EP0179095A1

EP0179095A1 - Gamma photon detection apparatus and method

Info

Publication number: EP0179095A1
Application number: EP19850901934
Authority: EP
Inventors: Meir Lichtenstein; Gregory Mack Jost
Original assignee: Individual
Current assignee: Individual
Priority date: 1984-04-26
Filing date: 1985-04-26
Publication date: 1986-04-30
Also published as: WO1985004959A1

Abstract

Un appareil de détection de photons gamma comporte un réseau (10) d'éléments allongés sensibles aux photons gamma (12) disposés chacun pour produire à l'intérieur de l'élément un courant primaire captif de photons lumineux en réponse à un photon gamma reçu. Un réseau (14) d'éléments scintillants allongés (16) s'étend angulairement par rapport aux éléments sensibles aux photons gamma. Chaque élément scintillant (16) répond à la réception des photons lumineux émis transversalement par les éléments sensibles aux photons gamma afin de produire dans l'élément scintillant respectif un courant secondaire captif de photons lumineux. Une série de ces courants indique conjointement l'emplacement d'un événement causé par le photon gamma reçu et ayant provoqué le courant détecté.A gamma photon detection apparatus has an array (10) of elongated elements responsive to gamma photons (12) each arranged to produce within the element a primary current captive of light photons in response to a received gamma photon . An array (14) of elongated scintillating elements (16) extends angularly relative to the elements sensitive to gamma photons. Each scintillating element (16) responds to the reception of light photons emitted transversely by the elements sensitive to gamma photons in order to produce in the respective scintillating element a captive secondary current of light photons. A series of these currents jointly indicates the location of an event caused by the gamma photon received and having caused the detected current.

Description

"GAMMA PHOTON DETECTION APPARATUS AND METHOD"

This invention relates to gamma photon detection apparatus, especially such apparatus for use in nuclear medicine.

Conventional systems for monitoring a radioactive agent administered to a patient are based upon the detection of photons emitted by the nuclei of the agent. However, these photons are emitted in all directions and a collimator is employed to detect the photons via a unidirectional window which collects only about .01 percent of the photons emitted. A theoretically better alternative is to detect the two oppositely emitted gamma photons which are the product of a positron annihilation in the body tissue. Positions emitted by the radioactive nuclei are found to travel between 2 and 4mm in tissue before they annihilate with an electron to produce two oppositely emitted gamma photons of 511KeV. However, such high energy photons are difficult to detect and according to conventional practice require a very expensive array of thick thallium activated sodium iodide crystals, each requiring in turn an individual photo-multiplier tube as a detector. For a complete tomographic system of satisfactory length, about one metre,, to.encircle the torso, it would be necessary to

2 utilize 15000 Nal or other crystals of 1cm area and a similar number of photomultipliers. International patent publication WO83/03683 by

Koslow Technologies Corporation discloses a radiation imaging system comprising an array of parallel conventional sσintillator detector crystals optically coupled at opposite ends to respective planar or curved arrays of parallel luminescent light-conducting channels. The channels of the respective arrays extend in mutually orthogonal directions: when the arrays are curved,, the system is said to have application as a positron ring camera. This imaging system has a number of disadvantages. The scintillator crystals are said to be conventional and may thus typically comprise thallium activated sodium iodide. Because of the length of these crystals and because of the tendency for annihilation photons (511keV) to cause more than one event in sodium iodide crystals, it is not possible to accurately determine the Z-axis co-ordinate or depth in the detector of the primary event. Moreover, it is important with this form of scan that a timing resolution be possible in order to pin-point the location of the annihilation event, but sodium iodide crystals are too slow; i any event the complete reliance on optical transfer to light channels increases the response time, which is important for time of flight. The distance of travel along the Z-axis to the light channels greatly reduces performance and makes it impossible to analyse multiple events. It is an objective.of this invention to provide practical gamma photon detection apparatus which may be adapted as a positron camera to monitor the gamma photon products of positron-electron annihilation. The invention accordingly affords, in its broadest aspect, gamma photon detection apparatus comprising: an ordered array of elongate gamma photonsensitive elements each arranged to generate within the element a confined primary current of light photons in response to a received gamma photon; and an ordered array of elongate scintillant elements extending angularly with respect to said gamma photonsensitive elements and each responsive to the receipt of light photons transversely emitted by said gamma photonsensitive elements to generate within the respective scintillant element a confined secondary current of light photons or electrons; whereby a set of_. said currents are jointly indicative of the location of an event which was caused by the received gamma photon and gave rise to the detected currents.

The gamma photonsensitive elements and/or the scintillant elements may comprise elongate solid bodies of scintillant or elongate tubes containing liquid scintillant. Preferably, the light photon absorption wavelength of the scintillant elements matches the light photon emission wavelength of the gamma photonsensitive elements.

Advantageously, said array of gamma photonsensitive elements is sandwiched between respective arrays of said scintillant elements. The scintillant elements of one of these arrays preferably extend at a substantial angle to the scintillant elements of the other of said arrays.

The invention further provides a method of detecting a gamma photon, comprising: arranging for said photon to cause a Compton event; confining a first proportion of the resultant light photons as a primary current; absorbing in a scintillant a second proportion of said resultant light photons emitted transversely of said current; emitting further light photons in response to such absorption; confining a fraction of the further light photons as a secondary current; and detecting said currents; whereby a set of said currents are jointly indicative of the location of an event which was caused by the respective gamma photon and gave rise to the detected currents .

The invention will now be further described, by way of exmple only, with reference to the accompanying drawings, in which:

Figure 1 is a schematic .representation of a segment of gamma photon detection apparatus according to the invention;

Figure 2 is a diagram illustrating detection of a gamma photon by the apparatus of Figure 1; and

Figure 3 is a partly sectioned schematic view of a positron camera comprising gamma photon detection apparatus according to the invention. A simplified orthogonal arrangement according to the invention is depicted in Figures 1 and 2 for purposes of illustration. This arrangement includes a primary array 10 of elongate gamma photonsensitive elements 12 consisting of co-planar parallel tubes filled with liquid gamma photonsensitive scintillant 13, viz, toluene containing 2, 5 diphenyloxazole, commonly referred to as PPO. A secondary array 14 of co-planar, parallel elongate scintillant elements 16 is arranged adjacent array 10 so that elements 16 extend orthogonally with respect to tubes 12. Each scintillant element 16 consists of a glass tube filled with the liquid light photonsensitive scintillant (17) 1, 4 bis(2-5(ρhenyloxazole) ) benzene, commonly referred to as POPOP. Tubes 12, 16 are conveniently 2cm in diameter and the respective concentrations of the scintillants are 5gms/l and lgm/1. The absorption wavelength of POPOP matches the emission band of PPO, so that the POPOP can absorb light emitted by PPO at a wavelength of interest and re-emit it equally in all directions at a longer wavelength, but the converse cannot occur. An example is the absorption of the PPO emission at 370nm and its re-emission by the POPOP at 440nm.

Each tube 12, 16 terminates at one end at a photo ultiplier tube (PMT) 18. Each PMT is coupled to its respective tube, using silicone grease, by a flat piece of glass glued at the end of the tube by means of epoxy resin. The output of each PMT 18 is fed to a fast co-incidence detector via an amplifier and also into a multi channel analyser. It will be noted that tubes 12, 16 are in close proximity but do not touch. It is also to be noted that tubes 12, 16 need not be glass but may be formed in a suitable plastics material, e.g. fused plastics film. 5 Furthermore, tubes 12 and 16 need not be of the same size or shape.

With reference to Figure 2 , when a Compton event is triggered at 31 in the scintillant 13 of a tube 12 by a positron annihilation gamma photon on path 30 at 511KeV,

^*Lo the gamma photon is scattered to path 30' and the average Compton recoil electron (170 eV) produces 2700 light photons 32 in all directions. A certain proportion of these, dependent on the critical angle c for the tube and thus on the refractive index of

"L5 scintillant 13, but typically about one third, is confined by total internal reflection within the respective tube 12 as a primary current 33 of light photons which travels away from the Compton event in both directions and is detected by the associated PMT

20 18. This indicates, e.g., the X coordinate of the event to a resolution of the tube diameter.

The light photons 32a not confined by total internal reflection escape transversely from the respective tube 12 and, on one side of the array, are

25 absorbed by the POPOP, by way of absorption/re-emission events 35, in only the tubes 16 adjacent the event. A certain proportion of re-emitted photons 34 in each tube 16 is again confined by total internal reflection within the tube as a secondary current of light photons,

3_Q detected by the respective PMTs 18. In this manner, the position of the original Compton event along the respective tube 12, i.e. the Y coordinate, is indicated by the PMTs 18 at its the peak reading, to the same resolution as before. The original gamma photon has been detected and the information about its location obtained.

It will be understood that the Compton reactions of the annihilation, as monitored with the detector configurations, and the gamma photons (if present) can be analysed in respect of either time or energy or both to determine the initial source of the radiation and eliminate misleading information due to, e.g. infra-patient scatter.

In a modification of the above arrangement, the secondary tubes 16 are of substantially smaller diameter than the primary tubes 12, say 1mm and 5mm respectively. Considering m pipes 12 as primary and n pipes 14 as _ secondary in this hypothetical two-dimensional detector the number of PMTs can be further reduced because the end of each tube end can be connected to a PMT. Many tube ends connect to each PMT. The number of PMTs can be reduced to 2 m + 2n . Consider the primary tubes and take the case of n = 100 tubes. At one end of the array, groups of 10 adjacent tubes can be coupled to each of 10 PMTs, and at the other end of the array, every 10th tube can be connected to a PMT. Hence an output in 2 of 20 tubes defines in which of the 100 tubes the event occurred. The same reasoning applies to the secondary scintillant tubes, hence a total of 40 PMTs can be used to uniquely address 10,000 pixels in the 100 x 100 array. If two Compton events occur (the first and second Comptons ^"of a single 511KeV gamma event) then usually not four, but eight PMTs register a pulse and the position of two events cannot be determined uniquely. In the two-dimensional detector this occurrence is rare, but in the three-dimentional detector to be considered later, it is common. One way to avoid this problem is to have a second layer of secondary tubes (on the other side of the primary tubes) and lay the tubes so that they extend at an angle of, say, 60° to each other.

This will produce extra positional information and allow the correct two positions to be determined. Alternately, time of flight resolution can uniquely identify the two scintillations in some cases. In a three-dimensional array -of about 20cms in depth of scintillator, most of the energy of 511KeV gamma photons will be deposited (half value layer = 7cms) . This takes place by a series of Compton events terminating in a photoelectric (PE) event. ^' For the purpose of illustration, if every Compton event is the average Compton event expected, then starting with a 51lKeV gamma ray, the Compton events yield 175, 95, 58, 37 and 26KeV electrons, leaving a 120KeV photon which undergoes many collisions before finally terminating in a PE event. If the total energy window for a detector is set just below the Compton edge, 350KeV, as is the practice for current detectors, then three to four average collisions will provide sufficient energy to exceed this value (Johns and Cunningham 1974) . -A practical detector would have the threshold window for each Compton event set at approximately 25 to 30KeV. Again, this would allow about four collisions yielding a total energy of 365 to 390KeV on average.

Once the threshold energy deposition is exceeded the detector must analyze the timing and location of the multiple Compton events to determine the position and timing of the first event.

If the timing resolution is 500 picoseconds then the chronological order of Compton scatters can be determined for distances greater than 15cm, by spacing out the layers of the detector such that inter Compton distances greater than 15cm become common.

Where inter Compton distances are shorter than 15cms then the energies deposited can be used to calculate scatter angles of the photons in the detector and hence a most likely path determined. The most likely paths of the two opposing annihiliation photons at their entry into the detector are compared and if both correspond to the real path then they must be co-planar. ^A graded absorber greatly reduces the body scattered radiation (low energy gammas) while affecting the 511KeV to only a slight extent. Such an absorber needs to be used to reduce detection of gamma rays already scattered in the body. A schematic diagram of an embodiment of positron camera according to the invention is depicted in Figure 3. This camera extends to three dimensions the principles described above with respect to Figures 1 and 2. The camera is arranged as a cylinder 8, 40cms long with a radius of 25cms and containing a depth of scintillator of 20cms. The tubes 12' of primary scintillant are arranged in concentric layers 10' and each tube runs parallel to the long axis of the cylinder. If n is the number of such tubes then the

-, number of PMTs is 2n* coding as explained above. If all

2 these primary tubes are made of 0.5cms solid plastics scintillant, then the camera would use 5600 separate tubes 12' and 150 high speed PMTs arranged into 40 layers for the primary photosensitive system alone. This number can be reduced. First, the average depth of the tubes (in the radial direction) is increased to 1cm so that only 20 layers are used. The number of tubes is halved. It may be useful to grade the tubes from less than 1cm in diameter on the inside to more than 1cm in diameter on the outside. This will lead to only a minor loss in resolution as geometric considerations dictate that very few gamma particles emerging from the patient and producing a scintillation in the detector, will enter at such an angle as to produce a significant error in the calculation of the flight path of the annihilation photons. Furthermore, the width (circumferential) of the primary tubes can be markedly increased if only the secondary tubes 16* are used to provide precise localization. If, for example, the average width is 5cms then the number of primary tubes is approximately 600 and the number of high speed PMTs is just 50.

Tubes 12* then provide precise energy and timing information, and only approximate positional information, enough to prevent ambiguity when reading the precise positions of multiple events using an array of secondary scintillant elements, as explained below.

The secondary scintillant tubes 16' need not be connected to the higher speed PMTs as they are not being used to provide fast timing information, thus reducing cost. In the illustrated arrangement, which is only one alternative, a layer 14' of secondary tubes 16' is placed between each pair of concentric layers 10' of primary tubes 12' and the tubes 16' extend at an angle of 45° to the long axis of the cylinder. Each secondary tube thus traces the path of a helix. The direction of the helix alternates from one layer 14* to the next so that each Compton event in a tube 12' contributes photons to two groups of secondary tubes, one above and one below, which are at 90° to each other and so provide precise positional information. The ideal form for a secondary tube 16' is that of a ribbon, 5mm wide and very thin, of solid plastics scintillant. There may be excessive attentuation of light transmission along such a thin layer of plastic so the alternative may be to use groups of thin glass or plastics tubes filled with liquid scintillator. The number of secondary units is then approximately 6200 and the corresponding PMT number is 158. A 40cm long x 25cm internal radius detector therefore uses 600 pieces of 40cm x 5cm x 1cm strips of plastics scintillant for the primary tubes 12* with just 50 high speed PMTs. It uses 6200 secondary scintillant units and 160 sensitive PMTs. Note that this does not involve sealing 6200 light tubes. Simply sharply bending (180°) the light tubes at high temperature will be adequate.

If the mechanical problems of curving and positioning the secondary tubes for a cylindrical detector are too difficult then the same result may be achieved by using a hexagonal detector with all pipes now being straight or of composite straight segments. There is little change in the number of tubes and PMTs. By comparison with other typical designs for positron cameras, the detector area is larger, being 40 x 2 x 25cms, which equals 0.62 square metres for the dimensions proposed above. This is more than twice that of other designs based on PMTs arranged in rings, but the major advantage with this detector is that there are no lead septa between rings. The detector is able to accept coincidences between any two regions of the detector, i.e. the detector unlike those with lead septa is able to detect photon pairs emerging from the body at any angle providing they are covered by the detector. This enables full use of the detector to be obtained. It has a large solid angle of acceptance, about 0.6 x 4/3τr , compared to conventional detectors.

Intra-detector resolution is equal to the separation between tube centres. This produces an intra-patient resolution of approximately half this, i.e. between 2.5mm and 3.7mm for 5mm tube thus enabling utilization of the small positron path lengths of 18p(15) positron. This is superior to the best figures quoted for ring detector systems with the order of 4.8mm (which is not in all axes) . Use of time of flight is highly advantageous and possible although this is not essential.

The number of PMTs can be around 300 or less, depending on design, which is less than the number in conventional multi ring systems. There is no loss of energy resolution compared to any of the conventionally used detectors other than Nal.

According to one alternative construction, it may be possible to make the front part of the detector using liquid or plastics scintillant, and have the back part of sodium iodide, and only choose these sequences that have one event in the organic scintillator and the remainder in the solid sodium iodide detector at the back. Most SllKeV photons are forward scattered after Compton reactions.

Another alternative approach to the use of thick primary and thin secondary tubes is to use tubes of the same diameter and fill them with a high concentration of primary scintillant together with a low concentration of thin secondary scintillant. If an event occurs in one tube, some of the light flash produced at the primary scintillant wavelength will be converted in the first tube to the longer wavelength and then transmitted along it, but provided the low secondary scintillant concentration is small enough, some of the primary scintillant light will escape to be converted in the adjacent tube which runs at an angle. Likewise, transfer can occur from the second tube to the first. This arrangement would simplify construction as all tubes are of the same thickness but the cost would be a loss of sensitivity and therefore count rate and energy resolution.

It may be possible to reduce the number of secondary scintillator tubes further by using long tubes and wrapping them around the primary scintillant tubes so that any one tube completes one entire circumference of the camera spiralling again at 45°. This would entail an average of 11 points of support (where each tube doubles back over the primary tubes or rods) per tube but reduce the number of 5.5 with a corresponding increase in the length of each individual pipe (to an average of 35 x 2 x 2 = 60cms) . The number of secondary PMTs is therefore reduced to approximately 70.

It is to be noted that some isotopes produce a photon as well as a positron. A high efficiency detector can detect this gamma photon as well as the annihillation photons and an approximate plane of its origin can be calculated. ^* In many cases this could supplement or replace TOF in providing positional information on the emitting isotope. This plane is in fact the surface of a cone.

Claims

1. Gamma photon detection apparatus characterized by: an ordered array of elongate gamma photonsensitive elements each arranged to generate within the element a confined primary current of light photons in response to a received gamma photon; and an ordered array of elongate scintillant elements extending angularly with respect to said gamma photonsensitive elements and each responsive to the receipt of light photons transversely emitted by said gamma photonsensitive elements to generate within the respective scintillant element a confined secondary current of light photons; whereby a set of said currents is jointly indicative of the location of an event which was caused by the received gamma photon and gave rise to the detected currents.

2. Apparatus according to claim 1 wherein said gamma photonsensitive elements comprise elongate solid bodies of gamma photonsensitive scintillant.

3. Apparatus according to claim 1 wherein said gamma photonsensitive elements comprise elongate tubes containing liquid gamma photonsensitive scintillant.

4. Apparatus according to claim 1, 2 or 3 wherein said scintillant elements comprise elongate solid bodies of light photonsensitive scintillant.

5. Apparatus according to claim 4 wherein said scintillant elements comprise tubes containing liquid light photonsensitive scintillant.

6. Apparatus according to any one of claims 1 to 5 wherein the light photon absorption wavelength of the scintillant elements matches the light photon emission wavelength of the gamma photonsensitive elements.

7. Apparatus according to any one of claims 1 to 6 wherein said scintillant elements are of substantially smaller cross-section than said gamma photonsensitive elements.

8. Apparatus according to any one of claims 1 to 7 wherein each of said arrays comprises a co-planar set of parallel, substantially equispaced elements, and wherein the planar sets are parallel and in mutually close proximity.

9. Apparatus according to claim 8, wherein said array of gamma photonsensitive elements is sandwiched between respective arrays of said scintillant elements.

10. Apparatus according to claim 9 wherein the scintillant elements of one of said arrays of scintillant elements extend at a substantial angle to the scintillation elements of the other of said arrays.

11. Apparatus according to any one of claims 1 to 8 wherein the gamma photonsensitive and scintillant elements are similar and comprise a mixture of a major proportion of gamma photon sensitive scintillant and a minor proportion of light photonsensitive scintillant.

12. A method of detecting a gamma photon, comprising: arranging for said photon to cause a Compton event; confining a first proportion of the resultant light photons as a primary current; absorbing in a scintillant a second proportion of said resultant light photons emitted transversely of said current; emitting further light photons in response to such absorption; confining a fraction of the further light photons as a secondary current; and detecting said currents; whereby a set of said currents is jointly indicative of the location of an event which was caused by the respective gamma photon and gave rise to the detected currents.

13. A method according to claim 12 wherein sets of currents for one or more Compton events, and the gamma photon(s) if present, are analysed in respect of time and/or energy to determine the initial source of the radiation.

14. A method according to claim 12 utilized to detect both a gamma photon emitted by an isotope and annihilation photon(s) generated by a positron emitted by the isotope.