Modular variable resolution full-ring spect imaging system and method

Abstract

The application relates to the technical field of medical imaging equipment, and discloses a modular variable-resolution full-ring SPECT imaging system and an imaging method thereof. The system comprises a ring-shaped gantry and a plurality of detector modules. The ring-shaped gantry is uniformly provided with a plurality of modular mounting interfaces along the inner circumference, and the plurality of detector modules are pluggably mounted on the mounting interfaces. The detector modules are divided into a basic detector module group and an extended detector module group. The basic detector module group comprises at least one basic detector module, and the extended detector module group comprises at least one extended detector module. The basic detector module has a first imaging characteristic, and the extended detector module has a second imaging characteristic superior to the first imaging characteristic. Through a modular expandable architecture and parallel acquisition fusion reconstruction, the application realizes low-cost upgrading and flexible configuration, and simultaneously obtains a global large field of view and a target area high-resolution image in a single scan, thereby improving imaging efficiency and diagnostic accuracy.

Description

Technical Field

[0001] This invention relates to the field of medical imaging equipment technology, specifically to a modular variable resolution full-ring SPECT imaging system and imaging method. Background Technology

[0002] A full-ring SPECT system typically comprises a ring gantry with multiple detector modules evenly spaced along its inner circumference. Each detector module consists of a scintillation crystal, a photoelectric conversion device or semiconductor detector, readout electronics, and a front-end collimator. The collimator is used to filter the direction of incident gamma photons, and its aperture size, aspect ratio, and arrangement directly determine the system's spatial resolution and detection sensitivity.

[0003] Currently, clinical SPECT using a full-ring design employs 12 detector modules arranged at equal intervals along the circumference. These modules are equipped with parallel-aperture collimators, covering routine examinations of the whole body, thyroid gland, and lung perfusion. However, the spatial resolution of this approach is limited by the collimator aperture, failing to meet the sub-millimeter precision requirements for areas such as the striatum and myocardial perfusion defects. Furthermore, the hardware configuration of such systems is fixed; performance improvements often require replacing the entire system, resulting in high upgrade costs and poor flexibility. To improve resolution, another approach proposes altering imaging characteristics within a fixed-number detector ring architecture by mechanically switching different collimators (e.g., switching between multiple pinhole arrays). However, in a single imaging event, all detector modules still use the same type of collimator for acquisition, making parallel mixed acquisition of general-purpose and high-resolution collimators impossible. This forces the system to make trade-offs between wide field-of-view coverage and high-resolution imaging, making it difficult to achieve both simultaneously. Additionally, the added precision mechanical switching structure increases system complexity, failure rate, and maintenance costs; the resolution improvement is limited by the physical limits of the collimator's internal switching capabilities.

[0004] Therefore, existing SPECT systems generally suffer from fixed configurations and uneconomical upgrades, and cannot simultaneously meet the requirements of large-area scan coverage and high-resolution imaging of local target areas in a single imaging event, which limits the breadth and depth of their clinical applications. Summary of the Invention

[0005] To address the shortcomings of the prior art, this invention provides a modular variable resolution full-ring SPECT imaging system and imaging method.

[0006] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows: A modular variable resolution full-ring SPECT imaging system includes a ring gantry and multiple detector modules. The ring gantry has multiple modular mounting interfaces uniformly pre-set along its internal circumference, and the multiple detector modules are pluggably mounted on these mounting interfaces. The detector modules are divided into a basic detector module group and an extended detector module group. The basic detector module group includes at least one basic detector module, and the extended detector module group includes at least one extended detector module. The basic detector module has a first imaging characteristic, and the extended detector module has a second imaging characteristic superior to the first imaging characteristic.

[0007] As a preferred technical solution, each of the detector modules includes a detector and a collimator; the first imaging characteristic of the basic detector module is large field-of-view imaging, and the collimator is a general-purpose collimator; the second imaging characteristic of the extended detector module is high-resolution imaging, and the collimator is a high-resolution collimator.

[0008] As a preferred technical solution, the general-purpose collimator is a parallel-hole collimator, and the high-resolution collimator adopts one or more of the following: multi-sub-hole collimator, trapezoidal sub-hole collimator, or multi-pinhole collimator.

[0009] As a preferred technical solution, all detectors use cadmium zinc telluride semiconductor detectors.

[0010] As a preferred technical solution, the installation interface is uniformly preset with N ports along the inner circumference of the ring frame, where N is a natural number ≥ 2, and the number of basic detector modules is 1 to 5 times that of extended detector modules.

[0011] As a preferred technical solution, it also includes a data acquisition and control system and a fused image reconstruction subsystem. The data acquisition and control system is used to automatically identify the number and type of detector modules installed on the installation interface, and automatically load the corresponding acquisition protocol based on the identification results to control multiple detector modules to acquire data in parallel. The fused image reconstruction subsystem receives the parallel acquisition data from the detector modules and generates a fused tomographic image.

[0012] As a preferred technical solution, the fused image reconstruction subsystem includes a preprocessing module, a spatial mapping module, a fusion reconstruction module, and a post-processing output module. The preprocessing module is used to preprocess the global projection data collected by the basic detector module and the target area high-resolution projection data collected by the extended detector module. The spatial mapping module is used to uniformly map the high-resolution projection data to the global projection image coordinate system. The fusion reconstruction module is used to fuse the complete field of view information provided by the global projection and the high-density sampling information provided by the target area high-resolution projection within the same reconstruction framework to obtain a fused reconstructed image. The post-processing output module is used to post-process the fused reconstructed image and output a standard fused tomographic image.

[0013] A full-ring SPECT imaging method, using the aforementioned modular variable resolution full-ring SPECT imaging system, includes: in a single imaging event, controlling the basic detector module group and the extended detector module group to acquire data in parallel to obtain global projection data and target area high-resolution projection data respectively; performing fusion reconstruction processing on the global projection data and target area high-resolution projection data; the fusion reconstruction processing includes: using the global projection data as the basic reconstruction framework, mapping the target area high-resolution projection data to the corresponding spatial position of the global projection data, and generating a fused tomographic image that combines the integrity of the whole-body field of view with the local high resolution of the target organ.

[0014] As a preferred technical solution, the steps of the fusion reconstruction process include: Step 1: Channel Projection Preprocessing The global projection data collected by the basic detector module and the high-resolution projection data of the target area collected by the extended detector module are preprocessed respectively. Step 2: Pixel Space Mapping Map the high-resolution projection uniformly to the global projection image coordinate system; Step 3: Integration and Reconstruction Within the same reconstruction framework, the complete field-of-view information provided by global projection and the high-density sampling information provided by high-resolution projection of the target area are fused to obtain a fused reconstructed image; Step 4: Post-processing and output The fused and reconstructed images are post-processed and adjusted to output fused tomographic images conforming to the DICOM standard.

[0015] As a preferred technical solution, in step 3, the method for implementing the fusion reconstruction is one of the following: (1) Step-by-step fusion: The whole body image is obtained by standard reconstruction using global projection, and the local image is obtained by high-resolution iterative reconstruction using target area projection; the local high-resolution image is embedded into the whole body image according to the spatial coordinates, and boundary smoothing is performed. (2) Joint optimization: Construct a weighted reconstruction objective function, and simultaneously fit two sets of data: global projection and high-resolution projection of the target area. Set a higher weight for the target area and obtain a unified reconstructed image through iterative solution; (3) Multi-scale nesting: First, global reconstruction is completed based on coarse grid to obtain prior information of the image, which guides high-resolution reconstruction of the target area under fine grid. Then, the high-resolution reconstruction result of the target area is downsampled and embedded into the global image. (4) Deep learning assistance: The neural network is trained to learn the mapping relationship from low-resolution image to high-resolution image. The global reconstructed image and target area projection are used as network inputs, and the fused reconstructed image is directly output end-to-end.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: The system configuration of this invention is flexible and has low upgrade costs: This invention adopts a modular, pluggable design, allowing users to incrementally upgrade system performance by adding several extended detector modules with superior imaging characteristics to the basic detector module configuration, according to their needs. Existing hardware can be fully reused, avoiding complete system replacement and significantly reducing upgrade costs. Simultaneously, the modular installation interface reserves access capabilities for future new detector modules or new types of collimators, ensuring the architecture remains relevant despite technological advancements and guaranteeing long-term return on investment for users.

[0017] This invention balances global field of view with local high resolution: by having a basic detector module group and an extended detector module group work in parallel during a single imaging event, the basic detector module provides global projection data to ensure imaging coverage, while the extended detector module synchronously outputs high-precision projection data of the target area. This increases the amount of information acquired in a single scan from coverage area × conventional resolution to global coverage area × conventional resolution + target area coverage area × high resolution, thus resolving the contradiction that existing technologies cannot meet both requirements in the same scan.

[0018] This invention improves diagnostic efficiency and patient experience: since a single acquisition can complete the information that previously required two independent acquisitions, this invention significantly shortens the total examination time, reduces the radiation dose received by the patient, and fundamentally eliminates the image registration error caused by inconsistent body position between two acquisitions. It achieves a highly efficient diagnostic process of "one injection, one scan, and one fused image", thus improving the accuracy and reliability of diagnosis. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the configuration of the present invention in the hybrid parallel acquisition mode; Figure 2 This is a configuration diagram of the present invention in general mode; Figure 3 These are schematic diagrams illustrating the structures of different types of collimators in embodiments of the present invention; Figure 4 This is a schematic diagram of the fusion reconstruction process in the method of the present invention.

[0021] Reference numerals: 1-ring frame, 2-basic detector module, 3-extension detector module. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0023] A modular variable resolution full-ring SPECT imaging system, such as Figure 1 , Figure 2 As shown, the system includes a ring-shaped frame 1 and multiple detector modules. The ring-shaped frame 1 has multiple modular mounting interfaces evenly pre-set along its internal circumference, and the multiple detector modules are pluggably mounted on these mounting interfaces. The detector modules are divided into a basic detector module group and an extended detector module group. The basic detector module group includes at least one basic detector module 2, and the extended detector module group includes at least one extended detector module 3. The basic detector module 2 has a first imaging characteristic, and the extended detector module 3 has a second imaging characteristic superior to the first imaging characteristic. Specifically, each detector module includes a detector and a collimator. The first imaging characteristic of the basic detector module is large field-of-view imaging, and the collimator is a general-purpose collimator. The second imaging characteristic of the extended detector module is high-resolution imaging, and the collimator is a high-resolution collimator.

[0024] In one specific embodiment, 18 mounting interfaces are pre-set along the inner circumference of the annular frame 1, with an angular spacing of 20° between adjacent mounting interfaces. There are a total of 18 detector modules, including a basic detector module group containing 12 basic detector modules and an expansion detector module group containing 6 expansion detector modules. The mounting interfaces integrate mechanical positioning pin holes, electrical connectors, and data link interfaces, supporting automatic identification of detector modules and ensuring complete compatibility of signal links. Several common arrangement methods are provided below, but the detector module arrangement is not fixed and should not be construed as limiting the arrangement to the following three methods.

[0025] The number of installation interfaces and the ratio of expansion detector modules are not limited to these. The number of basic detector modules can be designed to be 1 to 5 times that of expansion detector modules. For example, the ratio of basic detector modules to expansion detector modules can be 8:4, 10:5, 16:8, etc. In principle, this is applicable to any incremental expansion mode with the number of basic detector modules plus the number of expansion detector modules, and it is preferable that the number of expansion detector modules is one-half or one-third of the number of basic detector modules.

[0026] The detector employs a cadmium zinc telluride (CZT) semiconductor detector. The CZT anode collector is cut into an anode pixel array, divided into independent readout channels. The core advantages of CZT semiconductor materials are: high detection efficiency achieved through high atomic numbers (Cd:48, Te:52); a wide bandgap (~1.57 eV) that keeps the CZT dark current at an extremely low level, allowing for stable operation at room temperature; and high energy resolution achieved through an average ionization energy of 4.64 eV. The CZT crystal specifications and electronic readout architecture of the detectors in the basic detector module group and the extended detector module group are completely identical, ensuring the compatibility of the signal acquisition link. The only difference lies in the type of collimator used.

[0027] In one specific embodiment, the general-purpose collimator is a parallel-aperture collimator, with the aperture axis perpendicular to the crystal incident surface and pointing towards the center of the scanning field of view. The aperture diameter ranges from 1.5 to 2.5 mm, and the aperture aspect ratio balances detection sensitivity and large field of view coverage. The high-resolution collimator is a key component for achieving a significant resolution leap. Depending on the imaging scenario, one or more of the following types of collimators can be selected: multi-aperture collimators, trapezoidal sub-aperture collimators, or multi-pinhole collimators. Figure 3 As shown. The multi-aperture collimator's substrate is integrally machined from a tungsten-based high-density alloy, and its dimensions and mounting interfaces are fully compatible with those of universal parallel-aperture collimators. Taking a four-aperture collimator as an example, as... Figure 3 As shown in Figure A, each reference through-hole is embedded with a cross-shaped physical grid, and the grid wall thickness is controlled to accommodate commonly used clinical radionuclides. 99Effective shielding against gamma rays from nuclides such as Tc, 177Lu, and 131I is achieved, eliminating photon crosstalk between adjacent sub-apertures. Simultaneously, the light transmission area loss caused by the grid is minimized, maintaining the highest possible detection sensitivity while improving resolution. A standard aperture is divided into four optically isolated independent sub-apertures; the equivalent aperture of each sub-aperture is half that of the reference aperture, increasing the channel depth-to-width ratio to twice. In addition to the four-sub-aperture cross grid, two-segment (dual-sub-aperture), nine-segment (3×3 grid), and sector-shaped segmentation (such as...) can also be used. Figure 3 (As shown in B), different numbers and geometric arrangements of sub-apertures, such as ring-shaped division, are used. The division grid can be made of high-density materials such as tantalum, tungsten alloy, gold, or platinum. The grid can also be installed as an independent thin sheet (rather than integrally formed with the substrate), as long as it has the same photon isolation function between sub-apertures, it is considered an equivalent structure. The correspondence between multi-aperture collimators and CZT pixels: Multiple sub-apertures share the same anode pixel of the CZT crystal, sampling the area covered by this pixel from multiple different spatial angles. This achieves a 50% reduction in the equivalent sampling interval through physical collimation without changing the intrinsic pixel size of the detector, thus doubling the system's spatial resolution. The exit face of the multi-aperture collimator is aligned with the detector's CZT crystal through a precision positioning groove, controlling the assembly gap between the grid intersection line and the anode pixel center; an integrated shielding edge is provided around the collimator to eliminate radiation leakage at module gaps. The channels of the trapezoidal sub-aperture collimator are arranged in a trapezoidal pattern, such as... Figure 3 As shown in C, the focus is on a local field of view; a multi-pinhole collimator, with multiple pinholes arranged in an array, utilizes geometric magnification to achieve ultra-high resolution in a small field of view. Figure 3 D shows one of the pinhole shapes.

[0028] Different extended detector modules in the extended detector module group can employ the same or different high-resolution collimators as needed. Specifically, some extended detector modules can be equipped with multi-aperture collimators, some with trapezoidal collimators, and some with pinhole collimators, forming a combined multi-type hybrid acquisition scheme to meet the differentiated resolution and field-of-view requirements of different target organs in the same scan. The coexistence and collaborative operation of a general-purpose collimator and one or more high-resolution collimators in the same scan is the core of the hybrid collimator parallel acquisition of this invention.

[0029] Furthermore, it also includes a data acquisition and control system and a fused image reconstruction subsystem. The data acquisition and control system is used to automatically identify the number and type of detector modules installed on the installation interface, and automatically load the corresponding acquisition protocol based on the identification results to control multiple detector modules to acquire data in parallel. The fused image reconstruction subsystem receives the parallel acquisition data from the detector modules and generates a fused tomographic image.

[0030] The fused image reconstruction subsystem includes a preprocessing module, a spatial mapping module, a fusion reconstruction module, and a post-processing output module. The preprocessing module is used to preprocess the global projection data acquired by the basic detector module and the target area high-resolution projection data acquired by the extended detector module. The spatial mapping module is used to uniformly map the high-resolution projection data to the global projection image coordinate system. The fusion reconstruction module is used to fuse the complete field of view information provided by the global projection and the high-density sampling information provided by the target area high-resolution projection within the same reconstruction framework to obtain a fused reconstructed image. The post-processing output module is used to post-process the fused reconstructed image and output a standard fused tomographic image.

[0031] The system of this invention features a general mode and a hybrid parallel acquisition mode. In general mode, 12 basic detector modules, in conjunction with a parallel-aperture collimator, complete 360° full-body acquisition, outputting a large-field-of-view tomographic image after standard reconstruction. In hybrid parallel acquisition mode, the 12 basic detector modules use a parallel-aperture collimator for large-area conventional sampling, while 6 extended detector modules use a high-resolution collimator for high-density, high-precision sampling of target organs. Both types of detectors share a timing synchronization signal within the same rotation cycle. When switching between general mode and hybrid parallel acquisition mode, the operator only needs to insert the 6 extended detector modules into the reserved installation interface; the system automatically identifies the module identity and collimator type and switches to the high-resolution acquisition protocol, without requiring mechanical disassembly.

[0032] On the other hand, the present invention provides a full-ring SPECT imaging method, which applies the above-mentioned modular variable resolution full-ring SPECT imaging system. The method is as follows: in a single imaging event, the basic detector module group and the extended detector module group are controlled to acquire data in parallel to obtain global projection data and target area high-resolution projection data respectively. The global projection data and target area high-resolution projection data are fused and reconstructed. The fusion and reconstruction process includes: using the global projection data as the basic reconstruction framework, mapping the target area high-resolution projection data to the corresponding spatial position of the global projection data, and generating a fused tomographic image that combines global large-scale and target area high resolution.

[0033] The steps of the fusion and reconstruction process are as follows: Figure 4 As shown, it includes: Step 1: Channel projection preprocessing The global projection data collected by the basic detector module and the target area high-resolution projection data collected by the extended detector module are preprocessed respectively. For the high-resolution projection data using a multi-aperture collimator, an additional sub-pixel channel separation operation needs to be performed to decompose the multi-channel projection output by the multi-aperture collimator into independent sub-pixel projections corresponding to each sub-aperture.

[0034] Step 2: Pixel Space Mapping By mapping the high-resolution projection to the global projection image coordinate system, the spatial reference difference between the two detectors is eliminated, providing a unified spatial reference for multi-source projection fusion.

[0035] Step 3: Integration and Reconstruction Within the same reconstruction framework, the complete field-of-view information provided by global projection and the high-density sampling information provided by high-resolution projection of the target area are fused to obtain a fused reconstructed image; one of the following methods is selected for implementation: (1) Step-by-step fusion: The whole body image is obtained by standard reconstruction using global projection, and the local image is obtained by high-resolution iterative reconstruction using target area projection; the local high-resolution image is embedded into the whole body image according to the spatial coordinates, and boundary smoothing is performed. (2) Joint optimization: Construct a weighted reconstruction objective function, and simultaneously fit two sets of data: global projection and high-resolution projection of the target area. Set a higher weight for the target area and obtain a unified reconstructed image through iterative solution; (3) Multi-scale nesting: First, global reconstruction is completed based on coarse grid to obtain prior information of the image, which guides high-resolution reconstruction of the target area under fine grid. Then, the high-resolution reconstruction result of the target area is downsampled and embedded into the global image. (4) Deep learning assistance: The neural network is trained to learn the mapping relationship from low-resolution image to high-resolution image. The global reconstructed image and target area projection are used as network inputs, and the fused reconstructed image is directly output end-to-end.

[0036] Step 4: Post-processing and output The fused and reconstructed images are post-processed to eliminate artifacts caused by image stitching, and the resulting fused tomographic images conform to the DICOM standard are output after adjustment.

[0037] Of course, the present invention may have other various embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and modifications according to the present invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

1. A modular variable resolution full-ring SPECT imaging system, characterized in that: It includes a ring frame and multiple detector modules; the ring frame has multiple modular mounting interfaces evenly pre-set along its internal circumference, and the multiple detector modules can be plugged into and detached and mounted on the mounting interfaces; the detector modules are divided into a basic detector module group and an extended detector module group, the basic detector module group includes at least one basic detector module, the extended detector module group includes at least one extended detector module, the basic detector module has a first imaging characteristic, and the extended detector module has a second imaging characteristic that is superior to the first imaging characteristic.

2. The modular variable resolution full-ring SPECT imaging system according to claim 1, characterized in that: Each of the detector modules includes a detector and a collimator; the first imaging characteristic of the basic detector module is large field-of-view imaging, and the collimator is a general-purpose collimator; the second imaging characteristic of the extended detector module is high-resolution imaging, and the collimator is a high-resolution collimator.

3. The modular variable resolution full-ring SPECT imaging system according to claim 2, characterized in that: The general-purpose collimator is a parallel-hole collimator, and the high-resolution collimator is one or more of the following: multi-sub-hole collimator, trapezoidal sub-hole collimator, or multi-pinhole collimator.

4. The modular variable resolution full-ring SPECT imaging system according to claim 2, characterized in that: All detectors used are cadmium zinc telluride semiconductor detectors.

5. The modular variable resolution full-ring SPECT imaging system according to claim 1, characterized in that: The installation interface is uniformly preset with N ports along the inner circumference of the ring frame, where N is a natural number ≥ 2. The number of basic detector modules is 1 to 5 times that of extended detector modules.

6. The modular variable resolution full-ring SPECT imaging system according to claim 1, characterized in that: It also includes a data acquisition and control system and a fused image reconstruction subsystem. The data acquisition and control system is used to automatically identify the number and type of detector modules installed on the installation interface, and automatically load the corresponding acquisition protocol based on the identification results to control multiple detector modules to acquire data in parallel. The fused image reconstruction subsystem receives the parallel acquisition data from the detector modules and generates a fused tomographic image.

7. The modular variable resolution full-ring SPECT imaging system according to claim 6, characterized in that: The fused image reconstruction subsystem includes a preprocessing module, a spatial mapping module, a fusion reconstruction module, and a post-processing output module. The preprocessing module is used to preprocess the global projection data acquired by the basic detector module and the target area high-resolution projection data acquired by the extended detector module. The spatial mapping module is used to uniformly map the high-resolution projection data to the global projection image coordinate system. The fusion reconstruction module is used to fuse the complete field of view information provided by the global projection and the high-density sampling information provided by the target area high-resolution projection within the same reconstruction framework to obtain a fused reconstructed image. The post-processing output module is used to post-process the fused reconstructed image and output a standard fused tomographic image.

8. A full-ring SPECT imaging method, using the modular variable resolution full-ring SPECT imaging system according to any one of claims 1 to 7, characterized in that, The method is as follows: In a single imaging event, the basic detector module group and the extended detector module group are controlled to acquire data in parallel to obtain global projection data and target area high-resolution projection data respectively. The global projection data and target area high-resolution projection data are then fused and reconstructed. The fusion and reconstruction process includes: using the global projection data as the basic reconstruction framework, mapping the target area high-resolution projection data to the corresponding spatial position of the global projection data, and generating a fused tomographic image that combines the integrity of the whole-body field of view with the local high resolution of the target organ.

9. The full-ring SPECT imaging method according to claim 8, characterized in that: The steps of the fusion reconstruction process include: Step 1: Channel projection preprocessing The global projection data collected by the basic detector module and the high-resolution projection data of the target area collected by the extended detector module are preprocessed respectively. Step 2: Pixel Space Mapping Map the high-resolution projection uniformly to the global projection image coordinate system; Step 3: Integration and Reconstruction Within the same reconstruction framework, the complete field-of-view information provided by global projection and the high-density sampling information provided by high-resolution projection of the target area are fused to obtain a fused reconstructed image; Step 4: Post-processing and output The fused and reconstructed images are post-processed and adjusted to output fused tomographic images conforming to the DICOM standard.

10. The full-circumference SPECT imaging method according to claim 9, characterized in that: In step 3, one of the following methods is selected for the fusion reconstruction: (1) Step-by-step fusion: The whole body image is obtained by standard reconstruction using global projection, and the local image is obtained by high-resolution iterative reconstruction using target area projection; the local high-resolution image is embedded into the whole body image according to the spatial coordinates, and boundary smoothing is performed. (2) Joint optimization: Construct a weighted reconstruction objective function, and simultaneously fit two sets of data: global projection and high-resolution projection of the target area. Set a higher weight for the target area and obtain a unified reconstructed image through iterative solution; (3) Multi-scale nesting: First, global reconstruction is completed based on coarse grid to obtain prior information of the image, which guides high-resolution reconstruction of the target area under fine grid. Then, the high-resolution reconstruction result of the target area is downsampled and embedded into the global image. (4) Deep learning assistance: The neural network is trained to learn the mapping relationship from low-resolution image to high-resolution image. The global reconstructed image and target area projection are used as network inputs, and the fused reconstructed image is directly output end-to-end.

Images