by 15 mm
b~ 30 mm plates. Spacers 34 and rods 24 together form the two
other sides of frame 18 in addition to sides 22.
Sheets 20 are equidistantly spaced 0.85 mm from each
other to define fifty slits 36,-which are 50 mm by 30 mm by 0.85
mm. The previously described tensioning of sheets 22 maintains
these slit dimensions.
Detector 14 comprises fifty detector elements 38, which
are scintillating sheets made o commercially available scintil-
lating plastic composed mainly of polyvinyl toluene and manufac-
tured and sold by Nuclear Enterprises, San Carlos, CaliforDia.Each sheet 38 is 50 mm by 10 mm by 0.85 mm, and is fitted be-
tween each pair of adjacent tungsten sheets 20. With fram~ 18
directed toward a radiation source so that slits 36 are most
favorably positioned to receive radia~ion from the source, each
..''. ` .
.
:
~675~?'
slit has an opening closest to the source to receive radiation
and a base at the opposite end of the slit and the edges of
sheets 38 that are farthest from the radiation source are posi-
tioned flush with the edges of the tungsten sheets 20 that like-
wise are farthest from the radiation source (Figs. 4 and 5).
Attached by transparent epoxy to the rear face of eachscintillating sheet 38 is a ribbon 50 of optical fibers (diagram- .
matically shown in Figs. 4 and 5) havihg the same dimensions in
cross-sectional area as does the rear face of sheet 38 (50 mm by
0.85 mm). Each ribbon 50 is composed of approximately 8 layers
~_ ~rc~ of 0.1 mm diameter fibers, about 4~ fibers per layer, all pro-
duced by conventional fiber OptIC techniques. The fibers extend
perpendicularly away from the rear face of sheet 38. All ifty .
ribbons.50, one for each sheet 38, are potted together in trans-
parent epoxy, forming block 42, which e~tends 25 mm outwardly
. from the rear ~aces of sheets 38. Frame sides 22 likewise ex-
tend 25 mm below the rear faces of sheet 38, to provide a frame
. .~or the block of potted fibers. Conventional clamps (not shown) .
can be used to assist frame sides 22 in gripping block 42, which
~0 extends the length of sides 22. As they extend out- OL b10C1,; 42,
the fibers making up each ribbon 50 are independently flexible,
and are gradually drawn together into a circular bundle, main-
tained in that position by ferrule 52. The.ends of each circular
fiber bundle are bonded to the sensor face of photomultiplier
54, which converts light signals to electrical ones for trans-
mission to pream~lifier 56 (Fig. 7). The optical fiber connectio
between detector array 14 and each photomultiplier 54 is flexible
enough to accept a 180 rotation of plate 60.
' ' -10- ' I
~6757
Collimator 12, detector 14, fiber ribbons 50, photo-
multipliers 54, and preamplifiers 56 are all mounted in housing
1~. Housing 16 includes circular steel mounting plate 60, steel
tube 62, plate drive 64, and support arms 66. Frame 18 fits
within a square hole centrally placed in mountin~ plate 60,
elamps 68 holding frame 18 in place in plate 60. Plate 60 is
fitted within the fo~ard opening of tube 62, and is rotatable
with respect to tube 62, a conventionai bearing arrangement (not
shown) permitting the rotation. Plate drive 64, including a
reversible, indexed electric motor and timer, rotates plate 60,
whose outer rim is too-thed to provide a gear linkage (not shown)
with drive 64. Frame 18, collimator 12, and deteetor array 14
all rotate with plate 60, which is driven discontinuously by
drive 64.
Tube 62 is itself pivotally mounted on support arms
66 so that the tube ean be tilt~!d toward a particular radioactive
source. Locking knob 44 is adjusted to hold tube 62 in the po-
sition chosen. Support arms 66 are mounted on a base (not shown),
which conveniently includes casters, so that ca~era 10 as a whole
ean be wheeled to different positions.
Fig. 7 shows in bloek diagram eonventional eircuitry
for processing electrical signals from photomultipliers 54.
Signals from photomultipliers 54 are in the form o electrical
pulses, each pulse corresponding to the absorption of a gamma
`~ 25 ray by the eorresponding scintillating sheet 38. These electrical
pulses are transmitted to preamplifiers 56 through ifty leads
70 rom the fifty photomultipliers 54. Preamplifier556 Ci~ lo-
cated in the rear portion of tube 62, just forward of a steel
eireular backplate (not shown) covering the rear opening of tube
62. A hole throug}l the center of the tube backplate permits
fifty leads 72 ehanneled through fle~ible conduit 74 to pass
~ 1116757
from preamplifiers 56 out of tube ~2. Preamplifiers 56 amplify
all pulses coming from photomultipliers 54, and are placed in
tube ~2 to shield them from noise. The rest of the circuitry
of Fig. 7 is conveniently housed in the mobile base. The am-
plified pulses travel through leads 72 to fifty pulse amplifiers76, one for each lead, where the pulses are further amplified.
The pulses are then transmitted from the fifty pulse amplifiers
76 to fifty pulse-height discriminators 78, which reject pulses
having less than a predetermined amplitude ~such as pulses caused
by Compton effect photons), and permit the rest of the pulses to
pass. Finally, the pulses are counted, and their number entered,
in register 80. Register 80 is a S0 x 50 register, and is syn-
chronously connbcted to plate drive 64, so as to count the pulses
coming from the fi~ty pulse-heic~ht discriminators for each dis-
crete angular position of collirnator 12 and detector 1~. Thisdata made up of pulse counts is stored in register 80, and, ~hen
all counts have been made, is subsequently reduced by known com-
-puter techniques, to be explained in more detail below, to a form
which iden~ifies the two-dimensional location of the radiation
sources impinging on collimator 1~ and detector 14.
In opera~ion, camera 10 is positioned so that the front
of collimator 12 is as close to the radiation source as possible.
The source itsel~ is composed of Technetium 99, a radioisotope
that emits gamma radiation with a characteristic energy of 1~0
25 ~ Kev, or of some other radioisotope suited to clinical or other
useful service. If the source is within a patient, collimator
12 is preferab~y brought into ~ontact with the patient in the
. ,'
_ '
1~6~S7
vicinity of the source. The source itself, for purposes of data
analysis, may be regarded as a three-dimensional array of point
sources randomly emitting gamma photons. Camera 10 in effect
takes a picture of the two-dimensional array resulting from the
orthogonal projection of this three-dimensional source array
upon detector 14.
Collimator sheets 20 and collimator slits 36 are initi-
ally vertically aligned, and the axis of tube 62 is then aimed
directly at the source volume ~usually at the part of the patient's
body to which the isotope has traveled). Locking knob 44 main-
tains the tube orientation. Photomultipliers 54 are activated,
ready to respond to light signals from detector 14, and the signal
processing equipment of Figure 7 is energized. Collimator 12 remains
in the vertical position for of the order of 10 seconds, during
which time gamma photons travel from the source to collimator 12,
and enter the most Eavorably positioned slits 36. Plate drive 64
then rotates collimator 12 through 3.6, followed by a ten-second
pause, followed by another 3.6 step, and so on ulltil fifty
such steps of 3.6, amounting to 180, have been taken. The num-
ber of rotational steps for each 180 turn ~here, fifty) is chosen
to equal the number of slits 36 in collimator 12. During each
~ ten-second interval, tungsten sheets 20 absorb any photons hitting
; the sheets. Photons passing from a particular point source
through a particular slit solid angle subtended by the area
of the incident face of a scintillating sheet 38 will pass into
that scintillating sheet, and most will be absorbed therein, ` .
thereby exciting visible photons in the sheet. These visible
1~L16757 r
photons are transmitted through sheet 38, and illuminate optical
fibers in the ribbon 50 attached to the particular scintillating
sheet. All faces of scintillating sheet 38 except the face bonded
to ribbon 50 are coated with a substance reflective of visible
light so that all visible light pulses excited within sheet 38
are eventually transmitted toward the rear face of sheet 38 and
ribbon 50, though with higher attenuation of pulses undergoing
one or more reflections before reaching ribbon 50. Attenuation
of the visible light signals also occurs within ribbon 50, and
about 1~ of the visible light photons generated in sheets 38
reaches photomultipliers 5~. However, the signals reaching ths
photomultipliers are sufficient to provide a basis for accurately
determining the location of th'e soùrces o~ radiation. '
Scintillating sheet 38 gives an output indicative of
the energy of the incident gamm~l photon, thus photons of e,nergy
lower than the primary gamma photons arising from Compton
scattering and originating at sites remote from the primary
radiation sources can be rejected in pulse-height discrimina~or
78. During the ten-second interval before collimator 12 takes
its first 3.6 step, and during each of the forty-nine subsequent
- tèn-second intervals between subsequent 3.6 steps, each scintil-
lating sheet 38 is collecting gamma photons from all radioactive
point sources lying in the lamina of space created by projecting
the pair of tungsten sheets 20 bounding sheet 38 toward the
~5 sources. Likewise during that interval each sheet 38 is collect-
ing all the radiation emitted by each point source within the
solid angle subtended by the area of the frontal face of sheet
38 with respect to that point source. The output of each
scintillating sheet 38 at each rotational position is thus a
series of visible light pulses corresponding to a series of gamma
photon absorptions within the sheet, the photons coming from the
various point sources.
1116757
The procedure for processing the data stored in register
80 to construct an image of the radioisotope source distribution
is grounded preliminarily on the assumption that the distribution
to be imaged will appear in the plane perpendicutar to the first
direction as a two-dimensional array of 50 x 50 source elements.
The problem then becomes that of constructing an image with 50 x
50 resolution elements, i.e., solving for 50 x 50 = 2500 unknowns.
The isotope will be located wherever the value of the source
element is nonzero, and will be absent where the value of the
source element is zero. In principle, when there are 50 x 50
unknowns, it is always possible to solve for those unknowns with
a system of 50 x 50 simultaneous linear equations. The steps of
setting up the equations in matrix form and solving by the pro-
cedures of matrix algebra constitute a well-known direct approach
to finding the ~mknowns. Register 80 p:rovides the data from
which 50 ~slits) x 50 tangles) = 2500 equations can be constructed.
Computer-performed mathematical techniques are presently at a
point where expedited solution of this number of simultaneous
equations can be accomplished. The following references
set forth methods for such expedited solutions: Ramachandran -
and Lakshminarayanan, "Three Dimensional Reconstruction From
Radiographs and Electron Micrographs: Applications of Convolu-
~ tions Instead of Fourier Transforms," Proceedings of the National
; Academy of Sciences of the United States, 1968, pp. 2236-2240;
Gordon and Herman, "Three Dimensional Recons~truction from Projec-
tions: A Review of Algorithms," International Review of Cytology,
Vol. 38 ~1974); DeRosier and Klug, "Reconstruction of Three
Dimensional Structures from Electron Micrographs," 217 Nature
130 (1968); and M. M. Woolfson, An Introduction to X ray Crystal-
lography, Chapter 4, "Fourier Transforms" (Cambridge Univ.
Press 1970).
X ,~
~ 16757 -
It is clear that radiation emitted from the peripheral
regions of the source area will enter the camera with a somewhat
lower probability of being accepted by the detector array than
will radiation emitted from regions clos~ to the axis of rotation
of the collimator. Therefore it will be necessary to "weight"
the data processing procedure so as to compensate for this bias.
In the quoted references it is shown how this compensation may
be achieved.
Because the resultant image created by these computer
techniques is a two-dimensional one, though the radioisotope it-
self actually occupies three-dimensional space, the further tech-
nique, well-known in the field of radiation therapy, of taking
additional pictures from different spatial positions of the col-
limator.and combining them to produce a three-dimensional image
is used. In the case in which the radioisotope is in the brain,
the simplest method of carr~in~ out this technique would be
to take one picture with the axis of the collimator directed be-
tween the patient's eyes and to take a second picture with the
collimator axis.shifted 90 so as to be directed through the
patient's ears.
. ~ comparison o* camera 10 with a conventional static
channel collimator-detector system shows that sensitivity is im-
proved with camera 10 and thus exposure time reduced without
. any impairment of resolution.
Furthermore, integration of detector 1~ with the colli-
mator 12 improves the overall spatial resolution of the image
in two respects. First, if the detector 14 and collimator 12 be
considered as separately construc~ed devices ~ithout attempt at
. registration between individual slits 36 and detector element 38,
as is normally the case with existin~ channel collimatin~ struc-
~ 67S7 `
tures, then the effective resolution of the complete system is
, generally considered to be the square root of -the sum of the
squares of the minimum resolution distances of the two devices
(the collimator and the detector) taken separately. Therefore
if the spatial resolutions of the two devices are similar, in-
tegration of the detector with the collimator will yield an over-
all resolution improvement by a factor of the square root of two.
Second, and more important, integration of the detector 1~ with
collimator 12 will essentially elim,inate the loss of resolution
incurred through scattering of radiation either within the colli-
mator or within the detector array itself.
~ igs. 8 throug~ 10 show a second em~odiment of the
invention using the same collimator 12 but a different detector
46. Detector 46 is,co~posed of fifty detector elements 88, which
are strips of photoconductors 90, five of which are spaced by
polyethylene terephthalate insulation spacers 91 to form each
strip. Each cadmium telluride photoconductor 90`is a rectangular
crystal wafer measuring lO mm by 5 mm by 0.75 mm, and each strip
88 of wafers is thus approximately 50 mm long. Deposited on the
two largest faces of each photoconductor 90 is a pair of elec-
trodes 92. Each electrode 92 itself comprises a thin layer of
platinum deposited directly on the cadmium telluride body, thereby
forming an electrode-cadmium telluride-electrode sandwich. The
deposited layers 92 are each on the order of a micron in thickness.
25- Resting flush against one side of strip 88 and extending parallel
to the strip is ground connector 98. Ground connector 98 is a
strip of polyethylene terephthalate manufactured and sold by du
Pont under the trademark Mylar on which a thin aluminum coating
has been deposited. The thickness of this Mylar-aluminum strip
is approximately fifty microns, ~ith the Mylar accounting for most
of the thickness. An aluminum-Mylar ground finger lO0 e~tends
.
. . .~
1~16757
downward from strip 98 toward one end thereof. The alumin~m
face of ground connector 98 is positioned on one side of strip
88. The Mylar gives strength to the aluminum connector while
. serving as an insulator between an adjacent tungsten sheet 20
and the aluminum connector. A similar connector 102 is positioned
. on the other. side of strip 88, but the aluminum coating has been
divided, by prior removal of narrow vertical bands of aluminumr
into five electrically isolated regions 104 corresponding to .
. the five photoconductors 90 making.up strip 88. ~ach of the five
aluminum regions 104 has an aluminum-Mylar finger 106 extending
downward from the region. The w:~ole connector-electrode-photocon-
ductor-electrode-connector sandwich that results when all these
elements are brought together is positioned between èach pair of .
adjacent tungsten sheets 20 of collimator 12 as well as between .
one frame side 22 and an adjacent tungsten sheet 20. This.sand-
wich structure measures 50 mm by 5 mm by 0.85 mm, and takes the
place of scintillating sheet 38 in slit 36. A strip 88 of photo-
.. conductors 90 accounts for most of the thickness of the connector-
electrode-photoconductor sandwich in each slit 36 so that only
a small fraction of the incident gamma photons in slit 36 will
. be lost by entering the connectors 98, 102 or the electrodes 92.
Photoconductor strips 88 and conductors 98 and 102 are supported
by printed circuit board 108. Board 108 is slotted at appropriate
locations to receive ground fingers 100 and fingers 106. The fin-
gers pass to the underside of board 108, where a printed ground
lead (not shown) connects the ground fingers 100 for all the
ground connec~ors of the detector array and where separate printed
leads (not shown) are individually connected to each of fingers
106. Photoconductor strips 38 rest along their lower faces
on board 108. Tungsten sheets 20 are also gro~nded by connection
to the ground lead of board 108 (the con~ections are not shown).
~675~7
Strips 88, by virtue of Mylar spacers inserted at the strip ends,
. and the aluminum portions of connectors 98 and 102, by removal of
bands of aluminum at the ends, are also insulated from lead
spacers 34, which as before are fitted between sheets 20, and form
two borders ,or tne detector array. ~oard 108 is itself secured
to frame sides 22 by brackets (not shown).
Voltage source 110 (50 ~olts) (Fig. 9) is connected
through board 108 and connectors 98 and 102 so as to apply its
voltage through each pair of electrodes 92 across each photocon-
ductor 90 (there are 5 x 50 = 250 photoconductors 90). The out-
put signal from each photoconductor 90 is carried out through
connectors 98 and 102 to printed leads in board 108 and from
. there through flexible leads 112 to a preamplifier 56~ There axe .
. 250 preamplifiers in all, one for each photoconductor 90. The
I . 15 output signals fro~ five preamplifiers 56a corresponding respect-
ively to the five photoconductors 90 in a strip 88 are combined.
and fed into a pulse amplifier 76. As with the embodiment using .
detector array 14 and photomultipliers 54, there are fifty pulse .
amplifiers 76, fifty pulse-height discriminators 78, and one re~-
ister 80. Each preamplifier 56a is an operational amplifier with
. a field effect transistor input, a open-loop gain of 105, and an
input current sensitivity of 10-11 ampere.
An incident gamma photon producing a charge of 10 14
coulombs in the photoconductor crystal, and with the entire
charge collected at the pair o-f electrodes 92 in a micro-second,
the current output is 1O -6 10-8 ampere, which can be recognized
and amplified in preamplifier 5~. Because the Mylar insulation
on connectors 98 and 102 is kept thin so that the area oE the
radiation-exposed surface o.f photoconductor 90 can be ma~imized,
a large capacitance (in relation to the capacitance across the
photoconductor 90) is developed across the M~lar between each
tungsten sheet 20 and the aluminum coating of connectors 98 and
102. This capacitance might ordinarily reduce the voltage ex-
cursion of any signal coming from electrodes 92 below limits
detectable by preamplifier 56, but the segmenting of detector
strip 88 into five separate photoconductors 90 results in an
overall capacitance for each photoconductor 90 one-fif-th as large
as the total capacitance of strip 88, an acceptable value as far
as retrieval of signals from photoconductor 90 is concerned. In
regard to sensitivity, with a preamplifier 56a capable of detectin
incoming si.gnals down to 10-11 ampere, accurate measurement of`the
magnitude o~ expected signals on the order of 10-8 ampere is pos-
sible. Thus it is possible to screen out the lower energy Compton
photons downstream in the pulse-height discriminators 78. The
overall bacXground noise is approximatel~v equivalent to a 10 Kev
signal so that discrimination bel:ween genuine source photons and
noise is possible, using radiation sources of more than about 20
X~:v. ' .
The thin strip design of photoconductors -90, with elec-
trodes 92 positioned parallel to the collimator sheets 20, though
bringing with it a capacitance problem as just described, offers
simultaneously the advantages o~ a short interelectrode distance
and a long photon absorption distance. - By having a relatively
short distance be-tween opposing electrodes 92 ~about 0.75 mm),
the current carrier (electron or hole) collection efficiency of
the electrodes is improved, with resultant improvement in obtain-
ing uniformity of response for photon e~citations, and the carrier
collection time is reduced, with resultant improvement in temporal
resolution. B~ having a relatively deep photoconductor crystal
(5 mm~, most incoming photons from a 1~0 ~ev or less isotope
source will be absorbed b~ the cr~s a', providlng be~ter camera
~L16757
sensitivity. In general, ~ith use of photoconductors signal losses are far
lower than with scintillating sheets and optical fibers, and energy resolu-
tion is much improved. With more of the signal actually reaching preampli-
fier 56a, more accurate imaging is the result.
Methods for processing the data received in register 80 are as
previously described, and the improvement in sensitivity over the channel
collimator is also as previously described.
Regarding modifications in procedure and structure, collimator 12
can be continuously rotated rather than discontinuously rotated, or even
operated moving its axis along another curved or other configuration or
without rotation. Symmetry about the axis is preferred but not essential.
Appropriate modifications in the data reduction procedures will then be re-
quired, of course. If harder radiation sources than Technetium 99 are used,
scintillating sheets 38 and photoconductors 90 will need to be deeper from
top to bottom ~i.e., have a longer photon absorption distance). If, for
example, the source is on the order of Mev's, photoconductor 90 may need
to be 40 or 50 mm deep rather than only 5 mm. If greater resolwtion is
desired, the number of slits in collimator ]2 can be increased accordingly,
to at least a total of 250 slits; assembly of such a collimator would under-
standably be somewhat more involved than in the case of the flfty-slit col-
limator. Tantalum can be used in place of tungsten in sheets 20. Poly-
styrene can be used in place of polyvinyltoluene in scintillating sheets
38. Finally, in photoconductor detector 46, copper can be used insteacl of
aluminum in connectors 98 and 102, electrodes 92 can be made thicker, for
better uniformity of response ~though possibly at the sacrifice of effec-
tive photoconductor area), electrodes 92 can include a thin layer of a con-
ductor such as indium deposited on the platinum to improve contact between
the platinum and the metallic coating of connectors 98 and 102~ and photo-
conductor strips 88 can be continuous instead of segmented if the excited
signals are strong enough to overcome the capacitance problem. Addition-
i~ll6757
ally, instead of photoconductor detector 46, a detector comprlsing a con-
tinuous p]anar sheet of photoconductor material having electrodes formed
in strips and deposited on top and bottom instead of one of the sides, as
in detector 46, could be employed. Construction is made easier, but such
a detector does not have the combined advantage of both high photon col-
lection efficiency and improved charge carrier collection efficiency of
detector 46.
Other embodiments within the invention will be apparent to those ~-
skilled in the art.
~hile I have shown and described, for simplicity, a 50 mm by 50
mm device, my most preferred embodiment is a 250 mm by 250 mm device, each
slit being the same width as in the embodiment shown and described, but
five times as long, and there being 250 slits rather than fifty. ~y most
preferred detector is the semiconductor structure disclosed. Preferably
after stepping through 180 my camera through a flyback returns to its
initial position. In preferred embodiments slit length is at least ten
times slit width; in my most preferred embodiments it is at least fifty
times slit width.
~lile it migllt be thought, as it was by some to whom I initially
:. .
20 disclosed my invention, that the increased flux available in each position
with slits instead of holes would be an advantage neutralized by the in-
creased number of positions required to be used, surprisingly this proved
untrue, owing to improved signal to noise ratios, permitting increasing
speed as well as resolution.
' ~.
- 22 -
~'
