Ion beam emitting device and detection system thereof
By integrating a gamma-ray detection system into the ion beam therapy system, and utilizing a combination of scintillator plates and photon sensors, the real-time performance and cost issues of traditional PET scanners and PG detection are resolved. This enables efficient and accurate monitoring of proton beam range and dose, improving the safety and precision of treatment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TERAPET LTD
- Filing Date
- 2021-01-09
- Publication Date
- 2026-07-10
Smart Images

Figure CN114930195B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to an ion beam emitting device and its detection system. The ion beam emitting device may specifically relate to a medical device for ion beam therapy (e.g., proton beam irradiation of tumors). The detection system is used to detect gamma rays. The detection system can be used for dose and range monitoring during ion beam therapy. The use of the detection system is not limited to irradiation therapy in humans or animals; it can also be used as a conventional PET scanner or Compton camera for other applications. Background Technology
[0002] Proton or heavy ion / ion beam therapy is one of the most precise external radiation therapy methods. Unlike photon beams, which have a high incident dose that gradually decreases as they pass through the body, ions can penetrate tissue and deposit most of their energy near the end of their track, the well-known Bragg peak. In this text, the general use of the term "ion" should also be understood to include both negatively and positively charged ions, including protons.
[0003] In today's most advanced ion beam systems for radiotherapy, as shown in Figure 1, the radiation dose is typically delivered by a narrow (usually a few millimeters) ion beam 1, which has a specific energy tailored to the patient and is laterally deflected using rapidly tilting scanning magnets 2. The penetration depth of the beam is controlled by adjusting the beam's energy; the beam's intensity, lateral position, and size are recorded by beam intensity and profile monitors 3 before reaching the target area. In this way, the tumor is irradiated in three dimensions (3D). The target area can be divided into isoenergy layers 4 corresponding to the penetration depth of a given beam energy set. Each isoenergy layer is further divided into a series of "spots" with different lateral coordinates, where each spot will receive a certain number of particles.
[0004] In practice, a treatment protocol (as shown in Figure 2) is typically developed before ion beam therapy can begin. During this protocol, computed tomography (CT) scans are usually performed, possibly in conjunction with magnetic resonance imaging (MRI) and / or positron emission tomography (PET) scans of the patient and target tissue. CT / MRI / PET scans are used to map the target volume and define the required dose distribution. Then, the proton delivery method is calculated: the angle of incidence of the proton beam into the body, the beam energy used to position the Bragg peak at the predetermined location, and the beam shape and size before entering the patient, as well as the number of protons to be delivered per “spot,” as shown in Figure 1.
[0005] This process typically takes place a few days or weeks before the actual treatment begins (as shown at time t0 in Figure 2), and a patient's treatment may take several weeks, spread across several sessions. During this time period, the location and volume of the target tissue can change significantly. To verify the effectiveness of the treatment protocol, the patient is usually scanned before each treatment, ensuring the patient's position (relative to the treatment table and imaging equipment) is correct. Furthermore, anatomical changes that may affect dose distribution can be detected. As shown in Figure 2, in addition to the patient's position and anatomical changes, a range of other factors can cause discrepancies between the planned and delivered dose distribution (particularly the beam range). Imaging artifacts may be present, especially in patients with metallic implants, which is not uncommon in patients who have undergone prior surgery.
[0006] Tissue models generated from images can be affected by systematic errors and the conversion from CT images to proton ranges. Tissue heterogeneity along the beam direction from the skin to the target area can introduce significant uncertainty in beam distance calculations. Patient positioning or immobilization methods may vary between treatment courses. During irradiation, slow movements due to respiratory motion, heartbeat, peristalsis, or changes in position from standing to supine or prone can cause organ movement, potentially shifting the Bragg peak from its intended location. Of particular concern are cases where tumors are close to critical organs such as the spine, optic nerve, or brainstem. Due to the steep dose descent gradient in the Bragg peak region, range deviations in ion therapy can have more severe consequences than range deviations in photon therapy. Range errors may mean that a portion of the tumor receives no radiation dose at all (under-irradiation), or that normal tissue at the distal end of the beam receives a higher dose (over-irradiation).
[0007] When protons / heavy ions pass through tissues, nuclear reactions occur, some of which result in the emission of gamma rays. Two types of gamma rays are detectable for therapeutic monitoring: 1) coincident gamma rays generated by generating positron-emitting isotopes; and 2) transient gamma rays generated by exciting target nuclei. The first type can be detected using positron emission tomography (PET) scans, which are widely used in neurology and oncology because they can monitor glucose metabolism and the uptake of other targeted radiotracers in specific organs and tissues. A particular use case for PET scans is ion beam therapy, where the penetration depth of the beam within the patient may be uncertain due to tissue heterogeneity, and safety margins must be employed to avoid dose to critical organs and / or ensure an adequate dose to the entire tumor. PET scans can provide accurate information about the location of irradiated dose deposition within the patient. However, there are many practical and technical challenges in using conventional PET scanners in ion beam therapy.
[0008] An example of a traditional PET scan used in a clinical workflow (Shakirin 2011) is shown below:
[0009] Offline PET: PET scans are performed after irradiation, typically with a delay of several minutes when the patient is transported from the irradiation room to another room equipped with a PET scanner. Only isotopes with lifetimes of a few minutes can be detected. While offline PET can be performed using a conventional PET scanner, the relatively long delay in PET acquisition (depending on the distance between the treatment and imaging rooms) prevents the acquisition of emissions of shorter-lived radionuclide species. Offline PET can only measure long-half-lived radionuclides. Bioelution of proton-induced PET activity further degrades performance, reducing the activity level in the target area and resulting in a "blurred" image.
[0010] - Indoor PET: PET scans are performed shortly after irradiation using a PET scanner located in the treatment room. While the delay between irradiation and scanning is reduced compared to offline systems, some delay still exists. Furthermore, this method extends the occupancy time of the irradiation room, effectively reducing the overall patient turnover.
[0011] - Intra-beam PET: Measure positron annihilation activity during irradiation using a custom PET scanner integrated into the treatment site or placed directly into the gantry. Real-time data acquisition enables more precise dose and range control. For long half-lives ( 11 C 13 N, etc.) and short half-life ( 15 O、 10 PET activity levels are highest in tissues (e.g., C-type PET) with minimal effects from bioelution. However, integrating a dedicated PET system into a beam delivery system for real-time measurements is costly and technically challenging due to the geometric constraints of integration within the ion beam device in the treatment environment and the intensive computation required for real-time measurements based on the large number of signals output from the gamma-ray detector. Furthermore, the performance of in-beam PET devices is limited, particularly because the transient gamma emission during ion beam delivery overwhelms coincidence events, and there is a delay in positron annihilation relative to irradiation (random emission based on the lifetime of the generated isotopes). Nevertheless, PET scanning techniques and image reconstruction methods are mature and validated, and are advantageous for continuing image acquisition after ion beam emission, even when the patient is outside the room. Even after irradiation, image quality can be improved by increasing the imaging acquisition time. Typically, the idle time between portals can also be used for imaging. Another significant advantage of PET scanning is its ability to measure total dose.
[0012] WO 2018 / 081404 A1 discloses a scintillator detector for PET scanners, which features edge detection and may radially stack multiple sensor blocks to achieve inherent interaction depth resolution. Compared to conventional pixelated scintillator arrangements, the edge detection disclosed herein reduces the number of scintillator elements while improving interaction depth measurement. However, due to the use of multiple scintillator plates, the number of photon sensors remains high, and the associated signal processing requirements for real-time acquisition are also demanding.
[0013] Another known technique for verifying proton beam range is through measuring instantaneous gamma-ray (PG) emission (Knopf 2013). PG emission occurs essentially simultaneously with proton beam emission, so there is virtually no delay between emission and detection during processing. Furthermore, point-by-point imaging can be achieved, similar to imaging closer to the particle tip range. Therefore, PG detection can rapidly detect significant range deviations. However, PG detection is an immature technique, and image reconstruction is quite complex. Image quality cannot be improved by increasing imaging time; detector performance and gamma-ray absorption efficiency are key factors affecting image quality. Moreover, total dose reconstruction is also difficult to achieve.
[0014] CN 107544086 A[1] discloses a combined Compton-PET imaging device based on a scintillator. The gamma-ray detection element is of the forward ("top-up") or side-up type, such as Figure 39 As shown. Compton scattering between radially separated detection modules 50 enables Compton camera imaging, while coincident photoelectric absorption enables PET scanning. A detection probe comprising multiple scintillation crystal arrays 51 separated by radial gaps is disclosed. However, these radial gaps are obscured by a photon sensor array 52 or a light guide (e.g., an optical fiber). Therefore, the main surface of each scintillation crystal array (typically facing the imaging volume) is covered by the photon sensor. However, CN 107544086 A does not disclose how or whether the desired Compton scattering angle accuracy is achieved.
[0015] Shimazoe 2018[3] disclosed a similar device in which a 2D surface-coupled scintillator array 50 (GAGG:Ce) is coupled to a 2D array 52 of a photon sensor. Summary of the Invention
[0016] The overall objective of this invention is to provide a safe, reliable, and accurate ion beam emission device.
[0017] A specific objective of this invention in the medical field is to provide a safe, reliable, and accurate ion beam emission device for patient irradiation therapy.
[0018] Another object of the present invention is to provide a cost-effective detection system capable of accurate real-time imaging of the volume of interest (also referred to herein as the “target region”) emitting positrons and transient gamma rays.
[0019] Another object of the present invention is to provide a cost-effective detection system for integration into an ion beam emission device to safely, reliably and accurately control ion beam emission in real time.
[0020] One specific objective of this invention in the medical field is to provide a cost-effective detection system for safe, reliable, and accurate real-time control of ion beam emission therapy.
[0021] According to one aspect of the present invention, a gamma-ray detection system is disclosed herein, the gamma-ray detection system comprising a detection module assembly including at least two detection modules configured to perform positron emission tomography (PET) scanning on a target region; each detection module comprising: a plurality of stacked monolithic scintillator plates, each monolithic scintillator plate having a main surface oriented generally facing the target region and a lateral subsurface defining an edge of the scintillator plate, the surface area of the main surface being greater than the surface area of the lateral subsurface; and a plurality of photon sensors mounted on each of the edges, configured to detect and determine the location of scintillation events in the scintillator plates within the plane of the main surface from gamma rays incident on the main surface. The gamma-ray detection system is also configured to function as a Compton camera, at least one scintillator plate not closest to the target region being configured as an absorber scintillator plate of the Compton camera.
[0022] This document also discloses an ion beam therapy system for ion beam irradiation of tissue regions, the ion beam therapy system comprising: a patient support; an ion beam emitter movable relative to each other about at least one axis of rotation; and a gamma ray detection system configured to perform transient gamma ray detection and PET scanning during, between and after ion beam irradiation.
[0023] In an advantageous embodiment, a plurality of photon sensors of at least two radially stacked scintillator plates are connected to a processing circuit configured to multiplex the readouts (also known as readout information or readings) of the plurality of photon sensors.
[0024] In an advantageous embodiment, a plurality of photon sensors of at least two azimuth-axially arranged scintillator plates are connected to a processing circuit configured to multiplex the readouts of the plurality of photon sensors.
[0025] In an advantageous embodiment, at least one radial gap is provided, the at least one radial gap being located between at least two of the plurality of stacked scintillator plates or between the at least two detection modules.
[0026] In an advantageous embodiment, the relationship between the height H of the radial gap and the thickness T of one of the plurality of scintillator plates is typically in the range of 200>H / T>2, and preferably in the range of 50>H / T>10.
[0027] In an advantageous embodiment, the plurality of photonic sensors include at least one strip multilayer photonic sensor extending along the edges of the plurality of layers.
[0028] In an advantageous embodiment, a plurality of the strip multilayer photonic sensors are configured on each edge side of the plurality of stacked scintillator plates.
[0029] In an advantageous embodiment, the at least one strip multilayer photonic sensor is a dual-end strip detector configured to measure the arrival time of signals at both ends.
[0030] In an advantageous embodiment, the plurality of photonic sensors comprises at least one monolayer photonic sensor located on at least one edge of each scintillator plate. In a preferred embodiment, particularly for scintillator plates having four or more edges (e.g., square or hexagonal scintillator plates), a monolayer photonic sensor is located on two or more edges of each scintillator plate.
[0031] In an advantageous embodiment, the single-layer photonic sensors of the sensor board assembly are interconnected by cross-wire connections or resistor networks, such that the readout is the sum and / or weighted sum of signals from a plurality of interconnected single-layer photonic sensors, wherein each sensor board includes a scintillator board and an associated electronic sensor.
[0032] In an advantageous embodiment, the single-layer photonic sensor of the module is multiplexed such that the number of readout signals is a subset of the total number of photonic sensors in the module.
[0033] In an advantageous embodiment, the gamma ray detection system further includes a light-reflecting interface layer and a light-absorbing interface layer located between (or on top of) at least two scintillator plates in the scintillator plates.
[0034] In an advantageous embodiment, the gamma ray detection system further includes a low refractive index gap, such as an air gap, located between at least two scintillator plates in the scintillator plates.
[0035] In an advantageous embodiment, the gamma ray detection system further includes an electro-optic shutter located between the edge of at least one scintillator plate and the photon sensor.
[0036] In an advantageous embodiment, the electro-optic shutter includes a light diffuser material and thickness configured to diffuse light in a flashing event near the edge.
[0037] In an advantageous embodiment, the surface area S of the main surface of the scintillator plate and the thickness T of the scintillator plate are within the range of 100 mm. 2 <=S<=40000mm 2 , 0.5mm <= T <= 30mm.
[0038] In an advantageous embodiment, the detection module assembly surrounds the target region and includes at least one gap or orifice for emitting an ion beam through it.
[0039] In an advantageous embodiment, the photonic sensor, whose light edge is coupled to one or more sides of a stack of scintillator plates, is mounted on a support plate that includes an edge connector for coupling to a signal processing circuit board, the edge connector minimizing the gap between adjacent detection modules of the detection module assembly.
[0040] In an advantageous embodiment, the radial gap satisfies the relationship H / (T1+T2)>5, where T1 and T2 are the thicknesses of the two scintillators surrounding the radial gap, and H is the height of the radial gap.
[0041] In an advantageous embodiment, the total radial thickness of the plurality of stacked monolithic scintillator plates is less than 19 mm.
[0042] In an advantageous embodiment, the gamma ray detection system includes two radially stacked scintillator plates, wherein the ratio between the thickness of the inner radial scintillator plate and the total radial thickness of the scintillator is in the range of 0.2-0.6.
[0043] In one advantageous embodiment, the photon sensor bias voltage of the photon sensor on a single scintillator plate can be independently adjusted or enabled / disabled.
[0044] In an advantageous embodiment, a photon sensor coupled to at least two radially stacked scintillator plates is connected to a processing circuit configured to apply Compton kinematics rules to determine whether two overlapping block events correspond to Compton scattering of forward or backward scattering following absorption.
[0045] In an advantageous embodiment, the processing circuitry is configured to reject events that appear to originate from main gamma rays entering the detector in a radially outward direction.
[0046] In an advantageous embodiment, the processing circuit is configured to utilize the interaction coordinates of photoelectric absorption as the LOR endpoint of small-angle forward Compton scattering gamma rays originating from electron-positron annihilation.
[0047] In an advantageous embodiment, the processing circuitry is configured to discard Compton scattering events that exceed the configurable primary gamma-ray energy-dependent scattering angle to improve angular resolution.
[0048] In one advantageous embodiment, an analog signal from an adjacent photon sensor is added before digitization or other multiplexing circuitry.
[0049] In an advantageous embodiment, a two-stage Compton camera can be implemented through inter-module scattering between adjacent detection modules of the detection module assembly.
[0050] In an advantageous embodiment, a two-stage Compton camera can be implemented through inter-block scattering between adjacent sensor plates.
[0051] In an advantageous embodiment, a three-stage Compton camera can be implemented through inter-block scattering between adjacent sensor plates.
[0052] Compared to conventional PET scanners, the advantages of this invention are significantly higher spatial resolution and lower cost scalability. Lower cost scalability is crucial for achieving high sensitivity, which is particularly important in the medical field, especially for range and dose verification in proton therapy. According to one advantageous aspect of the invention, the combined PET scanner / Compton camera can simultaneously utilize the advantages of PET and transient gamma imaging (PGI). Furthermore, this technology also has significant implications for other applications (e.g., whole-body diagnostic PET or combined PET / SPECT scanners) or other nuclear imaging fields. This combination addresses the major limitations of intrabeam PET and PGI, enabling the fusion of these two imaging techniques into a single device.
[0053] Other objects and advantageous features of the invention will become apparent from the claims and the following detailed description of embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description
[0054] Figure 1 is a schematic diagram of a typical ion beam setup;
[0055] Figure 2 is a flowchart of a conventional ion beam irradiation preparation protocol, illustrating the errors and factors that may lead to range uncertainty in irradiation treatments performed according to conventional therapies;
[0056] Figure 3 This is a flowchart of an ion beam irradiation preparation procedure, illustrating errors and factors that may lead to range uncertainty during irradiation therapy, as well as corrective measures implemented according to embodiments of the present invention to reduce range uncertainty and improve dose accuracy;
[0057] Figure 4 This is a perspective view of an ion beam therapy system with a gamma ray detection system according to an embodiment of the present invention;
[0058] Figures 5a to 5e These are schematic diagrams of five variations of the detection module components of the gamma ray detection system according to various embodiments of the present invention;
[0059] Figure 6a This is a perspective view of the detection module of the gamma ray detection system according to an embodiment of the present invention.
[0060] Figure 6b It is similar to Figure 6a The image shows the removal of some photon sensor support plates and photon sensors to view the interior of the detection module;
[0061] Figure 6c This is a detailed cross-sectional schematic diagram of a portion of the scintillator plate of the detection module according to an embodiment of the present invention;
[0062] Figure 6d The image shows a prototype detection module according to an embodiment of the present invention, taken from the radially distal side, the prototype detection module having four radially stacked sensor plates;
[0063] Figure 7 This is a schematic diagram of the detection module of a gamma ray detection system according to an embodiment of the present invention. The detection module includes a sensor board and a signal processing and control module.
[0064] Figure 8a This is a simplified schematic diagram of a detection module component of a gamma ray detection system according to an embodiment of the present invention, illustrating the detection of positron annihilation (corresponding to the PET scanner function);
[0065] Figure 8b It is similar to Figure 8a The diagram illustrates the detection of transient gamma rays (corresponding to the Compton camera function);
[0066] Figures 9a to 9c This is a schematic perspective view showing stacked scintillator plates of different shapes for a detection module according to different embodiments of the present invention;
[0067] Figure 10This is a simplified schematic diagram of the detection module of a gamma ray detection system according to an embodiment of the present invention, showing the arrangement of the scintillator plates;
[0068] Figure 11a This is a simplified schematic diagram of the detection module, illustrating the arrangement of the photon sensor according to an embodiment of the present invention;
[0069] Figures 11b to 11e It is similar to Figure 11a Other views of other different embodiments of the photon sensor arrangement according to the present invention;
[0070] Figures 12a to 12c This is a simplified schematic diagram of the photon sensor arrangement of a detection module (especially used as a Compton camera) according to various embodiments of the present invention;
[0071] Figure 13a This is a simplified schematic diagram of the photon sensor arrangement according to another embodiment of the present invention (particularly used as a Compton camera);
[0072] Figure 13b It is similar to Figure 13a A view of another embodiment (used as a Compton camera);
[0073] Figure 14a and Figure 14b This is a simplified schematic diagram of a pair of detection modules of a detection module assembly according to various embodiments of the present invention, wherein adjacent modules serve as inter-module Compton cameras according to embodiments of the present invention;
[0074] Figure 15a and Figure 15b This is a simplified schematic perspective view of a detection module assembly (which can be used as a PET scanner and a Compton camera) according to various embodiments of the present invention;
[0075] Figure 16a and Figure 16b This is a simplified schematic perspective view of the detection module components according to different embodiments of the present invention. Figure 16a A dual-head assembly is shown. Figure 16b A three-head assembly is shown, which can be used as a Compton camera for transient gamma ray detection and as a PET scanner for positron annihilation detection.
[0076] Figure 17a An arrangement of photonic sensors with a photonic strip detector having a detection module is shown according to an embodiment of the present invention;
[0077] Figures 17b to 17e The arrangement of photons on the four sides of a stack of scintillator plates with single and strip photon sensors, depending on the variant, is shown.
[0078] Figure 18This is a simplified schematic diagram of the photon sensor arrangement of the detection module of a gamma ray detection system with a photon strip sensor according to another embodiment of the present invention;
[0079] Figure 19a and Figure 19b This is a perspective view of two different detection modules of a gamma ray detection system according to other embodiments of the present invention, showing a photon sensor configuration with a cross-line connection arrangement for cross-line readout;
[0080] Figure 19c yes Figure 19a and Figure 19b A simplified circuit diagram read from the cross-line connection of an embodiment;
[0081] Figure 20a and Figure 20b This is a simplified schematic diagram of a scintillator board that includes a detection module for an electro-optical shutter (EOS). Figure 20a The EOS is shown in the open state. Figure 20b The EOS is shown in a closed state;
[0082] Figure 20c This is a simplified schematic perspective view of an example of a scintillator board stack, where the top and bottom layers have open EOS and the three intermediate layers have closed EOS;
[0083] Figure 21a This is a simplified schematic perspective view of a scintillator plate with a single photon sensor;
[0084] Figure 21b This is a graph showing the average number of photons detected by a single photon sensor along the right edge after a flashing event;
[0085] Figure 22 This is a graph showing the error distribution in a conventional PET scanner using photon sensors, each with a receiving surface area of 1x1mm;
[0086] Figure 23 This illustrates triangles with three, four, and six sides respectively. Figure 9c ),square( Figure 9a ) and hexagon ( Figure 9b A graph showing the ratio between the number of photon sensors in a conventional PET scanner and the number of photon sensors (i.e., referred to as “pixels”) per edge of a scintillation plate in the shape of a scintillation plate; a conventional PET scanner; and the number of photon sensors according to the present invention.
[0087] Figures 24 to 26 This is a diagram illustrating the spatial accuracy of an embodiment of the present invention;
[0088] Figure 27This is a schematic diagram of the detection module components of a gamma ray detection system according to an embodiment of the present invention, used for triangulation of the light source position of a three-gamma event;
[0089] Figure 28 This is a schematic diagram of a sensor plate arranged in an azimuth-axial configuration according to an embodiment of the present invention;
[0090] Figure 29 This is a schematic diagram of a 2x2 azimuth-axial arrangement of multiplexed readouts of a sensor board according to an embodiment of the present invention;
[0091] Figure 30 This is a schematic diagram illustrating how to achieve a larger field of view by rearranging axially-oriented sensor plates;
[0092] Figure 31 This is a schematic diagram of advantageous arrangements of various embodiments of the present invention for achieving an extended axial field of view, compared to conventional detection modules;
[0093] Figure 32 This is a schematic diagram of the detector sensitivity of an axial line source used for different sensor board arrangements;
[0094] Figure 33 This is a schematic diagram of an annular assembly of a sensor plate having two radial gaps and two+2+2 radial layers arranged in an axial-orientation manner according to an embodiment of the present invention;
[0095] Figure 34 The detection probability diagrams for different energies, total radial thickness of the scintillator, and scattering / absorption relationships of the dual radial layer configuration according to various embodiments of the present invention are shown.
[0096] Figure 35 A graph showing the coordinate components of the angular accuracy of Compton reconstruction for various combinations of radial gaps and scintillator plate thicknesses according to embodiments of the present invention is illustrated.
[0097] Figure 36 A graph showing the probability of absorbing the main gamma ray directly through photoelectric absorption (“pe”) or through a single Compton scattering (“Compton”) or a combination of both (“pe or Compton”) is provided, with the left graph showing the probability of coincident absorption.
[0098] Figure 37 A graph showing the contribution of the angular accuracy of the Compton scattering angle reconstruction according to an embodiment of the present invention, as well as the energy resolution limitation and spatial resolution limitation, is presented.
[0099] Figure 38 A variant of a whole-body scanner (PET or Compton-PET, depending on the configuration) employing a rotating dual-head configuration with two radial gaps is shown according to an embodiment of the present invention;
[0100] Figure 39 This illustrates a front-coupled detection module with a scintillation crystal and a photon sensor array in the prior art;
[0101] Figure 40 A 16:4 multiplexing variant (common anode) of a photon sensor along the same edge according to an embodiment of the present invention is shown, whereby the sum of currents of adjacent photon sensors can be read.
[0102] Figure 41 The effects of energy and spatial uncertainties on the reconstruction of the Compton scattering angle are shown. Detailed Implementation
[0103] Referring to the attached diagram, from Figure 4 The text begins by illustrating an ion beam therapy system 6 according to an embodiment of the invention, particularly for ion beam radiotherapy or proton beam irradiation of a tissue region. In this embodiment, a patient 5 is positioned on a patient support 7, which is movable relative to an ion beam emitter 8 about at least an axis of rotation and translation. Specifically, the patient support 7 can translate at least along at least one axis, particularly a horizontal axis X relative to a fixed reference point (e.g., the ground); the ion beam emitter can rotate about the horizontal axis X relative to the fixed reference point (e.g., the ground). However, the patient support and / or the ion beam emitter can translate and / or rotate along and about multiple axes up to achieving full three-dimensional movement, thereby enabling the ion beam emitter to be positioned at any position and angle relative to the patient.
[0104] The ion beam therapy system also includes a gamma ray detection system 10. In some embodiments, the gamma ray detection system 10 can also be moved relative to the patient support along or around one or more axes. In one embodiment, the gamma ray detection system can be moved along at least one translational direction (especially along the axial direction), and in a variant, it can also be rotated in an azimuth direction in cooperation with the ion beam emitter 8.
[0105] However, in various variations (not shown), there may be a gamma-ray detection system that is stationary relative to a fixed reference point or a gamma-ray detection system that translates only relative to a fixed reference point (e.g., the ground).
[0106] In a preferred embodiment, the gamma-ray detection system 10 includes a detection module assembly 13, which is typically annular or polygonal. In one embodiment, the detection module assembly may include an opening 42 through which an ion beam emitter 8 can transmit ions (e.g., protons), such that the emission direction of the ion beam emitter 8 is substantially in the same plane as the detection module assembly. This allows for the efficient detection of gamma rays emitted from a target region receiving the ion beam simultaneously. For example, the detection module assembly 13 may have a general "C" shape to provide an opening between the opposite ends of the C-shape through which the ion beam emitter 8 can transmit ions. However, in various variations, a substantially closed annular / polygonal, such as a generally cylindrical detection module assembly, may be provided with an aperture through a portion thereof, allowing the ion beam to be transmitted through it (variants not shown).
[0107] The length of the detection module assembly 13 in the direction of the rotation axis X of the ion beam emitter 8 (also referred to herein as the axial direction) can range from about 5 cm to about 200 cm, depending on the variant. For detection configurations with a shorter axial length, translation of the detection module assembly 3 (possibly in conjunction with the ion beam emitter) can be achieved during ion beam therapy. According to one embodiment, the detection module assembly can also be translated for scanning the target area after ion beam emission or during diagnosis. Since the length of the detection module assembly is sufficient to extend across the entire target area, a detection module assembly that is static relative to the patient can be used, whereby the detection system may not track the displacement of the ion beam or ion beam emitter.
[0108] It can also be noted that the movement of the detection module components can be parallel to or correspond to the movement of the ion beam emitter, or can follow other movements configured to optimize the detection of transient gamma rays and positron annihilation gamma rays emitted from the target based on the target location, the target environment, and the tilt position and angle of the ion beam emitter 8. Optimal movement of the ion beam emitter and detection system can be achieved, in particular, by calibrating the system on the sample tissue.
[0109] The gamma-ray detection system 10 used in the ion beam therapy system 6 according to various embodiments of the present invention has an important advantage: detection can be performed in real time during proton beam emission, simultaneously capturing both transient gamma rays and positron annihilation gamma rays. Furthermore, positron annihilation gamma rays emitted within a specific time after proton beam emission or between consecutive proton beam emission pulses during treatment can be detected. This enables continuous monitoring of proton beam absorption relative to the target region and allows for adjustments based on feedback from the detection system to precisely locate the target region, taking into account any movement of the target region during or after treatment, and avoiding other problems previously discussed relative to conventional systems, such as elution effects.
[0110] Impulsive gamma rays emitted from the volume of interest can be detected by a detector used as a Compton camera, while positron annihilation gamma rays, typically with lower energies (511 keV), can be detected by a detection module using the operating principle of a PET scanner. Both detection methods are integrated into the detection module of a detection assembly according to embodiments of the present invention, as will be further described below. It can be noted that PET detection can be performed during ion beam emission, between two ion beam emitting pulses, and after ion beam emission; alternatively, PET detection can be initiated only between and after ion beam emission pulses. During ion beam emission, the rate of impulsive gamma rays is very high, which may make the measurement of overlapping gamma rays in positron emission annihilation less accurate and reliable. However, for a period of time after ion beam emission, the emission of impulsive gamma rays is low, and the emission of positron annihilation gamma rays persists for a certain period (as is known per se), making it possible to perform measurements during and after ion beam emission.
[0111] refer to Figure 5a and Figure 5b The illustrations schematically show two different embodiments of the detection module assembly 13 of the gamma-ray detection system 10 according to various embodiments of the present invention. Although the detection module assembly 13 in these illustrations is shown as a substantially completely closed annular / polygonal shape, it should be understood that a portion of it can be removed to provide a generally “C” shape with an opening for the ion beam emitter to transmit the ion beam to the target region. It should also be understood that the detection module assembly 13 may include spatially separated detection modules, such as a “dual-head” configuration. Figure 5e () or "four-head" configuration (Figure 5f).
[0112] Detection module component 13 includes multiple detection modules 14. Figure 5a In the illustrated embodiment, the detection modules 14 can be arranged in an aligned manner to form segments; or, in Figure 5b In another embodiment shown, the detection modules 14 can be radially staggered to take the radial beam into account from the rotation axis X. However, various other configurations may also be adopted, whereby the number of modules aligned to form segments or positioned in a generally circular or polygonal arrangement (as shown) may vary.
[0113] The detection module 14 is configured to function as both a Compton camera 11 and a PET scanner 12, as described in further detail herein. However, the detection module according to the invention may be used solely as a Compton camera or solely as a PET scanner, depending on the application.
[0114] Each detection module 14 includes a plurality of stacked scintillator plates 16 and a plurality of photon sensors 18. The scintillator plates have a primary surface 40a oriented generally toward the target region or axis X, and a lateral secondary surface 40b defining the edge or contour of the scintillator plate. For simplicity, the lateral secondary surface 40b should also be referred to herein as the “edge.” In various embodiments (not shown), one or more detection modules may also be added at the axial end of the target region or the imaging volume of interest, or at an intermediate location between the axial end and a radial position. In an advantageous embodiment, the stacking orientation of the scintillator plates in the detection module 14 is orthogonal to the primary surface. The photon sensors 18 are located on the edges of the scintillator plates 16.
[0115] According to various embodiments, the detection module may include a stack of scintillator plates without radial gaps; or, according to other embodiments, the detection module may include a stack of scintillator plates with at least one radial gap 17.
[0116] The radial gap 17 is particularly useful for the normal operation of the Compton camera 11, in which some of the scintillator plates act as scatterers for the Compton camera, while the other scintillator plate acts as an absorber. The scattering and absorbing layers can be determined by applying the Compton kinematics rules or timing.
[0117] The interface between the scintillator plates may include an interlayer reflector 28 that is optically reflective to conduct light during a scintillator event to the edge of the scintillator plate while allowing gamma rays to pass through it.
[0118] As an alternative to or complement to interlayer reflectors, scintillator plates can be separated by one or more low-refractive-index gaps 31 (e.g., air gaps) or low-refractive-index solids (e.g., polymer materials). The low-refractive-index gaps 31 have the effect that the surface of the scintillator plates acts as an internal reflector to improve light transmission from scintillation events to the edges of the scintillator plates, while allowing gamma rays to pass through these layers. As a complement to or alternative to interlayer reflectors, light-blocking layers or absorbing layers 29 can be inserted between the scintillator plates to combine with the low-refractive-index gaps to prevent interlayer light contamination.
[0119] The reflective or absorptive interface layer can be formed on one side of the scintillator plate or on both sides of a stacked scintillator plate.
[0120] The main surface 40a of the scintillator plate is the surface on which gamma rays are typically incident, and the edge 40b (e.g., typically orthogonal to the main surface and extending between opposite sides of the scintillator plate) forms the edge of the scintillator plate, along which the photon sensor 18 is arranged. In a preferred embodiment of the invention, the surface area S and thickness T of the main surface of the scintillator plate can be within the following ranges:
[0121] 100mm 2<=S<=40000mm 2 ,and
[0122] 0.5mm <= T <= 30mm;
[0123] More preferably, within the following range:
[0124] 400mm 2 <=S<=40000mm 2 ,and
[0125] 1mm <= T <= 10mm.
[0126] The preferred range aims to achieve the following effects: optimizing the relationship between the accuracy of interaction depth (DOI) measurement (Z-direction) and / or reducing the number of readout channels and the detection accuracy of the main surface of the scintillator plate (XY plane) at the scintillation location. The optimal range may vary depending on the application scenario.
[0127] An edge light diffuser material layer 26 can be disposed along the edge 40b of the scintillator plate. The function of the edge light diffuser material layer 26 is to diffuse gamma rays, so that light incident on the scintillator very close to one edge 40b is distributed across several adjacent photon sensors.
[0128] The edge 40b of the scintillator plate may also be provided with a detector-scintillator optical interface 22, which includes an interface material that optimizes light transmission to the photon sensors through the edge and / or enables consistent, predictable photon transmission across the layer to avoid inconsistencies that may occur due to non-constant interfaces (e.g., due to air, variable gaps, etc.). This optical interface also serves to diffuse light from scintillation events occurring near the scintillator edge onto multiple photon sensors to improve spatial resolution.
[0129] In addition, an electro-optical shutter (EOS) 24 can be provided along one or more scintillator plates. The electro-optical shutter is opened (optically transparent, 24a) or closed (absorbed or reflected, 24b) by electronic operation to allow photons to pass through the edge to reach the photon sensor or to prevent photons from passing through and reaching the photon sensor, depending on the operating state of the detection module 14.
[0130] The relationship between the height H of the radial gap 17 and the thickness T of a scintillator plate is generally in the range of 200>H / T>2, and preferably in the range of 50>H / T>10.
[0131] In one variant, different sensor plates can have scintillator plates of varying thicknesses. For example, the radially inner scintillator plate, which primarily acts as a Compton scatterer, can be thinner to reduce the probability of absorption or rescattering of Compton-scattered gamma rays in the radially inner layer. The radially outer scintillator plate can be thicker to increase the probability of complete absorption. The thickness of the scintillator plate can vary with radial position or radial sequence position.
[0132] The relationship between the height H of the radial gap 17 and the thicknesses T1 and T2 of the two scintillator plates radially surrounding the gap is generally in the range of 100>H / (T1+T2)>1, and preferably in the range of 25>H / (T1+T2)>5.
[0133] It can be noted that the radial direction referred to in this article corresponds to the direction Z shown in the attached figure illustrating the detection module.
[0134] Photon sensors 18, arranged along the edge 40b of the scintillator plate 16, can be mounted on a photon sensor support plate 20. This support plate can be, for example, a circuit board with circuitry for interconnecting the photon sensors to the signal processing and control system 30 of the detection module 14. The support plate 20 can also be a flexible or rigid-flexible circuit. Flexible circuitry can cover one or more edges of the module and fold around the edges of the radial stack of the scintillator plate, optically coupled to said edges. Thinning the photon sensor support plate is advantageous to minimize dead zones between the detection modules.
[0135] The support plate may include protruding guide elements to align the edge of the scintillator relative to the photon sensor.
[0136] For example, the signal processing and control system 30 of the detection module 14 may include a circuit board 32 and electronic components 34 mounted thereon, including analog components for signal filtering, signal shaping, multiplexing, and combining individual photon sensor signals, as well as photon sensor bias voltage components, a microprocessor, and memory for processing and controlling the detection module. The circuit board 32 may be mounted at the outermost radial end of the module and includes connectors 36a and 36b for connecting the circuit board 32 to the photon sensor support plate 20 and further to the electronic control system of the gamma-ray detection system 10 for image reconstruction, such as... Figure 7 Chinese combination Figure 6a and Figure 6b As shown. The support plate 20 can be configured as a silicon photomultiplier tube array plate with an edge connector 36b, which advantageously minimizes the gap between adjacent detection modules 14 of the detection module assembly 13.
[0137] In one variant, some analog components (e.g., signal shaping components, filtering components, or signal multiplexing components) are mounted directly on the photon sensor support plate, close to the photon sensor.
[0138] The signal processing components include components for digitizing analog signals, such as triggering, timestamping, and energy measurement (e.g., charge integration or time-out). Other processing components can be used for low-level event processing such as event verification or event rejection, using predetermined or configurable rules based on Compton kinematics, photon sensor thresholds, energy thresholds, the number of simultaneously triggered sensor boards or photon sensors, or other applicable rules determined according to previous detector calibration.
[0139] The distribution of analog and digital signal processing components allows them to be connected to a radial stack of multiple sensor boards, meaning that one or more radial sensor board stacks can “share” the analog and digital signal processing components.
[0140] Photon sensor 18 may include a single-layer photon sensor 18a and / or a strip multilayer photon sensor 18b. In some embodiments, photon sensor 18 may include a strip multilayer photon sensor 18b extending radially (Z-direction) across the edges of multiple stacked scintillator plates 16 and a single-layer photon sensor 18a located on a single scintillator plate. Detection module 14 may include multiple strip multilayer photon sensors 18b located on each side of the module, and sensors located on each side of the module, only on a portion of the side, or only on one side (depending on the specific location). Figures 17a to 17e A single-layer photon sensor 18a (as shown in the variant in the example) is used to determine one or more layers of absorbed incident gamma rays, while a strip multilayer photon sensor 18b (possibly in combination with the single-layer photon sensor shown) is used to determine the incident position of the absorbed gamma rays in a plane orthogonal to the radial direction (i.e., a plane parallel to the main surface 40a of the scintillator plate 16).
[0141] A key advantage of using the striped multilayer photonic sensor 18b is that it reduces the number of channels that need to be processed by the signal processing and control electronics for a given number of stacked scintillator plates without compromising measurement accuracy. This significantly reduces data processing requirements and associated equipment costs, or alternatively, achieves higher interaction depth measurement accuracy by using a larger number of stacked scintillator plates for a given number of readout channels.
[0142] In one variant, such as Figure 18As shown, instead of a single-layer photon sensor 18a, only strip multilayer photon sensors 18b extending between the stacks of scintillator plates can be set to determine the interaction depth. However, these strip multilayer photon sensors are configured to measure the time difference between the two ends of the strip multilayer photon sensor, which is related to the illumination position on the strip, from which the layer where the scintillation event occurred can be inferred.
[0143] In yet another embodiment, to reduce the number of readout channels, the single-layer photonic sensors 18a arranged in rows can be configured with cross-connected wires, or interconnected by a resistor network, or typically by multiplexing, such as... Figures 19a to 19c As shown, this reduces the number of channels. It can be noted that... Figure 19c The diode in the example shown represents a silicon photomultiplier tube (SiPM). Multiplexed readouts enable the determination of the scintillation position in the scintillator plate by the intersection of rows and columns of a single-layer photon sensor, while reducing the number of channels processed electronically. Figures 20a to 20c The electro-optical shutter, schematically shown, can optically block a number of scintillator plates during very high-rate instantaneous gamma emission to prevent photonic sensor signals from multiple simultaneously triggered sensor plates from being superimposed in the multiplexer readout, which would destroy information from a single triggered sensor plate.
[0144] By using digital silicon photomultiplier tubes as photon detectors, individual units of the strip photon detector can be enabled / disabled to shield (ignore) light originating from a selected scintillator layer. This is an alternative to achieving the same function as an optical shutter.
[0145] It can be noted that in the Compton camera arrangement, the detection module 14 may include multiple scattering sensor plates 14s, the number of which and the surface area of which are greater than [a certain value]. Figure 15a and Figure 15b The absorber modules 14s shown are illustrated. In this configuration, the function of the positron emission tomography (PET) scanner is performed by multiple larger modules 14s that are radially closer to the target area, while the absorber modules 14a that are radially farther away from the target area act as absorber modules to ensure the proper functioning of the Compton camera. This configuration also allows the Compton camera and PET scanner to function simultaneously, while further reducing the number of readout channels used for signal processing.
[0146] Therefore, the PET scanner functionality according to embodiments of the present invention can be implemented using a stack of monolithic scintillator crystals. The scintillating light propagates laterally from the interaction point, and is detected by multiple photon detectors. For example, the photon sensor can be a SiPM (analog or digital) or other types of detectors known per se.
[0147] To improve the spatial resolution of events near the scintillator side, as previously described, an optical (non-scintillator) "diffuser" material 26 can be inserted between the scintillator and the photon sensor. This causes the light emitted by the gamma interaction to diffuse across multiple pixels, even if the interaction occurs near the scintillator side. Examples of diffuser materials can include glass, silicone rubber, etc., and the thickness can vary, thereby enabling different diffuser shapes to optimize light yield on the photon sensor. As a complement or alternative to the diffuser material, as described above, a thin interface optical layer 22 can be provided between the scintillator plate edge 40b and the photon sensor, for example, comprising grease, adhesive, or sealing glue.
[0148] Light propagation from the interaction point to the photon sensor can be achieved through total internal reflection. This can be achieved by inserting a material with a low refractive index, such as air, between the scintillator plates. The advantage of using air is that it does not introduce any specific manufacturing limitations and does not degrade over time, with different applications, or due to radiation. Another feature is the stacking of highly reflective materials or thin films (e.g., enhanced specular reflectors (ESRs)) between the scintillator plates and the layers. It should be noted that the reflectivity needs to remain sufficiently stable during device operation and / or between foreseeable device calibrations.
[0149] Because signals from various layers can be resolved in the embodiments of the present invention, the PET scanner function inherently possesses interaction depth capability. Depth resolution is primarily determined by the thickness of the scintillator plates. The thicker the scintillator plate, the higher the interaction depth (DOI) resolution. However, as the number of scintillator plates 16 increases, the number of photon sensors 18 required also increases. To mitigate this problem, embodiments of the present invention include the use of elongated photon sensors extending to the edges of multiple scintillators, i.e., the previously mentioned strip multilayer photon sensors. Thus, a single channel can measure light from multiple scintillator plates. To distinguish interacting scintillator plates, single-plate pixels are contained in at least one column of photon sensors on each side, i.e., the previously mentioned single-layer photon sensors.
[0150] The number of layers covered by a strip detector can be adjusted based on a predictable range of event rates: for low count rate applications, the probability of multiple gamma interactions occurring "simultaneously" (e.g., during an overlap window or within the response time of a photonic sensor) across several layers is negligible. For SiPMs, the actual dead time between events is typically on the order of several hundred nanoseconds.
[0151] In Compton camera imaging, a first "scattering" layer is typically used, where X-rays / gamma rays interact through Compton scattering and deposit a small fraction of their initial energy, E1. The X-rays / gamma rays are emitted at an angle Θ slightly different from their initial direction, the angle variation being related to the deposited energy. This photon is absorbed by a second scintillator plate (i.e., an "absorber").
[0152] By calculating the energy deposition E1+E2 in the two scintillator plates and the interaction coordinates of the two layers, the initial position information of the initial ray can be inferred.
[0153]
[0154] Where m0 is the mass of the electron and c is the speed of light. For energy identification, it is generally required that the measured energy deposition E1+E2 be matched with the transient gamma emission peak of one (or more) isotope states of interest to reduce the probability that the gamma rays detected in the first layer have been previously Compton scattered, for example in a patient.
[0155] Interaction coordinates in the two layers need to be measured to determine the angle. Unlike PET reconstruction, where the LOR can be plotted between the interaction coordinates of coincident scintillation events, Compton imaging produces a “cone” emanating from the interaction point of the scattering layer, the orientation and aperture angle of which are derived from energy and coordinate information from the two independent layers.
[0156] When the detector acts as a Compton camera, its angular accuracy is primarily determined by two components: (1) the energy accuracy of E1 and E2 can be determined from them, and (2) the spatial coordinate accuracy of the line between absorption and scattering events is defined, through which a Compton cone with angle Θ is generated for image reconstruction, such as... Figure 41 As shown. Based on the following assumptions, the magnitudes of these two components were studied in detail, and summarized in... Figure 35 (Coordinate components only) and Figure 37 In (coordinates and energy components):
[0157] • The energy accuracy is roughly proportional to the square root of the deposition energy, and it is assumed that the energy accuracy of LYSO et al. is about 10% at 511 keV.
[0158] • The spatial coordinate accuracy Δx is ±0.7mm in both the azimuth and axial directions (XY).
[0159] Since these two components are largely independent, they can be added orthogonally. Energy accuracy is a fundamental limitation inherent in scintillation crystal materials and is a formidable challenge due to the energy resolution limitations of photon sensors. Therefore, the contribution of spatial coordinate accuracy can be appropriately considered, at least much less than that of the energy component. This is largely achievable at H / T >= 10, for example, for most energies of interest in transient gamma imaging during proton therapy.
[0160] Figure 37 The diagram shows the angular accuracy as a function of the scattering angle in the energy range from 0.511 MeV to 7.0 MeV at H / T = 10. For scattering angles above approximately 40 degrees, the overall angular accuracy (marked solid line) depends primarily on the contribution of energy resolution (solid line), while the contribution of coordinate accuracy is less significant (dashed line).
[0161] For large-angle scattering, especially backscattering events (Θ>90 degrees), angular accuracy is significantly reduced. Therefore, to improve image quality, it is advantageous to reject events where the reconstructed scattering angle exceeds a certain configurable value. Different upper limit thresholds can be used for different energies. This threshold can be implemented as an energy discrimination threshold in the absorber or scatterer scintillator plate, or as the actual angle threshold applied after scattering angle reconstruction.
[0162] To accurately determine the lateral coordinates of Compton scattering or photoelectric absorption, a minimum energy deposition is required to obtain a sufficiently accurate reading from the photonic sensor of the scintillator. For forward scattering events, the energy deposition in the scattering layer may not reach this threshold, resulting in indeterminate lateral coordinates. In PET scan mode, this means that the coordinates on at least one side of the response line (LOR) are indeterminate. In this case, it would be advantageous to use the coordinates of the absorption event instead, as it would carry most of the original energy. If scattering and absorption occur within radially sufficiently adjacent scintillator plates, the coordinates of the absorption event can be used as the endpoint of the LOR (given the smaller scattering angle and shorter drift length between Compton scattering and absorption).
[0163] Figure 6b , Figure 10 , Figures 12a to 12c , Figure 14b An example of the detection module 14 is shown, in which one or more scintillator plates 16 facing the target are separated from one or more other scintillator plates 16 by radial gaps 17. The scintillator plate closest to the target forms a "scattering layer". Angular resolution is improved by introducing radial gaps in the other layers. The length of the radial gap H can be adjusted to optimize the conflicting constraints of bringing the PET layers as close as possible (for compactness and imaging reconstruction accuracy) and as close as possible to the target, while maximizing angular resolution.
[0164] 3rd grade Compton camera
[0165] In one variant of the invention, at least one module may also be configured to additionally function as a 3-stage Compton camera, which requires at least two radial gaps (i.e., at least three radially separated sensor plates).
[0166] Cost considerations for photonic sensors
[0167] In the forward detection module including the main surface coupling ( Figure 39 In traditional PET scanners, the total area of the photon sensor components used for scintillating blocks or individual scintillating rods or pixels is approximately equal to L. 2 Where L is the length of the scintillator element perpendicular to the radial direction. When using a square edge-coupled detector, the total area of the photonic sensor is 4LT, where T is the radial thickness of the scintillator. To ensure that the area of the edge-coupled photonic sensor is smaller than that of the main-surface coupled photonic sensor, L > 4T is required.
[0168] With a conventional scintillator radial thickness of approximately 20 mm, the L-side needs to be at least 80 mm to make the photon sensor area equal to or smaller.
[0169] Considerations for count rate
[0170] Specifically, for range verification applications, the rate of emitted transient gamma rays can be very high. For example, at a high therapeutic proton rate of 1.2E10 protons / second, the rate of transient gamma rays would be on the order of 1E9 / s (Rohling 2017). At a radial distance of 30 cm, this corresponds to approximately 0.1 gamma / cm² / µs; or, for a 5 cm x 5 cm square detection module, this corresponds to approximately 2–2.5 gamma / µs. Care should then be taken to ensure that the detector is not saturated or masked by the high transient rate of gamma rays. Events where the deposited energy does not match the peak of the transient gamma ray of interest can be rejected to improve the ability to distinguish between Compton-scattered gamma rays from one layer to another and two independent gamma rays detected simultaneously in two layers.
[0171] Various embodiments of the present invention include different configurations of the detection module that can be adjusted according to the expected count rate.
[0172] Configuration 1 - Independent layer equipped with pixel detector
[0173] exist Figure 11b , Figure 11cIn the first configuration shown, each scintillator plate is equipped with an individual photon sensor at its edge. This allows for the measurement of the lateral position and energy of gamma rays photoelectrically absorbed in one layer, or the measurement of the lateral position and energy of gamma rays that are Compton scattered in one layer and absorbed in another. Of course, second Compton scattering in a second layer and other interactions can also be used, but for the sake of brevity, this document focuses on the aspects of using the invention as a combined single-scattering Compton camera / PET scanner.
[0174] Configuration 2 - Strip detector spanning multiple layers
[0175] The second configuration ( Figure 11a In this approach, some photon sensor "pixels" have been replaced with "strips" spanning multiple layers. This offers the advantage of reducing the total number of channels and readout complexity. To be able to identify the layer where a flicker event occurs, each layer is equipped with at least one pixel 18a, which can detect only the flicker light from that layer. This is a viable solution if the detector is used as a pure PET scanner. However, this functionality is more difficult to achieve with a Compton camera because the signal read from one strip is essentially the sum of two flicker events across two layers.
[0176] Configuration 3 - Shared detector strip + isolated Compton layer
[0177] In the third configuration ( Figure 11d , Figure 11e In ) one of the layers ( Figure 11d 16s in Figure 11e 16a) is optically, electrically, or electro-optically separated from other layers. For Compton camera functionality, this separating layer serves as the absorber portion ( Figure 11e 16a) or the scatterer section ( Figure 11d The specific value (16s) depends on the main direction of the incident high-energy photons. This layer is used in conjunction with a scintillator plate stack, employing the photon sensor described in configuration 2.
[0178] Configuration 4 - Split Block
[0179] In the fourth configuration ( Figures 12a to 12c In this configuration, the detection modules of the scintillator plate are stacked and spatially separated into two parts or blocks (15a, 15s). The arrangement of each module is shown in Configuration 2. Both modules can be used as PET detectors. For the Compton camera, one block 15s will be used as the "scatterer section", while the other 15a will be used as the "absorber section".
[0180] Configuration 5-Compton pixels
[0181] In the fifth configuration, which is primarily used for high-rate, transient gamma rays ( Figure 13bIn addition to any of the aforementioned configurations, a single photon sensor 18p (pixel) optically coupled to a scintillating crystal 16p is used as a Compton absorber (or Compton scatterer). This pixel should be small enough to achieve the desired spatial / angular resolution. The advantage of this configuration is that only a small number of additional channels are required to achieve the Compton function. In various variations, multiple individual pixels can be added as needed. It should also be understood that this configuration can be “reversed,” i.e., a single pixel (18p, 16p) can be used as a scattering layer, and a stack of scintillators can be used as an absorber. In either case, the readout chain can be as follows: if the invention is practiced in Compton mode, only the event that triggers the “Compton pixel” is further processed, and all other events are discarded.
[0182] Configuration of 6-Compton pixels + Compton layer
[0183] If the rate of the transient gamma rays is very high, configuration 5 may experience count rate saturation. To address this, configuration 3 can be modified so that the electrically isolated Compton layer is coupled to a scintillator that is potentially thinner than the PET scintillator. The reduced thickness decreases the probability of interaction, which lowers the overall count rate. Furthermore, the PET scintillator will absorb some of the transient gamma rays, which may further reduce the overall count rate. Figure 13a ).
[0184] Compton Cameras Between Modules
[0185] The Compton camera's functionality can also be achieved through inter-module scattering. One detection module 14a acts as a scatterer, and another detection module 14b (e.g., an adjacent module) acts as an absorber. Figure 14a , Figure 14b In this configuration, a separate layer is not required for identifying Compton scattering events, and the hardware changes compared to a pure PET scanner are minimal. For example, Compton scattering events can be identified by total energy discrimination and inter-module coincidence timing. Forward scattering dominates at the gamma energies of interest (several MeVs) primarily used for proton-range verification, and introducing spatial gaps 17 between two or more layers 16 would be advantageous to improve angular resolution and increase the probability of inter-module Compton scattering.
[0186] Traditional PET scans require circular components to ensure that most gamma rays enter the scintillation crystal element, which is substantially perpendicular to the crystal facet facing the light source. However, off-center emission can cause parallax. The interaction depth capability according to the invention alleviates this problem. Furthermore, the inventors have realized that it is possible to utilize components that are advantageous to non-circular shapes (e.g., by means of having such...) Figure 5aThe hexagonal components shown have enhanced interaction depth capabilities to improve the forward scattering probability between adjacent modules. The invention also enables the sensor plate to be positioned closer to the patient or near the object being scanned.
[0187] In another embodiment, such as Figure 5b As shown, the modules can be arranged in a radially staggered pattern, which also increases the probability of Compton scattering between modules. In this arrangement, it is less necessary to introduce radial gaps between the scintillator plates within each module.
[0188] In another embodiment, such as Figure 15a As shown, the detection modules 14s and 14a are arranged in a radial group. The inner group 14s, closest to the light source (target), will serve as the PET module and the diffuser module. The outer group 14a will serve as the absorber module for Compton imaging. Figure 15a An example of a 9:1 module arrangement is shown, in which an absorber detection module 14a is radially offset and centered on a 3x3 PET / scatterer detection module 14s. It should be understood that other arrangements and ratios between the scatterer / absorber modules can, of course, be used, such as 1:1, 9:4 (e.g., ...). Figure 15b (as shown in the image) etc.
[0189] Using essentially the same modules as scatterers and absorbers offers the advantage of simplifying the manufacture and readout of electronics, with these modules optionally being radially separated and optionally having different scintillator thicknesses.
[0190] Sensor board decommissioned
[0191] Read using cross lines ( Figures 19a to 19c ) or on floor 16 ( Figures 17a to 17e , Figure 18 The photon sensor strip 18b shared between the layers 16 can block scintillation light from one or more layers 16 to ensure that the column signal comes from only one layer. This can be achieved by a mechanical shutter around the edge of each layer. Alternatively, an electro-optical shutter 24 can be used. Figures 20a to 20c For example, a polarized liquid crystal or a transducer can be used to switch between a transmission state 24a and an absorption / reflection state 24b using a driving voltage. In the transmission state, the electro-optic shutter can also be used as a light diffuser between the scintillator 16 and the photon sensor 18.
[0192] In an advantageous embodiment, an alternative way to achieve similar functionality is to selectively enable or disable the photon sensor bias voltage of a single photon sensor or a group of photon sensors, for example, by... Figure 7The bias switching network in the diagram illustrates the sensor board grouping. Alternatively or additionally, the bias voltage can be adjustable, allowing the gain of an individual photon sensor, a group of photon sensors, a group of photon sensors optically coupled to a scintillator plate, or a group of sensor boards to be adjusted based on the expected primary gamma ray energy of interest.
[0193] Another alternative is to use a digital silicon photomultiplier tube as a photon sensor, which can enable / disable individual micro-units of the strip photon sensor in order to shield (ignore) light from a selected scintillator plate.
[0194] Read the chain
[0195] As an alternative to detector strips shared between layers, individual pixels along the sides of each scintillating block and multiplexed readout chains can be used.
[0196] Figure 19a An example of the first row / column readout (cross-line) of a PET scanner module is shown, with 5 square scintillator plates and 5 columns of photon sensors on each side. A total of 100 photon sensors can be read out through 25 channels. Channels Y1-Y5, corresponding to the sum of layers 1-5, provide information about the layer where the scintillation event occurred. For example, the readout electronics can be further simplified by using these channels as signal amplitude threshold triggers to evaluate whether a scintillation event has occurred in each layer. The PET scan data processing circuitry then further processes only events that involve exactly one layer, ensuring that the columns and signals X1,…,X… N It only corresponds to light emission within one layer.
[0197] Figure 19b and Figure 19c An example of the second row / column readout (cross-line) of the combined PET scanner and Compton camera module is shown, with four square scintillator plates and five columns of photon sensors on each side. A total of 80 photon sensors can be read out through 24 channels. Figure 19b Channels Y1 to Y4 provide information about the layer where the flickering event occurred and the energy (total amount of light). Energy suppression / filtering can be implemented using a discriminator. Figure 19b In this embodiment, a layer 16a is spatially separated from the other layers to serve as a Compton absorption layer. Channels Z1 to Z2 of the photon sensor are read out from this layer. 20 The Compton camera's data processing circuitry only handles events that involve exactly two layers, one of which is the absorption layer. In this example, 44 channels are required.
[0198] Other multiplexing schemes can be used, such as symmetric charge splitting.
[0199] Another possibility is to use the aggregated amount of information from the edges of the scintillator plate. One example is to use the centroid and the (charge) sum of each edge. If reducing the number of readout channels and / or digitizers is a high priority, this can be achieved, for example, before digitization via a resistor network / ASIC, as the sum of pixels along each edge and a weighted sum. Then, the output based on the edge i of the N photon sensor columns is reduced to two amounts for each scintillator edge.
[0200] {X1,…X N} i →S tot,i ,S weighted,i
[0201] (sum)
[0202] (Weighted sum)
[0203] To implement the centroid algorithm, the weighting coefficients can be evenly distributed (assuming all photon sensors have the same length along the edge), for example, λ. j =j.
[0204]
[0205] Then, based on centroid edge measurements, for example using cogs from a collimated gamma-ray source... i A calibration table composed of measured values and known flicker event coordinates allows for the reconstruction of the original coordinates of the flicker event. The total energy of the event is determined by the S-axis at all edges. tot,i The sum is obtained.
[0206] It can be noted that alternatively or additionally, other aggregation quantities besides the centroid can be used, such as the strip index with the maximum count, truncated centroid (discarding strips with a small number of counts), full width at half maximum, skew, or more complex functions.
[0207] Techniques for reconstructing events using aggregated edge quantities can be implemented in an analog manner (before digitization) or after digitization using resistive charge division circuits (CDC), where the purpose of digitization is to accelerate image reconstruction methods by reducing the dimensionality of the dataset for each flicker event.
[0208] Layer recognition based on time difference of dual-ended strip detector
[0209] Another method for layer identification is to read out the strip photonic sensors 18 located at both ends (41a, 41b), such as... Figure 18As shown. This dual readout of strip silicon photomultiplier tubes is known in itself and is described in Doroud2017
[11] (using differential readout for noise suppression), where the signal propagation speed on the strip is reported to be approximately v. prop ≈1E7m / s. The time difference between the arrival times of the pulse at both ends (i.e., the upper end (subscript u) and the lower end (subscript l)) is:
[0210]
[0211] Here, z is the distance between the scintillator layer and the intermediate layer (i.e., zero if the intermediate layer scintillates). For example, with a scintillator plate thickness of 3 mm, the ΔT difference between two adjacent layers is approximately 200 picoseconds (ps), which can be measured using state-of-the-art readout techniques. Furthermore, since light from any scintillator event will be distributed across multiple stripes along the edge, layer resolution can be improved by considering the time differences of several stripes used to identify the layer (e.g., by averaging the time differences of all stripes or all stripes with signal amplitudes above a threshold). The advantage of this method is that the layer identification photon sensor 18a can be replaced with a strip detector 18b, thus significantly reducing the total number of channels. The number of layers, and therefore the depth resolution of the interaction, is primarily limited by timing accuracy.
[0212] System components
[0213] Figure 16a A dual-head assembly is shown in a proton therapy environment. Figure 16b A three-head assembly is shown. In a two-head assembly, the ion beam 1 enters the target along the y-axis, and two detector assemblies (13a, 13b) are symmetrically arranged around the target 4 along the ±z-axis. The two assemblies (13a, 13b) will intercept a portion of the gamma rays emitted along the proton beam path: the transient gamma ray 21b and the positron annihilation gamma ray 21a. In a three-head assembly, an additional detector assembly 13c is arranged around the target, which is substantially oriented towards the proton beam emitter.
[0214] Layer recognition pixel - configuration
[0215] In a configuration where all sensors except one column of photon sensors are shared between two or more layers 16, a sufficient amount of light should reach the pixel 18a of the layer used to identify the flickering event. For example... Figures 17a to 17e As shown, these pixels 18a can be located at a corner, the center of an edge, or somewhere in between.
[0216] Figure 21bA contour plot showing the number of photons arriving at a pixel (as a function of the lateral coordinate of a flicker event) is presented, using relatively conservative assumptions about the crystal's light yield (30,000 photons / Melectron volt, gamma energy = 511 keV) and photon sensor parameters (photon detection efficiency = 0.2, dark count rate = 130,000 Hz / mm², excess noise factor = 0.2, signal integration time = 250 ns). In these plots, it is assumed that each edge is equipped with 5 photon sensors (20 in total), as... Figure 21a As shown in the figures, these graphs illustrate the light received by a single 10mm wide photon sensor located at the right edge (x = 25) from bottom to top (y = -25 to y = +25). For photon sensors located at the other edges, the situation is rotationally symmetric. For events originating near the right edge (x ~ 25mm, y > -15mm) and in the upper right quadrant, the photon sensor near the lower corner receives very little light. For all events originating on the left side (x <= 0), the photon sensor located at the center of the edge receives approximately 50 or more photons, but for events originating in the upper left corner, the sensor receives very few photons. Note that >50 photons is the average for simulated events. Events will occur where more or fewer photons are detected. However, even considering statistical fluctuations, this level is perfectly adequate to produce a signal well above the noise floor and the estimated dark count rate.
[0217] Therefore, to reliably determine which layer the event occurred on, it is only necessary to place the two layer identification photon sensors at the center of two opposite edges, such as... Figure 17e As shown. By considering the sum of two pixels, or only the pixel with the largest signal amplitude or integral, the layer of a flickering event can be reliably identified. Figure 17e As shown, in this case, a 4-layer square detection module (with 5 detector columns on two opposite edges) only requires 26 channels.
[0218] Detector geometry and pixel configuration
[0219] The geometry of the monolithic detection module, including its shape, area, and photon sensor configuration, impacts performance. Therefore, different polygonal shapes for the front and back sides, and varying numbers of photon sensors at each edge, were simulated. For each simulation scenario, a large number of events were simulated: isotropic photon emission from a randomly selected point (x0, y0, z0) within the scintillation block, ray tracing in the crystal and diffuser materials, and the response (number of detected photons) of each photon sensor. As mentioned above, conservative manufacturer-manufactured dark count rates, excess noise, and other figures were used to estimate the photon sensor responses, taking into account statistical fluctuations.
[0220] The detector mean and standard variations, varying with the transverse coordinates x0 and y0 within the scintillator, are estimated using photon sensor responses from a large set of events. This is either a training set or a calibration set. Depth z0 is not part of the calibration set. Then, another set of events is used as an evaluation set, where the extent to which the backtracking algorithm can predict the original transverse coordinates x0, y0 of scintillator events based on the training set is investigated. The predicted or fitted coordinates are denoted as x... fit ,y fit Then the lateral error ε is calculated as the Euclidean distance.
[0221]
[0222] For the entire evaluation set, the average error can then be calculated.
[0223] Average error itself is not a useful metric for measuring whether a monolithic edge detector configuration is superior to a conventional PET scanner configuration. To reduce the average error of a conventional PET scanner, one can simply reduce the size of the scintillator crystal and photon sensor, and increase the number of scintillators and detectors.
[0224] As a comparative metric, the inventors chose to compare the number of photon sensors according to the present invention with the number of photon sensors in a "conventional PET scanner" having a similar total lateral area and similar spatial resolution (same average error). A conventional PET scanner is defined as follows:
[0225] • A single scintillation crystal (bar), square in shape and frontally coupled to a single photon sensor in a 1:1 ratio.
[0226] To estimate the average error of the traditional configuration, simulations were performed, thereby obtaining... Figure 22 The error distribution is shown. Since a single-pixel crystal does not generate any information about the location of a flickering event within the crystal, the lateral coordinate of each event is assigned to the crystal center. Events originating from crystal corners exhibit the largest errors. The average error is approximately equal to:
[0227]
[0228] Where 's' represents the side of the crystal element facing the light source. The crystal size can be obtained by setting the average error according to various embodiments of the present invention:
[0229]
[0230] This allows us to calculate the total number of channels for a given total horizontal area.
[0231] A typical example of a conventional PET scanner configuration could be a scintillation crystal measuring 3.1mm x 3.1mm x 20mm (3.1mm in the X and Y directions, 20mm in the Z direction) coupled to a SiPM array of SiPM pixels measuring 3.3mm x 3.3mm. The fill factor of this configuration would be approximately 88%, regardless of the gaps between modules.
[0232] A monolithic detector with a lateral surface area of 2500 mm² was evaluated. 2 It is surrounded by a 4mm non-scintillation frame gap to account for space for the light diffuser and photon sensor. The fill factor of this detector is:
[0233] • Triangle: 72% Figure 9c )
[0234] • Square: 74% Figure 9a )
[0235] • Hexagonal: 76% Figure 9b )
[0236] Overall, when considering the non-flicker gaps between modules, we believe that the fill factor of a conventional PET scanner is similar to that of the embodiments of the present invention.
[0237] Figure 24 An example of the real and reconstructed interaction locations in a 50x50mm scintillator plate generated from simulation is shown. Figure 25 The average lateral (axial-azimuth) spatial reconstruction error on the scintillator plate is shown, while Figure 26 This is a histogram of the average lateral reconstruction position error. In this example, an average lateral error of 0.93 mm was achieved.
[0238] The embodiments of the present invention can increase the effective thickness of the scintillation material by adding more layers without significantly reducing image quality. Therefore, compared with conventional PET scanners, the lower fill factor can be compensated for by increasing the overlap probability of more scintillation material.
[0239] The optimization metric for each embodiment of the present invention can be defined as the ratio of the number of channels in a conventional PET scanner to the number of channels in a single layer of the device:
[0240]
[0241] Figure 23The ratio R of triangular, square, and hexagonal scintillator plates versus the number of photon sensors per edge is shown. Square detectors are likely the easiest to fabricate, and the optimal number of pixels per side is 5, although similar gains can be achieved with 4 or 6 photon sensors per side. If we consider a conventional PET scanner, where an array of M×M detectors is coupled to an array of N×N crystals (where N>M), and a monolithic light guide is inserted between the crystals and photon sensors, the ratio R will be even lower. A 4×4 detector array coupled to a 5×5 crystal array will reduce R by 36%. Clearly, increasing the number of layers will further decrease the ratio R. However, this metric does not account for DOI gain. R is effective if DOI information is not needed (e.g., the thickness of the scintillator plate is the same as the height of a single crystal in a conventional PET scanner).
[0242] Typically, for triangular, square, or hexagonal PET scanners, the number of channels can be reduced by 30 to 40 times compared to conventional 1:1 coupled PET scanners. This is a significant improvement, substantially reducing the cost of PET scanners. Alternatively, the gain from reducing the number of channels can be used to increase (especially) the axial field of view (FOV) of the PET scanner. This is particularly advantageous for whole-body PET scanner applications, where the axial FOV can span the entire patient's body.
[0243] Compton camera scintillator plate thickness optimization
[0244] The probability of a valid Compton scattering event (Compton scattering in one scintillator plate and photoelectric absorption in the other) typically depends on the thickness of the scintillator plates. This probability was studied in detail for a two-layer configuration with different scintillator plate thicknesses for both the scatterer and absorber, including varying primary gamma-ray energies and the total radial thickness of the scintillators, such as... Figure 34 This is a generalization. For lower energies (0.511 keV), the optimal ratio is in the range of 30%–50%, depending on the total scintillator thickness (6–20 mm). However, for higher energies, the optimal ratio is closer to 50%, i.e., the thicknesses of the scatterer and absorber are equal. Depending on the variant and the energy of interest, the scattering layer thickness is preferably between 20% and 60% of the total thickness.
[0245] Calibration - Reference Table
[0246] The calibration of the testing system can advantageously include the following steps:
[0247] Given the horizontal coordinate (x) cal ,y cal The detector is illuminated by a collimated light source at point z. Precise knowledge of the interaction depth z is not required. cal .
[0248] Record enough events for each detector layer and calibration location.
[0249] Calculate the average and standard deviation μ of the response of each photon sensor i at a given location. i (x,y) and σ i (x,y).
[0250] A reference table can be generated based on fine grid interpolation of the average and standard deviation of each photon sensor at the calibration location and optionally at any intermediate location not part of the calibration process.
[0251] μ as a single photon sensor i and σ i As a substitute or supplement, it can also be used to calculate any amount of aggregation (e.g., center of gravity).
[0252] Event Reconstruction
[0253] To reconstruct the interaction coordinates of a single event, the following methods can be used:
[0254] Digitize the response and / or aggregation of the photon sensor;
[0255] Find the position in the calibration table that best matches the response (this can be done efficiently using various methods known in themselves).
[0256] Event Rejection
[0257] In both PET scan mode and Compton camera mode, rejecting events based on energy deposition is advantageous. Furthermore, rejecting two-scattering or multi-scattering events within the same scintillator is also advantageous. One approach is to compare the event signature (the signal for each readout channel) with its nearest match in a reference table. "Nearest match" means the normalized Euclidean distance (sum of squared differences divided by the standard deviation of the pixel response or aggregation amount) between the event and the nearest reference match. If this difference is greater than a configurable threshold, the event can be rejected.
[0258] PET scan mode and Compton camera mode
[0259] When the detector assembly is operating in PET scan mode, only events triggering on the two sensor plates opposite the light source (i.e., along the LOR) should be considered. In proton therapy environments, due to the high potential trigger rate of transient gamma rays during proton delivery, PET scan mode can be selectively disabled entirely during ion beam delivery. Since the imaging volume of interest (defined by the processing volume and proton beam path) is explicitly defined (within the uncertainty margin), an effective detection module overlap set can be defined to discard any overlapping events outside the imaging volume of interest. Furthermore, one or more energy windows can be defined to reject events that do not correspond to the gamma energy of interest.
[0260] For example, in Compton camera mode, events triggered by two layers within a single module (Compton camera within a module) or by a single layer in two adjacent modules (Compton cameras within a module) can be accepted. Multiple energy suppression windows (where energy is the sum of signals from the two triggered layers) can be defined to accept only events corresponding to known transient gamma emission peaks. To further limit the data rate, in cases involving the proton beam range within the primary target of interest, only the module with the highest image reconstruction resolution in the proton beam direction can be enabled.
[0261] Multiplexed readout of azimuth-axially arranged sensor board
[0262] In previously disclosed arrangements, radially stacked blocks are read out via multiplexing or strip sensors. A major drawback of these arrangements is their inefficiency in detecting forward Compton-scattered gamma rays. Forward-scattered gamma rays may interact within a single sensor plate (if no scattered gamma rays are detected) or between two sensor plates that are typically radially aligned. In most cases, it is impossible to reconstruct the two interaction locations separately. The two interaction locations can only be reconstructed when Compton scattering causes the two interactions to occur in two independent modules (“inter-module scattering,” low probability), or when additional scattering / absorbing sensor plates or Compton pixels with independent photon sensors are added (increasing complexity). Specifically, at energies used to detect transient gamma rays (up to 7 MeV), small-angle forward scattering dominates.
[0263] To overcome the inherent drawbacks of combined readout in radially stacked multilayer configurations, a novel multiplexed readout scheme is proposed. The inventors have realized that arranging the common readout sensor board in a radial plane (rather than radially stacked), instead of combining information from compactly arranged, radially stacked sensor boards into a common readout channel (with the drawbacks discussed previously), yields significant functional and performance advantages. Several drawbacks can be overcome by arranging the common readout scintillator board in an azimuth-axial configuration instead of a radial one. An azimuth-axially arranged sensor board is shown below. Figure 28As shown.
[0264] The term "radial-azimuth-axial" in this article refers to the common cylindrical arrangement of the scintillator elements in a PET scanner. Alternatively, other arrangements not directly applicable to the term "cylindrical coordinate system" may be used, such as spherical, "box-shaped," or dual-headed arrangements. Figure 5e (e.g., four-headed, four-headed (Fig. 5f), or helmet-shaped (e.g., for dedicated brain PET scanners). However, they all have one thing in common: the scintillator elements are arranged in a certain way around the volume of interest.
[0265] Figure 29 An example of a multiplexed readout configuration for a 2x2 sensor board in an azimuth-axis configuration is shown. Each sensor board has multiple photon sensors, such as eight photon sensors 18. Depending on the configuration, the multiplexing circuit 33 can perform analog summations of the connected photon sensors 18. This allows for the individual readout of the summation for each sensor board (S1-S4), as well as the summation of all photon sensors located at specific positions (“top right,” “top left,” etc.) on the sensor board (E1-E8). The summation circuits S1-S4 can identify the scintillator board 18 where a flickering event occurred, while E1-E8 can provide spatial information about the flickering event.
[0266] By multiplexing the photonic sensor signals from azimuth-axis arranged blocks, all the benefits of reducing the number of readout channels are retained. Advantageously, this scheme also allows for independent readout of radially separated layers. The azimuth-axis multiplexing device is particularly well-suited for resolving 511 keV Compton-scattered gamma rays generated by electron-positron annihilation. If these layers are thin enough, a single 511 keV gamma ray is unlikely to be simultaneously Compton-scattered and absorbed on the same sensor plate. This advantage is particularly important because it allows for the acceptance of a larger portion of the incident 511 keV gamma rays. In conventional PET scanners, events with a deposition energy of 511 keV (within the instrument's energy acceptance window) are typically accepted to reject Compton-scattered gamma rays, and events where 511 keV gamma rays are directly photoelectrically absorbed are also accepted. One reason for this is that conventional pixelated PET scanners lack interaction depth resolution and cannot definitively determine the initial interaction location. However, in this invention, both scattering and absorption coordinates can be determined simultaneously. Based on the gamma ray energy of interest, Compton's kinematic rules can be applied to distinguish between scintillator plates that scatter and those that absorb. The time series can also be determined using timestamps. If the time series cannot be definitively determined, image reconstruction can be performed using probability-weighted LOR (PET) or Compton cone imaging (Compton camera imaging). Furthermore, prior knowledge about the spatial origin of the main gamma ray can be used to accept or reject LOR or Compton cone imaging.
[0267] Typically, the ability to resolve Compton-scattered gamma rays during PET scanning operations significantly improves the overall sensitivity of the detector: a small fraction of the 511 keV gamma rays detected simultaneously.
[0268] For example, consider the following configuration where the total radial thickness of the scintillator is 20 mm (e.g., LYSO), divided into five azimuth-axially stacked modules, each 4 mm thick.
[0269] The probability that the 511keV gamma rays about to appear on the device will interact in some way is about 80%. However, the probability of it being directly absorbed by photoelectric light is only 26%. Therefore, the probability that two overlapping, antiparallel gamma rays will be simultaneously absorbed by photoelectric light is only 0.26 * 0.26 = 6.7%. This fundamentally limits the sensitivity of traditional PET scanners.
[0270] To compare with the azimuth-axis multiplexed sensor board proposed by the inventors, for example, assuming that events E1 > 100 keV and E2 > 100 keV can be clearly distinguished. Besides direct photoelectric absorption events, the probability that 511 keV gamma rays are first Compton scattered and then photoelectrically absorbed in different radial layers is 11.4%. Therefore, the probability of detecting 511 keV gamma rays by direct photoelectric absorption or as a two-stage Compton event (scattering + absorption) is 26% + 11.4% = 37.4%. The probability of coincident detection is 0.374^2 = 14%.
[0271] Therefore, embodiments of the present invention can have approximately twice the overall sensitivity or effective coincidence detection rate of conventional detectors. Half the number of emitted gamma rays are required to form an image of equivalent quality. Alternatively, in a radiopharmaceutical environment, the injected tracer isotopes can be significantly reduced to minimize patient exposure.
[0272] Combined signal readout from adjacent photon sensors at the edge
[0273] Common photon sensors, such as silicon photomultiplier tubes, are typically square, for example, 1x1mm, 3x3mm, 4x4mm, or 6x6mm. The precise dimensions of the scintillator plate proposed in this invention can be adapted to these dimensions. In one variation, for example, the desired scintillator plate size could be 48x48x3mm (3mm being the radial thickness). To match the radial thickness, a 3x3mm photon sensor would be suitable. In one variation of the invention, the number of photon sensors per edge should be at least 2, 3, 4, 5, 6, 7, or 8. However, a 3x3mm photon sensor corresponds to 16 channels per edge. In an advantageous embodiment of the invention, signals from adjacent photon sensors can be added electronically, such that adjacent groups of photon sensors are read out as a whole before digitization, or as a whole readout connected to a multiplexing circuit. Figure 40 An example is shown where the number of channels at each edge is reduced by 4:1 using the summation of the common anode current. Other signal calculation techniques are known to address issues such as increased effective sensor capacitance and can be used in various embodiments of the invention.
[0274] Axial field of view extension
[0275] In one advantageous embodiment, at least two orientationally opposed sensor plate arrangements can rotate around the patient or scanned subject to acquire local or whole-body PET and / or SPECT imaging, such as Figure 38 As shown. This detector can be used in conjunction with CT imaging equipment. It can also be used with MRI. This detector can acquire dynamic images ("4D" images containing volumetric and temporal information).
[0276] The primary limitation of most conventional PET scanners is their limited axial field of view (FOV). Manufacturing costs typically increase linearly with the expansion of the FOV, as more photon sensors, readout channels, and scintillator material must be added. Costs can be partially reduced by decreasing the radial thickness of the scintillator, but this does not reduce the scintillator area that must be covered by the photon sensors. For example, consider line sources with a similar axial range to conventional PET scanners, such as… Figure 31 This is conceptually illustrated. To simultaneously detect annihilated gamma rays emitted near the end of a line source, the scanner or object being scanned must be moved axially, thus extending the imaging time and making dynamic imaging along the line source challenging. However, the axial FOV of this invention can be increased by simply rearranging the sensor plates in the axial direction, such as... Figure 30As shown, this does not increase cost or increases it very little. When the total radial thickness of the scintillator decreases, the coincidence probability decreases, which is compensated for by increasing the field of view (FOV). Furthermore, by employing an azimuth-axial multiplexing configuration, Compton scattering events can be received, thus significantly improving overall sensitivity. A comparison of axial sensitivities is shown below. Figure 32 As shown.
[0277] like Figure 36 As shown, changing the sensor plate configuration from a single radial layer to two or more radial layers can effectively increase the probability of coincidence detection, even with a reduction in the total thickness of the scintillator. For example, the effective coincidence absorption probability of a single-layer configuration with a scintillator radial thickness of 20 mm is similar to that of a two-layer configuration with a total thickness of 15 mm or a four- to six-layer configuration with a total thickness of 10 mm. Therefore, the advantage of this invention is that it can significantly reduce the overall volume of the scintillator and the total area that the photon sensor needs to cover, thereby reducing the overall cost and / or increasing the axial field of view (FOV).
[0278] For example, the radial thickness of the scintillator can be less than 40 mm, 30 mm, 20 mm, 15 mm or 10 mm, and it is distributed on at least two scintillator plates.
[0279] In front-coupled detection modules, considering the probability of coincidence detection, a scintillator radial thickness of approximately 20 mm is generally considered the optimal choice for cost-effective PET scanning. In an advantageous embodiment of the invention, since Compton scattered gamma rays are also acceptable as valid events, the scintillator radial thickness can be reduced, for example, to less than 19 mm.
[0280] In an advantageous embodiment, the azimuth-axial multiplexing configuration can also be adjusted to include multiple radial clearances, such as... Figure 33 As shown, a configuration with 2+2+2 radially stacked scintillator blocks multiplexed in a 3x3 configuration is illustrated. This configuration can function as a PET scanner, a Level 2 Compton camera, and a Level 3 Compton camera. The radial air gap between the sensor plates also improves heat dissipation.
[0281] For example, multiplexed sensor boards can be arranged in an azimuth-axis array in 1x2, 1x3, 1x4, 2x2, 2x3, 2x4, 3x3, 3x4, 4x4, 4x5, or 5x5 configurations.
[0282] Functional combination and uses
[0283] Features of different embodiments are interchangeable between embodiments and can be combined in different ways, unless otherwise specifically stated. Although numerous specific details are set forth in the following description to provide a more thorough understanding of the invention, it will be apparent to those skilled in the art that the invention can be practiced without these specific details. The basic and conventional techniques in electronic devices, sensor systems, image analysis, signal processing, data communication systems, image acquisition systems, and other components that implement the invention will be readily understood by those skilled in the art; therefore, for the sake of brevity, further explanations and details will be omitted in this specification.
[0284] The detection system according to embodiments of the present invention can also be used for other types of nuclear imaging, such as imaging of 3γ emission, or imaging of isotopes emitting one positron (producing a coincident, opposing 511 keV γ-s, which, when detected, generates an LOR 27 along the location of the light source) and another γ (producing a Compton cone 25 at the location of the light source). By combining information from the LOR and the Compton cone, triangulation of possible isotope positions can be performed with very high precision, especially in ion beam therapy environments where the volume of interest is well-known, such as... Figure 27 As shown.
[0285] WO 2018 / 081404 A1 discloses a radially stacked edge detection detector, with each layer having a separate photon sensor. Compared to a 1:1 master-plane coupled PET scanner, using a single layer significantly reduces the number of channels. However, the number of channels increases linearly with the number of layers. According to one aspect of the invention, by employing a combination of strip photon sensors spanning multiple layers and pixel photon sensors coupled to a single layer, even the number of channels in a multi-layer detector is significantly reduced, while enabling the identification of which layer a flicker event occurs in via a pixel detector.
[0286] According to another aspect of the invention, a dual-ended strip photonic sensor is used, thereby utilizing the time difference between the two ends to distinguish which layer the scintillation event occurred in. Therefore, the total number of readouts does not depend on the number of layers in the module. The advantage of this solution is that it does not require layer-specific pixels. The interaction depth resolution is limited only by timing accuracy and the number of scintillator plates.
[0287] According to another aspect of the invention, the detection system uses adjacent modules as scatterer / absorber modules for a Compton camera (“inter-module Compton camera”). This configuration offers greater robustness to parallax caused by gamma rays entering the detector at an angle, due to the interaction depth resolution. Therefore, a non-circular arrangement (e.g., hexagonal, octagonal, or other polygonal) of the detection modules around the scanned object increases the probability of detecting forward-scattered gamma rays in two different modules compared to a conventional circular arrangement.
[0288] According to another aspect of the invention, an electro-optical shutter is provided for temporarily blocking optical signals from selected layers. Specifically, this is advantageous when reading out stacked detectors using strip detectors spanning multiple layers. For example, the electro-optical shutter can be used as a method to functionally convert some layers radially closest to the imaged object into "gamma filters." This is a useful feature when the instantaneous rate of transient gamma rays is so high that the detector saturates when all layers are optically activated. By temporarily blocking light from the layers closest to the target, these proximal layers will absorb a portion of the transient gamma rays without obscuring the detector. The overall count rate of the detector will then decrease.
[0289] According to another aspect of the invention, an arrangement of conceptually identical modules (each module comprising a scintillator plate, a photon sensor, and readout electronics) can be configured to function as a combined PET scanner and a Compton camera, wherein radially offset groups of modules serve as absorbers in a two-stage Compton camera. This configuration is advantageous from a manufacturing and cost perspective, and is also easily customizable.
[0290] According to another aspect of the invention, an azimuth-axial arrangement of a sensor board with multiplexed readout is provided. Specifically, this facilitates the detection of Compton-scattered gamma rays for PET scanning and Compton camera functions over a large solid angle around the volume of interest and reduces parallax.
[0291] Existing technology reference
[0292] [1]CN 107544086 A
[0293] [2]WO 2018 / 081404 A1
[0294] [3] K. Shimazoe et al., 2020, Nuclear Instruments and Methods in Physics Research, A: https: / / doi.org / 10.1016 / j.nima.2018.10.177
[0295] [4]EP1617237 A1
[0296] [5]US 2018 / 172847 A1
[0297] [6]US 2005 / 116173 A1
[0298] [7] Georgy Shakirin et al., 2011, Phys. Med. Biol. 56 1281 (Shakirin 2011) https: / / doi.org / 10.1088 / 0031-9155 / 56 / 5 / 004
[0299] [8] Rohling et al., 2017, Phys. Med. Biol., at press (Rohling 2017) https: / / doi.org / 10.1088 / 1361-6560 / aa6068
[0300] [9] Antje-Christin Knopf and Antony Lomax, 2013, Phys. Med. Biol. 58 R131 (Knopf 2013)
[0301]
[10] Jan et al., Med. Phys. 44(12), December 2017 (Jan 2017) (Medical Physics, Vol. 44 (No. 12), December 2017) https: / / doi.org / 10.1002 / mp.12626 )
[0302]
[11] K. Doroud, MCSWilliams, K. Yamamoto (Doroud 2017), The Strip Silicon Photo-Multiplier: An innovation for enhanced time and position measurement, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, Volume 853, 2017, Pages 1-8, ISSN 0168-9002
[0303] List of elements referenced in the attached diagram
[0304] Patient 5
[0305] Target area (e.g., tumor) 4
[0306] Ion beam therapy system 6 (e.g., proton beam therapy system)
[0307] Patient stent 7
[0308] Ion beam emitter 8
[0309] Ion beam 1
[0310] Scanning Magnet 2
[0311] Beam intensity and profile monitor 3
[0312] Gamma Ray Detection System 10
[0313] Compton Camera 11
[0314] PET scanner 12
[0315] Detection module components 13, 13a, 13b, 13c
[0316] Opening 42
[0317] Detection modules 14, 14a, 14s
[0318] Scatterer section / block 15s
[0319] Absorber section / block 15a
[0320] Sensor boards 16 and 18
[0321] Scintillator plate 16
[0322] Scattering layer 16s
[0323] Absorbing layer 16a
[0324] Main surface 40a
[0325] Lateral subsurface 40b (also known as the "edge")
[0326] Flashing Body Stick 16p
[0327] Radial clearance 17
[0328] Photon sensor 18
[0329] Single-layer photonic sensor 18a (also known as "photonic sensor pixel" or simply "pixel")
[0330] Photon sensor 18p coupled to 16p scintillator rod
[0331] Crossover connection layout 18c
[0332] The 18b striped multilayer photonic sensor (also known as a "striped photonic sensor" or simply a "striped detector")
[0333] Photon sensor bracket (plate) 20
[0334] Detector-Scintillator Optical Interface 22
[0335] Electro-optical shutter (EOS) 24
[0336] Edge light diffuser 26
[0337] Interlayer reflector 28
[0338] Light blocking layer / absorber 29
[0339] Low refractive index gap 31
[0340] Signal processing and control systems 30
[0341] Circuit board 32
[0342] Multiplexing circuit 33
[0343] Electronic components 34 (e.g., microprocessors, memory, FPGAs, etc.)
[0344] Connectors 36a, 36b
[0345] Gamma rays 21
[0346] Positron gamma rays 21a
[0347] Instantaneous gamma ray 21b
[0348] Light source 23
[0349] Compton Cone 25
[0350] Response Line (LOR) 27
[0351] Compton Cone - LOR Intersection 27b
[0352] Volume of interest (target region) 27c
[0353] scintillation rays 53
[0354] Main surface coupling detection module 50
[0355] Scintillator array 51
[0356] Photonic sensor array 52
Claims
1. A gamma-ray detection system (10), the gamma-ray detection system (10) comprising detection module assemblies (13, 13a, 13b, 13c), the detection module assemblies (13, 13a, 13b, 13c) comprising at least two detection modules (14, 14a, 14s), the at least two detection modules (14, 14a, 14s) being configured to perform positron emission tomography (PET) scanning on a target region (4); each detection module comprising: - Signal processing and control system (30), the signal processing and control system including circuit board (32) and electronic components (34) mounted on the circuit board. - A plurality of stacked monolithic scintillator plates (16), each monolithic scintillator plate (16) having a main surface (40a) oriented toward the target region and a lateral subsurface (40b) defining an edge of the scintillator plate, the surface area of the main surface being greater than the surface area of the lateral subsurface; and a plurality of photon sensors (18), mounted on each of the edges, configured to detect and determine the location of a scintillator event in the scintillator plate in the plane of the main surface from gamma rays incident on the main surface; wherein the gamma ray detection system is also configured to function as a Compton camera, at least one scintillator plate not closest to the target region is configured as an absorber scintillator plate of the Compton camera, and wherein the circuit board (32) is mounted at the radially outermost end of the detection module, and the photon sensors (18) are disposed on a photon sensor support plate (20), the photon sensor support plate including an edge connector (36b) for connecting to a connector (36a) of the circuit board (32) of the signal processing and control system (30).
2. The gamma ray detection system according to claim 1, wherein, The photon sensor support plate (20) is configured as a silicon photomultiplier tube array plate.
3. The gamma ray detection system according to claim 1, wherein, Multiple photon sensors of at least two radially stacked scintillator plates are connected to a processing circuit configured to multiplex the readouts of the multiple photon sensors.
4. The gamma ray detection system according to claim 1, wherein, Multiple photon sensors of at least two azimuth-axially arranged scintillator plates are connected to a processing circuit configured to multiplex the readouts of the multiple photon sensors.
5. The gamma ray detection system according to claim 1, wherein the gamma ray detection system includes at least one radial gap (17), the at least one radial gap (17) being located between at least two scintillator plates in the plurality of stacked scintillator plates or between the at least two detection modules, wherein, The height of the radial gap H The thickness of one of the plurality of scintillator plates T The relationship between 200 > H / T > Within 2 range.
6. The gamma ray detection system according to claim 1, wherein, The plurality of photonic sensors include at least one strip multilayer photonic sensor (18b) extending along the edges of the plurality of layers.
7. The gamma ray detection system according to claim 6, wherein, The gamma ray detection system includes multiple strip multilayer photon sensors located on each edge side of the multiple stacked scintillator plates.
8. The gamma ray detection system according to claim 6, wherein, The at least one strip multilayer photonic sensor is a dual-ended strip detector configured to measure the arrival time of signals at both ends.
9. The gamma ray detection system according to claim 1, wherein, The plurality of photon sensors include at least one monolayer photon sensor (18a), which is located on at least one edge of each scintillator plate.
10. The gamma ray detection system according to claim 9, wherein, The single-layer photonic sensors in a row or column are interconnected by cross-line connections (18c) such that the readout is the sum and / or weighted sum of the signals from the multiple interconnected single-layer photonic sensors.
11. The gamma ray detection system according to claim 1, wherein the gamma ray detection system further comprises a light reflection interface layer (28) and a light absorption interface layer (29) located between at least two scintillator plates in the scintillator plate.
12. The gamma ray detection system according to claim 1, wherein the gamma ray detection system further comprises a low refractive index gap located between at least two scintillator plates in the scintillator plates.
13. The gamma ray detection system according to claim 1, wherein the gamma ray detection system further comprises an electro-optic shutter (24) located between the edge of at least one scintillator plate and the photon sensor.
14. The gamma ray detection system according to claim 13, wherein, The electro-optic shutter includes a light diffuser material and thickness, which is configured to diffuse light in a flashing event near the edge.
15. The gamma ray detection system according to claim 1, wherein, Surface area of the main surface of the scintillator plate S and the thickness of the scintillator plate T Within the following range: 100 mm 2 <= S <= 40000 mm 2 , 0.5 mm <= T <= 30 mm.
16. The gamma ray detection system according to claim 1, wherein, The detection module assembly surrounds the target area and includes a gap (42) or hole for emitting an ion beam through it.
17. The gamma ray detection system according to claim 5, wherein, The radial clearance satisfies the following relationship: H / (T1+ T2) > 5, of which, T1 and T2 It is the thickness of the two scintillators surrounding the radial gap. H It is the height of the radial clearance.
18. The gamma ray detection system according to claim 1, wherein, The total radial thickness of the multiple stacked monolithic scintillator plates is less than 19 mm.
19. The gamma ray detection system of claim 1, wherein the gamma ray detection system comprises two radially stacked scintillator plates, wherein the ratio between the thickness of the inner radial scintillator plate and the total radial thickness of the scintillator is in the range of 0.2-0.
6.
20. The gamma ray detection system according to claim 1, wherein, The photon sensor bias voltage of a single scintillator plate photon sensor can be adjusted or enabled / disabled independently.
21. The gamma ray detection system according to claim 1, wherein, Photon sensors coupled to at least two radially stacked scintillator plates are connected to a processing circuit configured to apply Compton kinematics rules to determine whether two overlapping block events correspond to Compton scattering of forward or backward scattering following absorption.
22. The gamma ray detection system according to claim 21, wherein, The processing circuitry is configured to reject events originating from the main gamma rays entering the detector in a radially outward direction.
23. The gamma ray detection system according to claim 21, wherein, The processing circuit is configured to use the interaction coordinates of photoelectric absorption as the LOR endpoint of small-angle forward Compton scattering gamma rays originating from electron-positron annihilation.
24. The gamma ray detection system according to any one of claims 21 to 23, wherein, The processing circuitry is configured to discard Compton scattering events that exceed the configurable primary gamma-ray energy-dependent scattering angle to improve angular resolution.
25. The gamma ray detection system according to claim 1, wherein, An analog signal from a neighboring photon sensor is added before digitization or other multiplexing circuitry.
26. The gamma ray detection system according to claim 5, wherein the height of the radial gap... H The thickness of one of the plurality of scintillator plates T The relationship between 50 and 50 is as follows: H / T > Within 10.
27. The gamma ray detection system according to claim 9, wherein, The plurality of photon sensors include at least one monolayer photon sensor (18a), which is located on at least two edges of each scintillator plate.
28. An ion beam therapy system (6) for irradiating a tissue region with an ion beam, the ion beam therapy system (6) comprising: Patient stent (7); An ion beam emitter (8) that can move relative to at least one axis of rotation; Furthermore, the gamma-ray detection system according to claim 1 is configured to perform transient gamma-ray detection and PET scanning during, between, and after ion beam irradiation.