Low-cost and high-dynamic-range gamma imaging method and system

Abstract

The invention belongs to the technical field of nuclear medicine imaging, and discloses a low-cost and high-dynamic-range gamma imaging method and system, and the system comprises a detector array which is formed by splicing a plurality of detector modules, the detector module comprises a silicon photomultiplier array, a pixelated scintillator array coupled to the front side of the silicon photomultiplier array and a front-end reading circuit electrically connected with the silicon photomultiplier array; the pixelated scintillator array is used for receiving and responding to the gamma ray to generate scintillation light; the silicon photomultiplier array is used for converting the scintillation light into an electric signal; the front-end reading circuit is used for preprocessing and reading out the electric signals; the upper computer is used for acquiring detection position information of a gamma event and a calibrated real energy value based on a piecewise polynomial calibration algorithm according to the read-out signal of each detector module, and performing image reconstruction in combination with an energy weight iteration reconstruction algorithm; therefore, the gamma imaging capability with low cost and high dynamic range is realized, and the imaging quality can be remarkably improved.

Description

Technical Field

[0001] This application relates to the field of nuclear medicine imaging technology, and more specifically, to a low-cost, high dynamic range gamma imaging method and system. Background Technology

[0002] In the field of nuclear medicine imaging, gamma imaging technology is widely used in disease diagnosis and treatment monitoring. Currently, medical gamma imaging detectors on the market mainly adopt two technical approaches: one is the traditional combination of scintillation crystal and photomultiplier tube (PMT), and the other is a novel compound semiconductor detector approach.

[0003] While traditional methods offer a relatively wide energy linearity response, their energy response range is limited by the dynamic range of the subsequent electronic hardware system, thus restricting the system's gamma-ray imaging capabilities, especially in the high-energy range. When the gamma-ray energy exceeds the detector's energy response limit, the system cannot accurately acquire the energy information of the high-energy gamma rays, severely impacting imaging quality.

[0004] While novel compound semiconductor detectors (such as CZT and CdTe detectors) offer better energy resolution, their energy response range is actually narrower than traditional methods due to the inherent properties of the materials themselves, and their high manufacturing costs hinder large-scale clinical applications. Furthermore, there is room for improvement in the stitching methods of detector modules, signal readout mechanisms, and image reconstruction algorithms in existing technologies. These factors collectively limit the performance and cost control of gamma imaging systems over a wide energy range.

[0005] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0006] The purpose of this application is to provide a low-cost, high dynamic range gamma imaging method and system, which has low-cost, high dynamic range gamma imaging capabilities, can accurately acquire energy information of high-energy gamma rays, and significantly improve imaging quality.

[0007] In a first aspect, this application provides a low-cost, high dynamic range gamma imaging system, comprising: The detector array is composed of multiple detector modules. Each detector module includes a silicon photomultiplier array, a pixelated scintillator array coupled to the front side of the silicon photomultiplier array, and a front-end readout circuit electrically connected to the silicon photomultiplier array. The pixelated scintillator array is used to receive and respond to gamma rays to generate scintillation light. The silicon photomultiplier array is used to convert the scintillation light into an electrical signal. The front-end readout circuit is used to preprocess and read out the electrical signal. The host computer is used to obtain the detection location information and calibrated true energy value of the gamma event based on the readout signals of each detector module and the piecewise polynomial calibration algorithm, and to perform image reconstruction by combining the energy weight iterative reconstruction algorithm.

[0008] Secondly, this application provides a low-cost, high dynamic range gamma imaging method, based on the aforementioned low-cost, high dynamic range gamma imaging system, comprising the following steps: A1. Obtain the readout signals of each of the detector modules; A2. Calculate the detection location information and initial energy information of the gamma event based on the four signals in the readout signal; A3. The initial energy information is calibrated using a piecewise polynomial calibration algorithm to obtain the calibrated true energy value; A4. Based on the detection location information and the true energy value, an energy weight iterative reconstruction algorithm is used to reconstruct the image and generate a gamma-ray image.

[0009] Beneficial Effects: This application provides a low-cost, high dynamic range gamma imaging method and system. Utilizing a silicon photomultiplier as the photoelectric conversion device, its inherent cost advantage, combined with the modular splicing design of the detector array, significantly reduces the overall hardware manufacturing cost. At the data processing level, the host computer applies a piecewise polynomial calibration algorithm, which can accurately compensate for the nonlinear response of the detector in the high-energy range, thereby extending the energy response range to a wider interval and achieving accurate acquisition of high-energy gamma-ray energy information. Furthermore, by combining an energy weighted iterative reconstruction algorithm, the system can effectively distinguish and suppress noise caused by scattering events during image reconstruction, thus significantly improving the reconstruction quality of gamma-ray images. The comprehensive application of these technologies enables this system to possess low-cost, high dynamic range gamma imaging capabilities, accurately acquire high-energy gamma-ray energy information, and significantly improve imaging quality. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of a low-cost, high dynamic range gamma imaging system provided in this application.

[0011] Figure 2 This is a schematic diagram of the detector module.

[0012] Figure 3 A flowchart of a low-cost, high dynamic range gamma imaging method provided in this application.

[0013] Labeling Explanation: 1. Detector Array; 2. Detector Module; 201. Silicon Photomultiplier Array; 202. Pixelated Scintillator Array; 2021. Lateral Scintillator Strip; 2022. Vertical Scintillator Strip; 203. Front-end Readout Circuit; 2031. Channel Compression Readout Network; 2032. Preprocessing Module; 3. Host Computer. Detailed Implementation

[0014] 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 a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally 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 inventive effort are within the scope of protection of this application.

[0015] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0016] Please refer to Figures 1-2 A low-cost, high dynamic range gamma imaging system according to some embodiments of this application includes: The detector array 1 is composed of multiple detector modules 2. Each detector module 2 includes a silicon photomultiplier array 201, a pixelated scintillator array 202 coupled to the front side of the silicon photomultiplier array 201, and a front-end readout circuit 203 electrically connected to the silicon photomultiplier array 201. The pixelated scintillator array 202 is used to receive and respond to gamma rays to generate scintillation light. The silicon photomultiplier array 201 is used to convert the scintillation light into an electrical signal. The front-end readout circuit 203 is used to preprocess and read out the electrical signal. The host computer 3 is used to obtain the detection location information and the calibrated true energy value of the gamma event based on the readout signals of each detector module 2 and the piecewise polynomial calibration algorithm, and to perform image reconstruction by combining the energy weight iterative reconstruction algorithm (the specific process can be referred to the steps of the low-cost high dynamic range gamma imaging method in the following text).

[0017] The detector array 1 is composed of multiple independent detector modules 2, which aims to achieve large-area gamma ray detection, while reducing the overall system manufacturing cost and maintenance difficulty through modular design.

[0018] Among them, detector module 2 is the basic component of detector array 1. Each detector module 2 integrates gamma ray detection, photoelectric conversion and signal preprocessing functions, and can independently receive gamma rays and output corresponding electrical signals.

[0019] Among them, the silicon photomultiplier array 201 is a semiconductor photodetector composed of multiple silicon photomultipliers (SiPMs) (such as...). Figure 2 In the diagram, the portion indicated by label A is the unblocked part of a silicon photomultiplier (SiPM) scintillator array 202, used to convert the weak flicker light generated by the scintillators into a measurable electrical signal. Compared to traditional photomultiplier tubes (PMTs), SiPMs have advantages such as small size, low power consumption, insensitivity to magnetic fields, and relatively low cost.

[0020] The pixelated scintillator array 202 generates scintillating light when it receives gamma rays. Its structure can take various forms, such as being composed of multiple independent scintillator blocks or scintillator strips in both the horizontal and vertical directions. This pixelated design helps improve the localization accuracy of gamma events.

[0021] The front-end readout circuit 203 is electrically connected to the silicon photomultiplier array 201 and is responsible for performing preliminary processing on the electrical signals output by the silicon photomultiplier array 201, such as signal amplification, filtering and analog-to-digital conversion, so as to provide high-quality digital signals for subsequent processing by the host computer 3.

[0022] In this context, the host computer 3 typically refers to a computer or server responsible for receiving and processing readout signals from each detector module 2. The host computer 3 obtains detailed information about gamma events and ultimately reconstructs gamma-ray images by running specific algorithms, such as piecewise polynomial calibration algorithms and energy weight iterative reconstruction algorithms.

[0023] The piecewise polynomial calibration algorithm is used to calibrate the energy value of gamma events. Since silicon photomultipliers may exhibit nonlinear responses in the high-energy range, this algorithm expands the system's energy dynamic range by segmenting the energy response curve and applying different polynomial functions for fitting and calibration at different energy levels, ensuring accurate energy information is obtained even in the high-energy range.

[0024] Among them, the energy-weighted iterative reconstruction algorithm is an image reconstruction algorithm that introduces energy weights during the iterative reconstruction process. By assigning different weights to gamma events of different energies, the impact of scattering events on image quality can be effectively suppressed, improving the signal-to-noise ratio and sharpness of the image.

[0025] This application proposes a low-cost, high dynamic range gamma imaging system, aiming to solve the problems of limited energy response range, inability to perform normal imaging, and high cost of existing gamma imaging systems in the high-energy range. The system includes a detector array 1 and a host computer 3.

[0026] The detector array 1 is composed of multiple detector modules 2. This modular design allows the system to flexibly configure the detection area according to actual needs; for example, the size of the detector can be adjusted by increasing or decreasing the number of detector modules 2. Each detector module 2 includes a silicon photomultiplier array 201, a pixelated scintillator array 202 coupled to the front side of the silicon photomultiplier array 201, and a front-end readout circuit 203 electrically connected to the silicon photomultiplier array 201. The pixelated scintillator array 202 is used to receive and respond to gamma rays to generate scintillation light. For example, when gamma rays are incident on the scintillator array, the scintillator material absorbs the energy of the gamma rays and emits visible or ultraviolet light. The silicon photomultiplier array 201 is used to convert the scintillation light into an electrical signal. For example, after the scintillation light is received by the silicon photomultiplier array 201, it excites the avalanche effect inside the silicon photomultiplier, generating a charge pulse proportional to the number of photons. The front-end readout circuit 203 is used to preprocess and read out the electrical signal. For example, the front-end readout circuit 203 can amplify and filter the weak electrical signal output by the silicon photomultiplier array 201 and convert it into a digital signal for subsequent transmission and processing.

[0027] The host computer 3 is used to obtain the detection location information and calibrated true energy value of the gamma event based on the readout signals of each detector module 2 and a piecewise polynomial calibration algorithm. It then combines this with an energy weighted iterative reconstruction algorithm to reconstruct the image. For example, after receiving the digital signal from the front-end readout circuit 203, the host computer 3 first calibrates the original energy reading using a piecewise polynomial calibration algorithm to eliminate the nonlinear response of the silicon photomultiplier in the high-energy range. The calibrated energy value, along with the detection location information of the gamma event, is then fed into the energy weighted iterative reconstruction algorithm to ultimately generate a high-quality gamma-ray image.

[0028] The gamma imaging system of this application achieves a low-cost hardware foundation through a modularly designed detector array 1, combined with a silicon photomultiplier array 201 and a pixelated scintillator array 202. Compared to traditional PMT or CZT / CdTe detectors, the silicon photomultiplier array 201 is lower in cost and less sensitive to magnetic fields, making it more suitable for complex medical environments. Simultaneously, through a piecewise polynomial calibration algorithm, the system effectively addresses the nonlinear response problem of silicon photomultipliers in the high-energy range, extending the energy dynamic range from a limited linear response region to a wider energy range. This allows for accurate energy information even in the high-energy range, solving the problem of traditional systems failing to image properly in high-energy conditions. Furthermore, the introduction of an energy-weighted iterative reconstruction algorithm enables the system to distinguish between valid events and scattering events during image reconstruction, suppressing noise and improving image quality by assigning different weights. This overall technical solution, combining hardware innovation and algorithm optimization, not only significantly reduces the cost of the gamma imaging system but also greatly improves its energy dynamic range and image reconstruction quality, providing a more economical and accurate imaging method for medical diagnosis.

[0029] In some implementations, see Figure 2 Each silicon photomultiplier is considered as a pixel unit. The silicon photomultiplier array 201 includes multiple horizontal pixel columns and multiple vertical pixel columns. The pixelated scintillator array 202 includes multiple horizontal scintillator strips 2021 and multiple vertical scintillator strips 2022. The projection of each horizontal scintillator strip 2021 along the normal of the detector module 2 coincides with the projection of the horizontal center line of each horizontal pixel column along the normal of the detector module 2. The projection of each vertical scintillator strip 2022 along the normal of the detector module 2 coincides with the projection of the vertical center line of each vertical pixel column along the normal of the detector module 2.

[0030] Specifically, treating each silicon photomultiplier as a pixel unit means that each silicon photomultiplier independently receives and responds to the scintillation light in its corresponding area, thereby achieving precise detection of the location of gamma events. This modular design facilitates independent processing and analysis of the output signal of each silicon photomultiplier, providing fundamental data for subsequent location and energy measurement. For example, each silicon photomultiplier can be an independent chip or a miniature photodiode array integrated on a larger chip, with its output signal transmitted via independent pins or a shared readout line. The silicon photomultiplier array 201 consists of multiple silicon photomultipliers arranged in a specific two-dimensional grid structure. By organizing these pixel units into multiple horizontal and vertical pixel columns, a regular two-dimensional detection plane can be formed. This columnar structure helps to quickly determine the horizontal and vertical coordinates of a gamma event by analyzing the signal responses of different columns, thereby achieving precise location. For example, these pixel columns can be silicon photomultiplier units physically connected together or logically defined groups of silicon photomultipliers, with their signals read out using multiplexing or channel compression techniques. The pixelated scintillator array 202 is composed of multiple independent scintillator strips arranged in a specific pattern. When gamma rays interact with the scintillator strips, scintillating light is generated. By designing the scintillator strips into transverse and longitudinal stripe structures, the diffusion range of the scintillating light can be effectively limited, ensuring that it is primarily received by the silicon photomultiplier pixel unit directly below it. This pixelated design helps improve the spatial resolution of gamma events. For example, the scintillator strips can be made of scintillating materials such as sodium iodide (NaI(Tl)), lanthanum bromide (LaBr3(Ce)), or LYSO, and formed into independent strips through cutting or growth processes. These strips can be encapsulated in a reflective material to guide the scintillating light downwards.

[0031] This application employs an ingenious structural design to precisely align the silicon photomultiplier array 201 and the pixelated scintillator array 202 within the detector module 2. Specifically, each silicon photomultiplier is treated as an independent pixel unit, and these pixel units are organized into multiple horizontal pixel columns and multiple vertical pixel columns, forming a regular two-dimensional detection grid. Correspondingly, the pixelated scintillator array 202 is also composed of multiple horizontal scintillator strips 2021 and multiple vertical scintillator strips 2022. Crucially, the projection of each horizontal scintillator strip 2021 along the normal of the detector module 2 is designed to precisely coincide with the projection of the horizontal center line of each horizontal pixel column along the normal of the detector module 2. Simultaneously, the projection of each vertical scintillator strip 2022 along the normal of the detector module 2 also precisely coincides with the projection of the vertical center line of each vertical pixel column along the normal of the detector module 2. This bidirectional, centerline-aligned projection coincidence mechanism ensures that when gamma rays act within a scintillator crystal and generate scintillation light, this light can be efficiently and accurately received by the precisely aligned silicon photomultiplier pixel unit or pixel column directly below it. In this way, the center of the scintillation light signal is aligned with the centerline of the silicon photomultiplier pixel unit, greatly reducing positional information deviations caused by light signal diffusion or misalignment. This precise geometric correspondence allows the system to more accurately identify the location of gamma events within the scintillator, thus providing high-precision positional input for subsequent energy calibration and image reconstruction.

[0032] Furthermore, the pixelated scintillator array 202 can be directly coupled to the front side of the silicon photomultiplier array 201 via optical adhesive or light guide.

[0033] This application's solution aims to improve the overall performance of the gamma imaging system by optimizing the coupling method between the pixelated scintillator array 202 and the silicon photomultiplier array 201, thereby addressing the problems of low scintillation light transmission efficiency, signal loss, and severe scattering. Specifically, when gamma rays are incident on the pixelated scintillator array 202 of the detector module 2, the scintillator crystal units absorb the gamma ray energy and generate scintillation light. To ensure that this scintillation light can be efficiently collected and converted into an electrical signal, the pixelated scintillator array 202 is designed to be directly coupled to the silicon photomultiplier array 201 via optical adhesive or a light guide. This direct coupling method physically achieves a close contact between the scintillator array and the silicon photomultiplier array 201, minimizing the air gap and interface reflection between them. The optical adhesive or light guide, as an intermediate medium, has a refractive index that matches that of the scintillator and the silicon photomultiplier, effectively guiding the scintillation light and reducing scattering and absorption losses in the transmission path. Simultaneously, the pixelated scintillator array 202 is placed in front of the silicon photomultiplier array 201, ensuring the shortest transmission distance and optimal incident angle of the scintillation light from the source to the detector. This structural optimization enables the silicon photomultiplier array 201 to receive stronger and more complete scintillation light signals, thereby converting them into more accurate electrical signals with a higher signal-to-noise ratio. These high-quality electrical signals are then preprocessed and read out by the front-end readout circuit 203 and finally transmitted to the host computer 3. In this way, this scheme significantly improves the energy measurement accuracy and detection efficiency of gamma events, especially in high-energy gamma-ray detection scenarios. It can effectively maintain the dynamic range of the system, providing reliable raw data for the host computer 3 to obtain the detection location information and calibrated true energy value of the gamma event based on the piecewise polynomial calibration algorithm, and to perform image reconstruction by combining the energy weight iterative reconstruction algorithm, thereby improving the quality and reliability of the final gamma-ray image.

[0034] In some implementations, see Figure 2 The front-end readout circuit 203 includes a channel compression readout network 2031, which is used to output the electrical signals of all pixel units of the same silicon photomultiplier array 201 from four common output buses to obtain four analog signals corresponding to the horizontal positive, horizontal negative, vertical positive, and vertical negative directions, respectively (e.g., ...). Figure 2 In the diagram, X+ represents a positive analog signal in the horizontal direction, X- represents a negative analog signal in the horizontal direction, Y+ represents a positive analog signal in the vertical direction, and Y- represents a negative analog signal in the vertical direction.

[0035] The front-end readout circuit 203 is an electronic circuit directly connected to the detector array 1. Its main function is to acquire, condition, and digitize the raw electrical signals generated by the detector module 2 (e.g., silicon photomultiplier array 201). As the interface between the detector's sensitive element and subsequent data processing units (e.g., host computer 3), it is responsible for converting the weak analog signals output by the silicon photomultiplier array 201 into a format suitable for further analysis, typically involving signal amplification, noise suppression, and analog-to-digital conversion. This circuit can be implemented using a highly integrated application-specific integrated circuit (ASIC) to achieve low power consumption and miniaturization, or it can be implemented using discrete component circuit boards to provide greater design flexibility. The channel compression readout network 2031 is a specially designed electronic network designed to reduce the number of output channels required to read signals from the multi-pixel detector array 1. It avoids setting up a separate readout channel for each pixel by combining the signals of multiple pixels into a smaller set of common output lines. The core function of this network is to simplify the readout electronics, reduce hardware complexity, and reduce manufacturing costs, especially for large pixel arrays. This is achieved by encoding spatial information into a reduced number of output signals, for example, using various resistor or capacitor weighted networks, or through time-division multiplexing schemes. The four common output buses are four shared electrical paths or lines carrying compressed analog signals from the entire silicon photomultiplier array 201. These buses serve as the integrated output interface of the channel compression readout network 2031, transmitting the spatially encoded signals to the next stage of the front-end readout circuit 203. By using a fixed and limited number of buses (four in this application), the system significantly reduces the number of physical connections and subsequent processing channels, thereby reducing overall complexity. The four analog signals, corresponding to the lateral positive, lateral negative, longitudinal positive, and longitudinal negative directions respectively, are specific output signals generated by the channel compression readout network 2031, where each signal represents a weighted sum or difference of pixel signals along a specific spatial direction (X+, X-, Y+, Y-). These four analog signals collectively encode the two-dimensional spatial coordinates (X and Y) of the gamma event and its energy information. By analyzing the relative amplitudes of these four signals, the interaction positions of the gamma rays within detector array 1 can be precisely determined. This directional encoding is crucial for subsequent position reconstruction.

[0036] The solution of this application achieves efficient processing of the signal from the silicon photomultiplier array 201 by introducing a channel compression readout network 2031 into the front-end readout circuit 203. In the gamma imaging system, the detector array 1 is composed of multiple detector modules 2 spliced ​​together. Each detector module 2 includes a silicon photomultiplier array 201 and a pixelated scintillator array 202. Each silicon photomultiplier is considered as a pixel unit. The silicon photomultiplier array 201 includes multiple horizontal pixel columns and multiple vertical pixel columns. The pixelated scintillator array 202 includes multiple horizontal scintillator crystal strips 2021 and multiple vertical scintillator crystal strips 2022. The projection of each horizontal scintillator crystal strip 2021 along the normal of the detector module 2 coincides with the projection of the horizontal center line of each horizontal pixel column along the normal of the detector module 2. Similarly, the projection of each vertical scintillator crystal strip 2022 along the normal of the detector module 2 coincides with the projection of the vertical center line of each vertical pixel column along the normal of the detector module 2. This precise alignment of pixels and scintillator stripes lays the foundation for the channel compression readout network 2031 to effectively encode spatial information. The channel compression readout network 2031 integrates the electrical signals of all pixel units in the silicon photomultiplier array 201 and outputs them through four common output buses, forming four analog signals corresponding to the lateral positive, lateral negative, vertical positive, and vertical negative directions, respectively. This design significantly reduces the number of readout channels required, thereby simplifying the circuit structure and reducing hardware cost and power consumption. By analyzing the relative amplitudes of these four analog signals, the position of the gamma event within the detector can be accurately calculated, and the energy information of the gamma event can be obtained by summing these signals. This compression readout method reduces the number of channels while maintaining signal integrity, ensuring accurate energy measurement over a wide energy range, and effectively solving the problems of circuit complexity, high cost, and limited dynamic range in high-energy segments caused by independent pixel readout in traditional schemes.

[0037] In some possible implementations, the channel compression readout network 2031 is a resistance-weighted network, and the output of each pixel unit in the silicon photomultiplier array 201 is connected to the four common output buses through the resistance-weighted network.

[0038] Specifically, the resistor-weighted network is a passive circuit composed of resistive elements. Its main function is to combine multiple input signals according to preset weights and output a small number of synthesized signals. This network allows for precise configuration of resistor values, making the contribution of each input signal to the final output controllable. Besides the common R-2R ladder network, the resistor-weighted network can also be implemented using direct resistance summation, where multiple input signals are connected to a common node through resistors of different values, thus obtaining a weighted sum signal at the common node. This design has significant advantages in applications such as signal compression and position coding due to its simple structure, low cost, and high stability. Meanwhile, the output of each pixel unit in the silicon photomultiplier array 201 is connected to four common output buses through the resistor-weighted network, describing how the electrical signals generated by the silicon photomultiplier array 201 are effectively routed and processed. Specifically, each pixel unit in the silicon photomultiplier array 201 generates an electrical signal after receiving flash light. These independent electrical signals are not directly output but are guided to the resistor-weighted network. Within this network, the signal of each pixel unit is assigned a specific weight based on its physical position in the array, and then weighted and summed through a resistor network. Finally, these weighted signals are converged and output to four common output buses, representing the lateral positive, lateral negative, longitudinal positive, and longitudinal negative signal components, respectively. This connection and weighting mechanism ensures that the precise location information of gamma events within the detector can be effectively encoded into these four analog signals, providing a foundation for subsequent signal processing and position calculation.

[0039] The solution in this application solves the aforementioned problem by specifically implementing the channel compression readout network 2031 as a resistance-weighted network and connecting the output of each pixel unit in the silicon photomultiplier array 201 to four common output buses through this network. When gamma rays act on the pixelated scintillator array 202 and generate scintillation light, the corresponding pixel unit in the silicon photomultiplier array 201 converts the scintillation light into an electrical signal. These electrical signals generated by each pixel unit are then input into the resistance-weighted network. This resistance-weighted network weights and sums the electrical signals from different pixel units according to a preset resistance value. For example, the signals of pixel units located at different positions in the array are connected to the four common output buses (lateral positive, lateral negative, vertical positive, and vertical negative) through resistors with different resistance values. This weighted summation mechanism allows the relative amplitude of the analog signals on the four common output buses to accurately reflect the position of the gamma event within the detector module 2. In this way, the originally large number of pixel unit signals are efficiently compressed into four analog signals, greatly simplifying the complexity of the front-end readout circuit 203 and reducing hardware costs. Simultaneously, since the resistor-weighted network is passive and stable, it ensures the accuracy of signal processing and the stability of the output, avoiding positioning errors or noise introduction caused by improper signal compression. This design is highly compatible with the functional requirements of the channel compression readout network 2031 in the front-end readout circuit 203, providing high-quality input signals for subsequent preprocessing modules 2032 (such as filtering and amplification, bias processing, pulse waveform adjustment, and analog-to-digital conversion), thereby ensuring that the entire gamma imaging system achieves high dynamic range performance at low cost.

[0040] As a specific implementation, the resistor-weighted network can adopt a matrix structure based on resistor voltage division and summation. For example, for a two-dimensional silicon photomultiplier array 201, the output of each pixel unit can be connected to four common output buses through a set of resistors with preset resistance values. Specifically, for any pixel unit in the array, its output signal is fed to the common buses in the horizontal positive, horizontal negative, vertical positive, and vertical negative directions through different resistor paths. The resistance values ​​of these resistors can be designed according to the coordinate position of the pixel unit in the array, so that the closer the pixel is to the horizontal positive edge, the greater its contribution weight to the horizontal positive bus, and the smaller its contribution weight to the horizontal negative bus, and vice versa. The weighting method of the vertical signal follows a similar principle. For example, a "cross-shaped" or "quadrangle" weighted network can be used, in which the signal of each pixel is connected to four output buses through four resistors, and the resistance value is inversely proportional or directly proportional to the distance from the pixel to the "center" of the corresponding bus. This resistor network can effectively integrate the signals from all pixel units and output four analog signals. The relative magnitudes of these signals directly encode the two-dimensional location information of the gamma event.

[0041] Furthermore, see Figure 2 The front-end readout circuit 203 further includes a preprocessing module 2032, which is used to preprocess the four analog signals to obtain and output the readout signal; the preprocessing includes filtering and amplification, bias processing, pulse waveform adjustment and analog-to-digital conversion processing.

[0042] The preprocessing module 2032 is a dedicated functional unit within the front-end readout circuit 203. Its main function is to refine and optimize the raw analog signal output from the channel compression readout network 2031 to improve signal quality and provide a reliable data foundation for subsequent digital processing and gamma event analysis. This module can be implemented as a custom integrated circuit (ASIC), a field-programmable gate array (FPGA) with analog front-end components, or a circuit board composed of discrete components. The preprocessing module 2032 receives four analog signals and performs a series of operations on them to enhance signal quality, eliminate artifacts, and prepare for digital conversion and subsequent analysis. After preprocessing, the module generates and outputs a clean, digitized, and standardized readout signal for further processing by the host computer 3.

[0043] The filtering and amplification aims to remove unwanted noise components from the signal while increasing the signal amplitude to improve the signal-to-noise ratio and fully utilize the dynamic range of the analog-to-digital converter. This can be achieved by combining active filters (e.g., Butterworth or Chebyshev filters implemented by operational amplifiers) or passive filters (e.g., RC filters) with low-noise amplifiers (LNAs) or variable-gain amplifiers (VGAs).

[0044] The biasing process is used to adjust the DC offset of the signal, ensuring that the signal falls within the optimal input range for subsequent processing stages (such as analog-to-digital converters) and eliminating baseline drift. This can involve DC coupling using a voltage reference, AC coupling using a DC recovery circuit, or digital offset correction after analog-to-digital conversion.

[0045] The purpose of pulse waveform adjustment is to optimize the shape of the signal pulse for optimal detection, timing, and energy measurement, typically achieved by optimizing rise / fall times or pulse width. This can be accomplished using pulse shaping amplifiers (e.g., Gaussian shapers, CR-RC shapers) or differentiating / integrating circuits.

[0046] The analog-to-digital conversion process is a crucial step in converting analog signals into digital format, enabling the host computer 3 to perform digital processing. This can be accomplished by a successive approximation (SAR) analog-to-digital converter, a flash analog-to-digital converter, or a Sigma-Delta analog-to-digital converter, depending on the required accuracy, speed, and power consumption.

[0047] The solution proposed in this application effectively solves the problems of noise, distortion, and baseline drift that may exist in the four analog signals output by the channel compression readout network 2031 by introducing a preprocessing module 2032 into the front-end readout circuit 203. Specifically, the four analog signals output from the channel compression readout network 2031 first enter the preprocessing module 2032 for filtering and amplification to remove high-frequency noise and enhance signal strength, thereby improving the signal's anti-interference capability and the effective dynamic range of the analog-to-digital converter. Subsequently, these filtered and amplified signals undergo bias processing to precisely adjust their DC offset, ensuring signal baseline stability and placing them within the optimal input range of the analog-to-digital converter. Next, the shape of the signal pulses is optimized through pulse waveform adjustment, which is crucial for accurately extracting the occurrence time, amplitude, and duration of gamma events, directly affecting the accuracy of detection location and energy information. Finally, the refined analog signals are converted into digital signals through analog-to-digital conversion. This digitized readout signal has higher noise immunity and can seamlessly interface with the digital processing system of the host computer 3. Through this series of preprocessing steps, the scheme of this application ensures that the readout signal provided to the host computer 3 is of high quality, stable and accurate, thereby significantly improving the calculation accuracy of gamma event detection location information and initial energy information, laying a solid foundation for subsequent piecewise polynomial calibration and image reconstruction, and ultimately improving the overall performance and reliability of the gamma imaging system.

[0048] Secondly, refer to Figure 3 This application provides a low-cost, high dynamic range gamma imaging method based on the aforementioned low-cost, high dynamic range gamma imaging system, comprising the following steps: A1. Obtain the readout signals of each of the detector modules 2; A2. Calculate the detection location information and initial energy information of the gamma event based on the four signals in the readout signal; A3. The initial energy information is calibrated using a piecewise polynomial calibration algorithm to obtain the calibrated true energy value; A4. Based on the detection location information and the true energy value, an energy weight iterative reconstruction algorithm is used to reconstruct the image and generate a gamma-ray image.

[0049] The method first acquires the readout signals from each detector module 2 in step A1. This step aims to obtain pre-processed electrical signals from the detector array 1 of the gamma imaging system. The readout signals are the data processed and output by the front-end readout circuit 203 after the detector module 2 responds to a gamma event. Their function is to provide the raw data basis for subsequent gamma event analysis and image reconstruction. The readout signals are typically digital signals obtained by analog-to-digital conversion.

[0050] After acquiring the readout signals, step A2 calculates the detection location and initial energy information of the gamma event based on four of these signals. This step utilizes the four signals provided by the front-end readout circuit 203 to determine the location of the gamma event within the detector module 2 and its corresponding initial energy. These four signals typically correspond to lateral positive, lateral negative, longitudinal positive, and longitudinal negative signals, respectively, and contain spatial and energy information about the gamma event occurring within the detector. By processing these four signals using specific algorithms, the precise detection location of the gamma event can be calculated, for example, using a centroid positioning algorithm. Simultaneously, by summing the amplitudes of these four signals, the initial energy information of the gamma event can be obtained.

[0051] To overcome the inaccuracy of initial energy information in the high-energy range, this application introduces a piecewise polynomial calibration algorithm in step A3. This step aims to address the nonlinearity of the detector's response across different energy ranges, particularly the signal saturation phenomenon that may occur in silicon photomultipliers at high energy levels. The piecewise polynomial calibration algorithm establishes a mapping relationship between the initial energy information and the true energy value, converting the nonlinear initial energy information into an accurate true energy value. The core of this algorithm lies in dividing the entire energy range into different segments based on the energy response characteristics and applying a different polynomial function to each segment for fitting and calibration; for example, a linear function is used for the low-energy range, and a higher-order polynomial function is used for the high-energy range.

[0052] Based on this, step A4 uses the calibrated true energy value and detector location information to perform image reconstruction using an energy-weighted iterative reconstruction algorithm, ultimately generating a gamma-ray image. This step utilizes the calibrated true energy value and detector location information to generate a high-quality gamma-ray image. The energy-weighted iterative reconstruction algorithm introduces energy information as weights to distinguish the reliability of different gamma-ray events during image reconstruction. Its function is to suppress the impact of noise such as scattering events on image quality, improving the signal-to-noise ratio and sharpness of the image. This algorithm is typically based on the principle of iterative optimization, continuously updating the image pixel values ​​to match the actual detection data and energy weights, ultimately obtaining a high-fidelity gamma-ray image.

[0053] The method first acquires the readout signals from each detector module 2 in step A1. These signals are generated by the pixelated scintillator array 202 after gamma rays interact with the detector array 1, and the silicon photomultiplier array 201 converts them into electrical signals. These signals are then processed by the channel compression readout network 2031 of the front-end readout circuit 203 and the preprocessing module 2032. Specifically, the channel compression readout network 2031 of the front-end readout circuit 203 outputs the electrical signals of all pixel units of the silicon photomultiplier array 201 from four common output buses, resulting in four analog signals corresponding to the lateral positive, lateral negative, vertical positive, and vertical negative directions, respectively. Subsequently, the preprocessing module 2032 performs filtering, amplification, bias processing, pulse waveform adjustment, and analog-to-digital conversion on these four analog signals, finally outputting the readout signals. After acquiring the readout signals, step A2 calculates the detection location information and initial energy information of the gamma event based on the four signals. While this step can initially determine the location and energy of the gamma event, the silicon photomultiplier array 201 may approach signal saturation when receiving high-energy gamma rays, causing its output initial energy information to exhibit a nonlinear response in the high-energy range, thus lacking accuracy. To overcome the inaccuracy of the initial energy information in the high-energy range, this application introduces a piecewise polynomial calibration algorithm in step A3. This algorithm finely calibrates the initial energy information obtained in step A2, converting it into a calibrated true energy value. By pre-calibrating and establishing a piecewise polynomial mapping relationship, this algorithm can effectively compensate for the nonlinear response of the detector in different energy ranges, especially in the high-energy range, "straightening" the nonlinear initial energy information into an accurate true energy value. This calibration process is crucial for achieving high dynamic range imaging, ensuring that the energy data used for subsequent image reconstruction has high accuracy and high reliability. Based on this, step A4 uses the calibrated true energy value and detector location information to perform image reconstruction using an energy weighted iterative reconstruction algorithm, ultimately generating a gamma-ray image. This reconstruction algorithm fully utilizes the accurate true energy value provided in step A3. For each gamma event, a corresponding confidence weight is calculated based on the deviation between its true energy value and the preset photoelectric peak energy. These weights are cleverly integrated into the image pixel value update formula during the iterative reconstruction process. This results in "good events" with energies near the photoelectric peak receiving higher weights, thus dominating image reconstruction, while "bad events" containing scattering components are assigned lower weights, effectively suppressing their negative impact on image quality. In this way, this method not only solves the problem of inaccurate initial energy information in the high-energy range but also further improves the signal-to-noise ratio and sharpness of the image during reconstruction through the energy weight mechanism, balancing spatial resolution and energy resolution, thereby achieving low-cost, high dynamic range gamma imaging.

[0054] In some implementations, step A2 includes: A201. Based on the four signals in the readout signal, the detection location information of the gamma event is calculated using the centroid positioning algorithm; A202. Calculate the sum of the energies corresponding to the four signals in the readout signal to obtain the initial energy information of the gamma event.

[0055] Specifically, the amplitudes or integral values ​​of the four signals in the readout signal reflect the energy deposition distribution of the gamma event within detector module 2. The centroid localization algorithm is a technique widely used in position-sensitive detectors. Its core idea is to accurately determine the position of the gamma event within the pixelated scintillator array 202 by weighted averaging the signal intensities received at different locations of the detector. For example, a two-dimensional weighted averaging method, such as an algorithm based on Gaussian fitting or template matching (which is existing technology and will not be detailed here), can be used to further improve the localization accuracy. This step aims to obtain the precise two-dimensional coordinates of the gamma event within detector module 2, such as (X, Y) coordinates (the position of the corresponding pixel unit). This positional information is the basis for subsequent image reconstruction, and its accuracy directly affects the spatial resolution of the final gamma-ray image. Simultaneously, the sum of the energies corresponding to the four signals in the readout signal is calculated. This step involves summing the amplitudes or integral values ​​of the four pre-processed signals from the channel compression readout network 2031 to obtain the total energy response of the gamma event. This summation operation effectively collects all energy information deposited by the gamma event within the pixelated scintillator array 202. This sum represents the raw energy response deposited in the detector by the gamma event, which is usually positively correlated with the true energy of the gamma ray and serves as the input for subsequent piecewise polynomial calibration algorithms.

[0056] To address the accuracy and efficiency issues that traditional methods may encounter when calculating the location and energy of gamma events, this application proposes a refined processing procedure. First, a centroid localization algorithm is used to process the four readout signals. By analyzing the relative intensity of the signals in different directions, the detection location information of the gamma event within detector module 2 can be accurately calculated. The centroid localization algorithm fully utilizes the spatial distribution characteristics of the signals, effectively improving the location resolution and providing more accurate input for subsequent image reconstruction. Second, by simply and efficiently calculating the sum of the corresponding energies of the four readout signals, the initial energy information of the gamma event can be quickly obtained. This summation method directly reflects the total energy deposited by the gamma event in the detector, laying the foundation for subsequent energy calibration. Through the above technical means, this application's solution can simultaneously and accurately extract the detection location and initial energy information of the gamma event from the four signals efficiently acquired at the front end. This processing method not only improves the accuracy of location and energy information acquisition but also simplifies the calculation process and enhances processing efficiency. This, combined with the efficient generation mechanism of the channel compression readout network 2031 in the front-end readout circuit 203, enables the entire system to achieve high dynamic range gamma imaging at a lower cost, providing high-quality data input for subsequent piecewise polynomial calibration and energy weight iterative reconstruction, thereby significantly improving the quality of the final gamma-ray image.

[0057] In some implementations, step A3 includes: A301. Compare the initial energy information with the preset energy threshold; A302. If the initial energy information is not greater than the preset energy threshold, the initial energy information is calibrated according to the pre-calibrated linear mapping function to obtain the calibrated true energy value; A303. If the initial energy information is greater than the preset energy threshold, the initial energy information is calibrated according to the pre-calibrated higher-order polynomial mapping function to obtain the calibrated true energy value; the linear mapping function and the higher-order polynomial mapping function smoothly transition at the preset energy threshold.

[0058] Specifically, when calibrating the initial energy information, it is first necessary to compare the initial energy information with a preset energy threshold. This preset energy threshold is used to divide the energy response range of the gamma event into different segments, such as low-energy and high-energy segments, thus providing a basis for subsequently selecting a suitable energy calibration function. This comparison process can be implemented through a hardware comparator circuit or completed through software logic judgment in the host computer. The preset energy threshold can be a fixed value preset according to the detector characteristics, such as 400keV, or it can be a parameter that is dynamically adjusted according to actual application requirements.

[0059] If the initial energy information is not greater than a preset energy threshold, the initial energy information is calibrated according to a pre-calibrated linear mapping function to obtain the calibrated true energy value. The linear mapping function is typically suitable for low-energy ranges where the detector response is relatively linear. Its parameters can be obtained by testing and calibrating detector module 2 using a standard radiation source with known energy (such as Am-241), and then fitting the data using linear regression methods such as least squares (for example, the linear mapping function can be...). Where E is the calibrated true energy value and A is the initial energy information. , These are polynomial coefficients, which can be obtained through fitting. This calibration process is typically implemented in a host computer by executing a corresponding software algorithm.

[0060] If the initial energy information is greater than the preset energy threshold, the initial energy information is calibrated according to the pre-calibrated higher-order polynomial mapping function to obtain the calibrated true energy value. The higher-order polynomial mapping function is used to handle high-energy ranges where the detector response exhibits saturation or nonlinearity, and can more accurately compensate for nonlinear effects. Its parameters are also obtained by testing and calibrating detector module 2 using various high-energy standard radioactive sources (such as Cs-137 and Co-60), and then fitting the data using a polynomial regression method (for example, the higher-order polynomial mapping function can be...). , , , These are polynomial coefficients, which can be obtained through fitting. This calibration process is also typically implemented in the host computer by executing the corresponding software algorithm.

[0061] To ensure the continuity and consistency of calibration results and avoid data jumps at energy thresholds, the linear mapping function and the higher-order polynomial mapping function are designed to have a smooth transition at the preset energy threshold. This can be achieved by introducing specific mathematical constraints when fitting the two functions, such as requiring that the function values ​​and first derivatives (or higher-order derivatives) of the two functions at the preset energy threshold be equal. This smooth transition constraint is processed by an optimization algorithm during the calibration phase, thereby ensuring the continuity and robustness of the entire energy calibration curve.

[0062] This application's solution addresses the nonlinear response problem in energy calibration by refining the implementation of the piecewise calibration algorithm, ensuring accurate true energy values ​​across different energy ranges and thus improving the overall accuracy of image reconstruction. In gamma imaging methods, the initial energy information obtained in step A2, especially for detectors like the silicon photomultiplier array 201, may exhibit nonlinear responses due to signal saturation in the mid-to-high energy range. This solution addresses this nonlinearity by introducing a piecewise polynomial calibration algorithm. Specifically, by comparing the initial energy information with a preset energy threshold, the energy range is divided into low-energy and high-energy ranges. In the low-energy range, utilizing the relatively linear nature of the detector response, a pre-calibrated linear mapping function is used for calibration, simplifying the process and maintaining accuracy. In the high-energy range, considering the saturation characteristics of the detector response, a pre-calibrated higher-order polynomial mapping function is used for compensation fitting, effectively handling the complexity of high-energy nonlinearity and significantly improving calibration accuracy. Furthermore, by ensuring a smooth transition between the linear mapping function and the higher-order polynomial mapping function at the preset energy threshold, error accumulation due to discontinuities is avoided, enhancing the robustness and reliability of the calibration process. This precise energy calibration enables the subsequent image reconstruction using the energy weighted iterative reconstruction algorithm in step A4 to be based on more reliable true energy values, thereby significantly improving the quality and accuracy of gamma-ray images, especially in wide-spectrum imaging applications.

[0063] In some implementations, step A4 includes: A401. For each gamma event, calculate the confidence weight of the gamma event based on the deviation between the actual energy value of the gamma event and the preset photoelectric peak energy; A402. Based on the array geometric parameters of the pixel units of detector module 2 and the detection position information of each gamma event, establish a system response matrix and initialize the image matrix to be reconstructed; each element of the system response matrix is ​​the response probability of each image pixel to each gamma event detection position, and each element of the image matrix to be reconstructed is the image pixel value of the image to be reconstructed. A403. Based on the confidence weight and the system response matrix, iteratively reconstruct the image matrix to be reconstructed using the OSEM algorithm to update the image pixel values ​​in the image matrix to be reconstructed; A404. Repeat step A403 until the preset iteration termination condition is met, and convert the final image matrix to be reconstructed into a gamma-ray image.

[0064] In the above method, calculating the confidence weight of gamma events is a crucial step. This confidence weight quantifies the reliability of each gamma event, i.e., how close its energy value is to the ideal photoelectric peak energy (i.e., the preset photoelectric peak energy). Generally, events with energies near the photoelectric peak are considered "good events" with high confidence, while events deviating from the photoelectric peak may contain scattering components and have lower confidence. The confidence weight can be calculated using various mathematical models; for example, it can be based on the Gaussian distribution function, with the following formula: , Let be the credibility weight of the i-th gamma event. Let i be the true energy value of the i-th gamma event. To preset the photoelectric peak energy, This is the system's energy resolution parameter. Alternatively, the Lorentz function or other forms of bell-shaped functions can be used to describe the energy spectrum distribution and calculate the weights accordingly. The preset photoelectric peak energy... The energy resolution parameter is usually determined by systematically calibrating a known radioactive source. It can be derived from the full width at half maximum (FWHM) of the photoelectric peak in the energy spectrum obtained from the calibration experiment.

[0065] During image reconstruction, a system response matrix needs to be established and the image matrix to be reconstructed needs to be initialized. Each element of the system response matrix represents the response probability of each image pixel (i.e., each pixel unit) to each gamma event detection location (i.e., the probability that each pixel unit will respond when a gamma event occurs at its location). This matrix can be precisely calculated using Monte Carlo simulations (e.g., using the GEANT4 toolkit) to simulate the interaction of gamma rays within the geometry of detector module 2, or through experimental measurements, i.e., obtaining response data by scanning the detector with a point source at a known location. Each element of the image matrix to be reconstructed represents the pixel value of the image to be reconstructed, representing the spatial distribution of the gamma ray source. This matrix is ​​typically initialized with uniform values ​​(usually values ​​greater than zero, e.g., all pixel values ​​are 1 or other positive numbers) to avoid division-by-zero errors during iterative calculations, or it can be initialized based on prior information.

[0066] Subsequently, based on the calculated confidence weights and the established system response matrix, the image matrix to be reconstructed is iteratively reconstructed using the OSEM algorithm to update the image pixel values. OSEM (Ordered Subset Expectation Maximization) is an iterative reconstruction algorithm widely used in nuclear medicine imaging. It accelerates the convergence of the reconstruction process by dividing the detected gamma event data into multiple subsets and updating the image for each subset. In each iteration, the algorithm updates the image pixel values ​​based on the current image estimate (i.e., the current image matrix to be reconstructed), the system response matrix, and the confidence weight of each gamma event. For example, the update formula can be expressed as: ,in, Let j be the value of the j-th image pixel after the nth iteration. This represents the value of the j-th image pixel after the (n+1)-th iteration. Let be the element in the i-th row and j-th column of the system response matrix, representing the response probability of the j-th image pixel to the detection location of the i-th gamma event. This represents the value of the k-th image pixel after the nth iteration update. Let be the element in the i-th row and k-th column of the system response matrix, representing the response probability of the k-th image pixel to the detection location of the i-th gamma event. This is the effective count for the i-th gamma event, typically set to 1. Besides the OSEM algorithm, other iterative reconstruction algorithms can also be used, such as the Maximum Likelihood Expectation Maximization (MLEM) algorithm or the Algebraic Reconstruction Technique (ART).

[0067] The iterative reconstruction process described above is repeated until preset termination conditions are met. These termination conditions may include reaching a preset maximum number of iterations (e.g., 100 iterations), or when the change in the image matrix to be reconstructed between consecutive iterations is less than a certain preset threshold (e.g., the relative rate of change of all image pixel values ​​is less than 0.1%), or when the quality metrics of the reconstructed image (such as signal-to-noise ratio, contrast) tend to stabilize. Once the termination conditions are met, the final image matrix to be reconstructed is converted into a gamma-ray image, which typically involves normalizing pixel values, applying pseudo-color mapping, and saving the image in a standard format (such as DICOM, PNG, etc.).

[0068] The proposed solution, through the aforementioned steps, uses the detection location information of gamma events and the true energy values ​​obtained through a piecewise polynomial calibration algorithm as input. First, it calculates the confidence weight for each gamma event, effectively distinguishing between "good events" at the photoelectric peak and "bad events" containing scattering noise. Then, based on the array geometry parameters of detector module 2 and the detection location information of gamma events, it constructs a system response matrix and initializes the image matrix to be reconstructed, providing an accurate mathematical model for image reconstruction. On this basis, iterative reconstruction is performed using the OSEM algorithm, incorporating the confidence weight into the iterative update formula. This ensures that high-weighted, high-quality data dominates image formation while reducing the interference of low-weighted scattering data on the image background. This mechanism significantly suppresses noise, improves image clarity, and accelerates the convergence of the reconstruction process. Finally, by setting reasonable iteration termination conditions, it ensures the stability and reliability of the reconstruction results, thereby generating high-quality gamma-ray images.

[0069] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A low-cost, high dynamic range gamma imaging system, characterized in that, include: The detector array (1) is composed of multiple detector modules (2). Each detector module (2) includes a silicon photomultiplier array (201), a pixelated scintillator array (202) coupled to the front side of the silicon photomultiplier array (201), and a front-end readout circuit (203) electrically connected to the silicon photomultiplier array (201). The pixelated scintillator array (202) is used to receive and respond to gamma rays to generate scintillating light. The silicon photomultiplier array (201) is used to convert the scintillating light into an electrical signal. The front-end readout circuit (203) is used to preprocess and read out the electrical signal. The host computer (3) is used to obtain the detection location information of the gamma event and the true energy value after calibration based on the readout signal of each detector module (2) and the piecewise polynomial calibration algorithm, and to perform image reconstruction in combination with the energy weight iterative reconstruction algorithm.

2. The low-cost, high dynamic range gamma imaging system according to claim 1, characterized in that, Each silicon photomultiplier is considered as a pixel unit. The silicon photomultiplier array (201) includes multiple horizontal pixel columns and multiple vertical pixel columns. The pixelated scintillator array (202) includes multiple horizontal scintillator strips (2021) and multiple vertical scintillator strips (2022). The projection of each horizontal scintillator strip (2021) along the normal of the detector module (2) coincides with the projection of the horizontal center line of each horizontal pixel column along the normal of the detector module (2). The projection of each vertical scintillator strip (2022) along the normal of the detector module (2) coincides with the projection of the vertical center line of each vertical pixel column along the normal of the detector module (2).

3. The low-cost, high dynamic range gamma imaging system according to claim 1, characterized in that, The pixelated scintillator array (202) is directly coupled to the front side of the silicon photomultiplier array (201) via optical adhesive or light guide.

4. The low-cost, high dynamic range gamma imaging system according to claim 2, characterized in that, The front-end readout circuit (203) includes a channel compression readout network (2031), which is used to output the electrical signals of all pixel units of the same silicon photomultiplier array (201) from four common output buses to obtain four analog signals corresponding to the horizontal positive direction, the horizontal negative direction, the vertical positive direction and the vertical negative direction respectively.

5. A low-cost, high dynamic range gamma imaging system according to claim 4, characterized in that, The channel compression readout network (2031) is a resistance-weighted network, and the output of each pixel unit in the silicon photomultiplier array (201) is connected to the four common output buses through the resistance-weighted network.

6. A low-cost, high dynamic range gamma imaging system according to claim 4, characterized in that, The front-end readout circuit (203) further includes a preprocessing module (2032), which is used to preprocess the four analog signals to obtain and output the readout signal; the preprocessing includes filtering and amplification, bias processing, pulse waveform adjustment and analog-to-digital conversion processing.

7. A low-cost, high dynamic range gamma imaging method, characterized in that, The low-cost, high dynamic range gamma imaging system according to any one of claims 4-6 includes the following steps: A1. Obtain the readout signals of each of the detector modules (2); A2. Calculate the detection location information and initial energy information of the gamma event based on the four signals in the readout signal; A3. The initial energy information is calibrated using a piecewise polynomial calibration algorithm to obtain the calibrated true energy value; A4. Based on the detection location information and the true energy value, an energy weight iterative reconstruction algorithm is used to reconstruct the image and generate a gamma-ray image.

8. The low-cost, high dynamic range gamma imaging method according to claim 7, characterized in that, Step A2 includes: A201. Based on the four signals in the readout signal, the detection location information of the gamma event is calculated using the centroid positioning algorithm; A202. Calculate the sum of the energies corresponding to the four signals in the readout signal to obtain the initial energy information of the gamma event.

9. A low-cost, high dynamic range gamma imaging method according to claim 7, characterized in that, Step A3 includes: A301. Compare the initial energy information with the preset energy threshold; A302. If the initial energy information is not greater than the preset energy threshold, the initial energy information is calibrated according to the pre-calibrated linear mapping function to obtain the calibrated true energy value; A303. If the initial energy information is greater than the preset energy threshold, the initial energy information is calibrated according to the pre-calibrated higher-order polynomial mapping function to obtain the calibrated true energy value; the linear mapping function and the higher-order polynomial mapping function smoothly transition at the preset energy threshold.

10. A low-cost, high dynamic range gamma imaging method according to claim 7, characterized in that, Step A4 includes: A401. For each gamma event, calculate the confidence weight of the gamma event based on the deviation between the actual energy value of the gamma event and the preset photoelectric peak energy; A402. Based on the array geometric parameters of the pixel units of the detector module (2) and the detection position information of each gamma event, establish a system response matrix and initialize the image matrix to be reconstructed; each element of the system response matrix is ​​the response probability of each image pixel to each gamma event detection position, and each element of the image matrix to be reconstructed is the image pixel value of the image to be reconstructed. A403. Based on the confidence weight and the system response matrix, iteratively reconstruct the image matrix to be reconstructed using the OSEM algorithm to update the image pixel values ​​in the image matrix to be reconstructed; A404. Repeat step A403 until the preset iteration termination condition is met, and convert the final image matrix to be reconstructed into a gamma-ray image.

Images