A high-precision detection method and system for ionization chambers based on gas detectors

Abstract

This invention relates to the field of radiation detection technology, specifically to a high-precision detection method and system for ionization chambers based on gas detectors. The method determines a spatial distribution basis matrix by analyzing the correlation of current values ​​in different ionization chamber units, and extracts the spatial linkage characteristics of the detector array during movement. It utilizes the local mapping characteristics of non-radiative zone units in the basis matrix and introduces a regularization term to determine the derivation matrix. Simultaneously, it determines an edge reference signal vector based on the historical relative deviation of the original current in the non-radiative zone units, and couples this vector with the spatial distribution basis matrix and the derivation matrix to determine the overall array interference estimation vector. Finally, it uses the relative deviation between the original total current vector and the overall array interference estimation vector to determine the dose response signal vector, accurately identifying the dynamic mechanical interference current that fluctuates in real time with movement. This results in higher accuracy in detecting radiation dose distribution based on the dose response signal vector.

Description

Technical Field

[0001] This invention relates to the field of radiation detection technology, specifically to a high-precision detection method and system for an ionization chamber based on a gas detector. Background Technology

[0002] During the acceptance and quality control process of radiotherapy equipment, a three-dimensional water tank scanning system drives an ionization chamber detector array to move in the water to collect radiation dose distribution data at different locations. Existing technologies typically use smoothing filtering algorithms to denoise the collected current signals and use the obtained ionization radiation intensity values ​​to detect radiation dose distribution.

[0003] However, existing technologies neglect the mechanical deformation interference caused by fluid resistance during dynamic scanning of the detector circuit board, and fail to consider the spatial linkage characteristics of the circuit board as a continuous elastic body. As a result, they cannot accurately extract the dynamic mechanical interference current that fluctuates in real time with the scanning speed from the complex mixed signal, leading to low accuracy in detecting low-dose regions and high gradient dose distribution regions (such as Bragg peaks). In other words, the accuracy of radiation dose distribution detection is low when existing technologies use ionizing radiation dose values ​​obtained by denoising the acquired current signal through smoothing filtering algorithms. Summary of the Invention

[0004] To address the issue of low accuracy in radiation dose distribution detection using existing technologies that rely on denoising algorithms to process acquired current signals, the present invention aims to provide a high-precision detection method and system for ionization chambers based on gas detectors. The specific technical solution adopted is as follows: The first aspect of this invention provides a high-precision detection method for an ionization chamber based on a gas detector, comprising: The original current value of each ionization chamber unit on the detector array in motion trajectory at each sampling time is obtained by a three-dimensional water tank scanning system; wherein, each ionization chamber unit includes at least two non-radiative zone units; Based on the temporal distribution correlation of the original current values ​​among different ionization chamber units, the spatial distribution basis matrix is ​​determined; based on the local mapping characteristics of each non-radiative region unit in the spatial distribution basis matrix, a regularization term is introduced to perform inverse mapping calculation to determine the derivation matrix. Based on the overall magnitude of the original current values ​​of each ionization chamber unit at each sampling time, the original total current vector is determined; based on the historical relative deviation of the original current corresponding to each non-radiation zone unit at each sampling time, the edge reference signal vector is determined; based on the coupling of the spatial distribution basis matrix, the inference matrix, and the edge reference signal vector, the full array interference estimation vector is determined. Based on the relative deviation between the original total current vector and the full array interference estimation vector, the dose response signal vector at each sampling time is determined; and the radiation dose distribution is detected based on the dose response signal vector.

[0005] Furthermore, the process of obtaining the spatial distribution basis matrix includes: Arrange the original current values ​​of each ionization chamber unit at all sampling times in chronological order to determine the original current time sequence. Construct a spatial correlation matrix, where the rows of the spatial correlation matrix represent each ionization chamber unit, the columns of the spatial correlation matrix represent each ionization chamber unit, and each element in the spatial correlation matrix is ​​the Pearson correlation coefficient between the original current time series of the ionization chamber unit in the corresponding row and the original current time series of the corresponding column. The spatial correlation matrix is ​​decomposed into eigenvalues ​​to obtain the eigenvectors corresponding to each eigenvalue; all eigenvalues ​​are arranged in descending order to determine the corresponding eigenvalue sequence; the eigenvectors corresponding to the first preset number of eigenvalues ​​in the eigenvalue sequence are arranged in columns to construct the spatial distribution basis matrix.

[0006] Furthermore, the process of obtaining the deduction matrix includes: Arrange the row vectors of the corresponding rows of all non-radiative region units in the spatial distribution basis matrix by row to construct a partial observation matrix; The transpose of the partial observation matrix is ​​multiplied with the partial observation matrix to determine the partial observation reference matrix; the regularization compensation matrix is ​​determined based on the product of the preset regularization parameter and the identity matrix; the dimension of the identity matrix is ​​the same as the dimension of the partial observation reference matrix. The partial observation reference matrix and the regularization compensation matrix are added together and then inverted to determine the corrected inverse matrix; the corrected inverse matrix is ​​multiplied by the transpose of the partial observation matrix to determine the derivation matrix.

[0007] Furthermore, the process of obtaining the original total current vector includes: The original current values ​​of all ionization chamber units at each sampling time are arranged sequentially according to the ionization chamber unit index to determine the original total current vector.

[0008] Furthermore, the process of obtaining the edge reference signal vector includes: The average of the original current values ​​of each non-radiative region element at a predetermined second number of sampling times before each sampling time is used as the historical current reference value of each non-radiative region element at each sampling time. The corresponding edge reference signal value is determined based on the difference between the original current value of each non-radiative region element at each sampling time and the corresponding historical current reference value. The edge reference signal values ​​of all non-radiative region elements at each sampling time are arranged sequentially according to the ionization chamber element index to determine the edge reference signal vector.

[0009] Furthermore, the process of obtaining the full array interference estimation vector includes: The inference matrix is ​​multiplied by the edge reference signal vector to determine the feature space coefficient vector; the spatial distribution basis matrix is ​​multiplied by the feature space coefficient vector to determine the full array interference estimation vector.

[0010] Furthermore, the process of obtaining the dose response signal vector includes: The dose response signal vector is determined by subtracting the full array interference estimation vector from the original total current vector.

[0011] Furthermore, the process of detecting radiation dose distribution based on the dose response signal vector includes: At each sampling time, the ionizing radiation dose value of each ionization chamber unit is determined based on the dose response signal; the radiation dose distribution is detected based on the ionizing radiation dose value.

[0012] Furthermore, the preset regularization parameter is set to .

[0013] Secondly, the present invention provides a high-precision detection system for an ionization chamber based on a gas detector, the system comprising: The data acquisition and preprocessing module is used to acquire the raw current values ​​of each ionization chamber unit on the detector array in motion trajectory at each sampling time through a three-dimensional water tank scanning system; wherein, each ionization chamber unit includes at least two non-radiative zone units; The first determining module is used to determine the spatial distribution basis matrix based on the temporal distribution correlation of the original current values ​​between different ionization chamber units; and to determine the derivation matrix by introducing a regularization term and performing inverse mapping calculation based on the local mapping characteristics of each non-radiation zone unit in the spatial distribution basis matrix. The second determining module is used to determine the original total current vector based on the overall magnitude of the original current values ​​of each ionization chamber unit at each sampling time; to determine the edge reference signal vector based on the historical relative deviation of the original current corresponding to each non-radiation zone unit at each sampling time; and to determine the full array interference estimation vector by coupling the spatial distribution basis matrix, the inference matrix and the edge reference signal vector. The radiation detection module is used to determine the dose response signal vector at each sampling time based on the relative deviation between the original total current vector and the full array interference estimation vector; and to perform radiation dose distribution detection based on the dose response signal vector.

[0014] Thirdly, the present invention provides a computer device including a memory and a processor. The memory is used to store computer program code, and the processor is used to call and run the computer program code from the memory to perform the method as described in the first aspect or any embodiment of the first aspect of the present invention.

[0015] Fourthly, the present invention provides a computer program product comprising computer program code, which, when executed, performs the method as described in the first aspect or any embodiment of the first aspect of the present invention.

[0016] Fifthly, the present invention provides a computer-readable storage medium storing computer program code that, when executed, performs the method as described in the first aspect or any embodiment of the first aspect of the present invention.

[0017] This application has the following beneficial effects: This application determines the spatial distribution basis matrix by analyzing the temporal distribution correlation of the original current values ​​between different ionization chamber units, effectively extracting the spatial linkage characteristics of the detector array during movement. It utilizes the local mapping characteristics of the non-radiative zone units in the basis matrix and introduces a regularization term to determine the derivation matrix, constructing a stable and reliable inverse mapping relationship from local to global. Simultaneously, it determines the edge reference signal vector based on the historical relative deviation of the original current of the non-radiative zone units, coupling it with the spatial distribution basis matrix and the derivation matrix to determine the full array interference estimation vector, achieving accurate simulation and restoration of the dynamic mechanical interference characteristics of the entire array. Finally, it uses the relative deviation between the original total current vector and the full array interference estimation vector to determine the dose response signal vector, accurately extracting the dynamic mechanical interference current that fluctuates in real time with movement from the complex original mixed signal. This effectively avoids the erroneous damage to the real signal by traditional smoothing filtering algorithms, making the obtained dose response signal purer and more realistic, thereby improving the accuracy of radiation dose distribution detection based on the dose response signal vector. Attached Figure Description

[0018] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating a high-precision detection method for an ionization chamber based on a gas detector, provided in one embodiment of the present invention; Figure 2 This is a structural diagram of a high-precision detection system for an ionization chamber based on a gas detector, provided in one embodiment of the present invention. Figure 3 This is a schematic diagram of a computer device structure provided in one embodiment of the present invention. Detailed Implementation

[0020] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a high-precision detection method and system for an ionization chamber based on a gas detector proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment, and specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0022] The following detailed description, in conjunction with the accompanying drawings, illustrates the specific scheme of a high-precision detection method and system for an ionization chamber based on a gas detector provided by this invention.

[0023] This invention provides a high-precision detection method for ionization chambers based on gas detectors. Please refer to [link to relevant documentation]. Figure 1 The diagram illustrates a flowchart of a high-precision detection method for an ionization chamber based on a gas detector, according to an embodiment of the present invention. The method includes: Step S101: Obtain the original current value of each ionization chamber unit on the detector array in motion trajectory at each sampling time through a three-dimensional water tank scanning system; wherein, all ionization chamber units include at least two non-radiative zone units.

[0024] This invention first uses a three-dimensional water tank scanning system to acquire the raw current values ​​of each ionization chamber unit on the detector array at each sampling time, within its moving trajectory; wherein all ionization chamber units include at least two non-radiation zone units. In routine dose verification scenarios for radiotherapy equipment (such as medical linear accelerators or proton and heavy ion therapy systems), the three-dimensional water tank scanning system drives the detector array, carrying high-density ionization chamber units, to dynamically move along a planned measurement trajectory inside a water-filled tank. During this dynamic scanning process, the detector array and its circuit board (PCB) are continuously subjected to the impact of water flow eddies and fluid resistance, resulting in minute mechanical deformation. To accurately capture the noise fluctuations caused by this deformation, the hardware layout of the detector array specifically includes at least two non-radiation zone units. These non-radiative region units are typically physically distributed at the edges or corners of the array, and their surfaces are covered with high atomic number shielding materials. Alternatively, they are strictly confined to outside the effective radiation field (such as outside the penumbra of the photon beam) during scanning path planning to ensure that their output signal reflects only pure mechanical vibration and electronic noise, without any real radiation dose signal. In one specific implementation of this invention, the ionization chamber units located at the four vertices of the detector array are used as non-radiative region units, which can be adjusted according to the specific implementation environment.

[0025] During the specific acquisition of raw current values, as the detector array moves along its trajectory in the water, the synchronous multi-channel electrometer at the back end continuously acquires signals from all ionization chamber units (including radiative and non-radiative units) on the array. Since the induced current in the ionization chamber is in the pA range, direct measurement is susceptible to noise interference. Therefore, this embodiment of the invention employs a charge integration method to improve sensitivity. Specifically, firstly, the accumulated induced charge within each sampling period is counted, and the end of the sampling period is taken as the corresponding sampling time. At each sampling time, the accumulated charge within the corresponding sampling period is read, and the ratio between the accumulated charge of the corresponding sampling period and the sampling period duration is calculated as the raw current value for each sampling time.

[0026] In one specific implementation of this invention, the preset scanning speed of the three-dimensional water tank scanning system driving the detector array is set between 5 mm / s and 20 mm / s to match the standard motion conditions for obtaining percentage depth dose curves or field off-axis ratio curves in actual clinical quality control. This speed can be adjusted according to the specific implementation environment. In this embodiment, it is set to 10 mm / s. The preset sampling frequency of the multi-channel electrometer for charge integration is set between 100 Hz and 1000 Hz, corresponding to a sampling period of 1 ms to 10 ms. This frequency can be adjusted according to the specific implementation environment. In this embodiment, the preset sampling frequency is set to 500 Hz, corresponding to a sampling period of 2 ms. This high-frequency sampling parameter ensures that the system can fully capture the high-frequency mechanical vibration current fluctuations caused by fluid turbulence and avoid spectral aliasing of interference features in the time domain. At the same time, in conjunction with the preset scanning speed, this sampling frequency ensures that the spatial sampling step size is much smaller than the physical size of the ionization chamber unit, thereby providing a sufficient density of spatiotemporal samples for constructing the spatial distribution basis matrix and ensuring real-time tracking and high-precision stripping of dynamic interference in complex flow field environments. It should be noted that the sampling time window in this embodiment of the invention is set to 5 to 30 seconds (preferably 10 seconds in this embodiment of the invention), which can be adjusted according to the stability of the flow field disturbance and the real-time requirements of the calculation in the specific implementation environment. For sampling moments where the number of previous sampling moments does not meet the sampling time window, the corresponding sampling moment and the whole of all previous sampling moments are used as the corresponding sampling time window for analysis, which will not be elaborated further here.

[0027] Step S102: Determine the spatial distribution basis matrix based on the temporal distribution correlation of the original current values ​​between different ionization chamber units; based on the local mapping characteristics of each non-radiation region unit in the spatial distribution basis matrix, introduce a regularization term to perform inverse mapping calculation and determine the derivation matrix.

[0028] After acquiring the raw current values, mixed time-series data containing radiation dose and mechanical interference were obtained. Considering that the detector circuit board, as a continuous elastic body, undergoes mechanical deformation due to water flow resistance during dynamic scanning, this deformation is not isolated but exhibits significant spatial linkage characteristics following specific mechanical laws. By analyzing the temporal distribution correlation of raw current values ​​between different ionization chamber units and determining the spatial distribution basis matrix, this physical spatial linkage law is quantified into a feature space describing the main interference mode. This spatial distribution basis matrix provides a reliable inverse mapping model for subsequently extrapolating the global interference distribution of the entire array using observation data from local non-radiative zone units. Combined with a low-frequency constraint mechanism in the spatial dimension, it enables accurate simulation and reconstruction of dynamic interference characteristics, effectively avoiding the false damage to high-gradient dose distribution details caused by traditional filtering algorithms, thereby improving the accuracy and fidelity of radiotherapy dose detection.

[0029] After determining the spatial distribution basis matrix that characterizes the mechanical linkage law of the entire array, the system obtains the basis vector set describing the typical deformation mode of the detector array. In actual radiotherapy scanning, the signal of the radiated zone element is simultaneously coupled with the actual dose and mechanical interference, while the physically shielded non-radiated zone element only reflects the pure dynamic vibration state. By analyzing the local mapping characteristics of each non-radiated zone element in the spatial distribution basis matrix and introducing a regularization term for inverse mapping calculation to determine the inference matrix, a linear correlation operator from sparse edge observation data to the global feature space can be constructed. The determination of this inference matrix eliminates the numerical instability caused by uneven distribution of observation points and multicollinearity at the algorithmic level, ensuring the numerical robustness and reliability of the subsequent full-array interference estimation process. Its implementation effect lies in the ability to reproduce the mechanical interference field covering the entire field and possessing physical consistency in real time and robustly using extremely small amounts of edge reference information, laying the foundation for high-fidelity dose stripping from the original mixed current.

[0030] Step S103: Determine the original total current vector based on the overall magnitude of the original current values ​​of each ionization chamber unit at each sampling time; determine the edge reference signal vector based on the historical relative deviation of the original current corresponding to each non-radiation zone unit at each sampling time; determine the full array interference estimation vector by coupling the spatial distribution basis matrix, the derivation matrix and the edge reference signal vector.

[0031] After completing the feature space modeling and pre-calculation of the extrapolation matrix, this embodiment of the invention enters the real-time measurement stage. The original current values ​​of each ionization chamber unit at each sampling time are synchronously summarized to determine the original total current vector. This vector records the full-field response intensity of the detector array in the current time slice, constituting the original observation source to be decoupled. This processing achieves spatial alignment and vectorized expression of multi-channel physical signals, ensuring that subsequent matrix operations can be performed under a unified mathematical framework; and provides a complete data carrier for real-time removal of full-field interference, ensuring the spatiotemporal consistency of radiation dose detection during dynamic scanning, thereby providing an accurate reference for reconstructing the true dose field details.

[0032] This invention further extracts the real-time observation signal of the non-radiative region unit at each sampling moment, i.e., the edge reference signal vector. Given the slight differences in sensor manufacturing processes and the low-frequency background drift caused by scanning ambient temperature, by calculating the historical relative deviation of the original current corresponding to each non-radiative region unit at each sampling moment and determining the edge reference signal vector, background noise such as DC leakage current can be effectively filtered out from the original observation values, thereby extracting the dynamic vibration characteristics purely excited by the current mechanical deformation.

[0033] Since the edge reference signal vector only reflects the local physical vibration state of the detector array, in order to quantify the global background noise, the system determines the full array interference estimation vector by coupling the spatial distribution basis matrix, the inference matrix, and the edge reference signal vector. This coupled calculation process accurately maps local edge fluctuations to weight coefficients in the feature space through the inference matrix, and then expands and reconstructs them by the spatial distribution basis matrix to form an interference distribution model covering the entire detector array. This process utilizes the unique low-frequency smoothing constraint mechanism of the basis matrix, naturally excluding high-frequency local features in the synthesized full array interference estimation vector. This ensures that it only contains overall mechanical interference conforming to the physical and mechanical laws of the circuit board, achieving accurate and real-time simulation of dynamic fluid interference across the entire array. This provides a highly reliable calculation benchmark for the subsequent lossless removal of background noise from the original mixed signal, thereby minimizing misjudgment and wear of the true radiation field characteristics.

[0034] Step S104: Determine the dose response signal vector at each sampling time based on the relative deviation between the original total current vector and the full array interference estimation vector; detect the radiation dose distribution based on the dose response signal vector.

[0035] After successfully constructing a full-array interference estimation vector that closely matches the current physical deformation state, the system has the prerequisites to quantify and separate background noise. Given that the original total current vector simultaneously couples the actual radiation field excitation signal and hydrodynamic interference, to restore the pure spatial distribution information of the radiation field, the system determines the dose response signal vector at each sampling moment based on the relative deviation between the original total current vector and the full-array interference estimation vector. This calculation process precisely decouples the mixed signal point-by-point along the physical dimension, accurately subtracting the dynamic baseline drift caused by water flow resistance and mechanical vibration. This processing step directly achieves the core objective of dynamic interference separation in this application. Its beneficial effects are that it not only significantly improves the measurement signal-to-noise ratio of the system in weak dose regions (such as the beam tail or penumbra), but also, thanks to the low-frequency smoothing constraint mechanism established by the previous algorithm, this stripping operation perfectly preserves the high-frequency spatial gradient details in the real dose field, thereby greatly improving the detection accuracy and high fidelity of three-dimensional water tank scanning dose verification for radiotherapy.

[0036] Since the dynamic mechanical interference current has been successfully extracted from the dose response signal vector, in order to verify the dose of radiotherapy, this embodiment of the invention finally performs radiation dose distribution detection based on the dose response signal vector. This detection process accurately reconstructs the radiation field intensity distribution on the scanning trajectory by spatiotemporally correlating the pure dose current with the coordinate position, ensuring the authenticity and high signal-to-noise ratio of the characteristics of high gradient dose regions (such as Bragg peaks).

[0037] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the spatial distribution basis matrix includes: The original current values ​​of each ionization chamber unit at all sampling times are arranged in chronological order to determine the original current time series. Through this vectorized arrangement, the system completely preserves the synchronous noise trajectory of the detector array within the sampling time window, providing the necessary time-dimensional data carrier for subsequent cross-channel correlation modeling. These time series not only record the disturbance amplitude of a single unit caused by fluid impact, but also, through their relative phase relationships, implicitly reveal the temporal evolution law of the detector array as a physical whole during the scanning process.

[0038] A spatial correlation matrix is ​​constructed, where each row and column represents an ionization chamber unit. Each element in the spatial correlation matrix is ​​the Pearson correlation coefficient between the original current time series of the ionization chamber unit in the corresponding row and the original current time series in the corresponding column. This process uses the Pearson correlation coefficient to quantify the statistical correlation strength of the stress deformation between observation points of the detector circuit board, which is a continuous elastic body. The magnitude of the Pearson correlation coefficient directly reflects the tightness of the mechanical linkage between two detector units, thus constructing a correlation topology network describing the mechanical linkage characteristics of the entire array. This provides a data foundation for subsequently converting isolated readings from discrete sensors into a spatial field with structural constraints. It should be noted that the row and column order of the spatial correlation matrix in this embodiment corresponds to the ascending order of the ionization chamber unit indices. The index values ​​of the ionization chamber units are determined a priori by the geometric arrangement sequence of each ionization chamber unit in the detector array, which will not be further elaborated here.

[0039] Eigenvalue decomposition is performed on the spatial correlation matrix to obtain the eigenvector corresponding to each eigenvalue. All eigenvalues ​​are arranged in descending order to determine the corresponding eigenvalue sequence. The eigenvectors corresponding to the first preset number of eigenvalues ​​in the eigenvalue sequence are arranged column-wise to construct the spatial distribution basis matrix. Utilizing the principle of principal component extraction through eigenvalue decomposition, the complex background fluctuations are deconstructed into a low-dimensional orthogonal basis describing the typical deformation modes of the plate (such as overall bending, torsion, or offset). Larger eigenvalues ​​represent the energy proportion of the corresponding deformation mode in the overall dynamic interference, while the first preset number of eigenvectors are retained, which function as a "low-frequency spatial filter" for the interference signal. This ensures that when reconstructing the full-field interference, the system can effectively remove random high-frequency electronic noise without spatial linkage characteristics through low-rank constraints. Thus, while ensuring the accuracy of mechanical interference reconstruction, it minimizes the damage to the true dose gradient distribution with high-frequency spatial characteristics (such as Bragg peaks), achieving high-fidelity signal stripping in physical logic.

[0040] In one specific implementation of this invention, the preset first quantity ranges from 3 to 10 and can be adjusted according to the specific implementation environment. In this embodiment, it is set to 5 to fully cover the main fluid deformation modes of the detector array and effectively filter high-frequency unstructured noise interference. Furthermore, the preset first quantity in this embodiment is less than or equal to the number of non-radiative region units to avoid rank deficiency caused by subsequent matrix multiplication. This will not be elaborated further here.

[0041] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the deduction matrix includes: Arrange the row vectors of the corresponding rows in the spatial distribution basis matrix of all non-radiative units to construct a partial observation matrix; multiply the transpose of the partial observation matrix with the partial observation matrix to determine the partial observation reference matrix; determine the regularization compensation matrix based on the product of the preset regularization parameter and the identity matrix; the dimension of the identity matrix is ​​the same as the dimension of the partial observation reference matrix; add the partial observation reference matrix and the regularization compensation matrix and perform matrix inversion to determine the corrected inverse matrix; multiply the corrected inverse matrix with the transpose of the partial observation matrix to determine the derivation matrix.

[0042] By extracting the row vectors corresponding to all non-radiative region units and constructing a partial observation matrix, the system establishes the local physical representation of the global spatial master mode at the array edge, and establishes a mapping bridge between local observations and global deformation features. The calculated partial observation reference matrix characterizes the information overlap and autocorrelation of local observation data in the feature space. However, given the limited number of non-radiative region units and the potential clustering effect at their physical locations, multicollinearity is easily generated among local observation data, causing the partial observation reference matrix to approach a singular state (i.e., ill-conditioned rank deficiency). Directly inverting it can lead to severe numerical instability or even computational deadlock. To address this computational challenge, the system introduces a Tikhonov regularization mechanism. By constructing a regularization compensation matrix and adding it to the partial observation reference matrix, a small numerical penalty term is essentially applied to the main diagonal of the ill-conditioned matrix. This operation forcibly raises the lower bound of the matrix's eigenvalues, ensuring that the corrected inverse matrix obtained during subsequent matrix inversion operations has extremely high numerical stability. The final inference matrix, as a robust linear inverse mapping operator, is related to the overall objective by effectively overcoming the risk of divergence in solving underdetermined equations caused by a small number of edge observations. It endows the algorithm with strong robustness in the face of different water flow impact environments, ensuring that the system can accurately and without deadlock inverse the mechanical interference coefficient of the entire array of the covering detector with only a very small number of non-radiative vibration signals.

[0043] It should be noted that, according to the calculation process of the spatial distribution basis matrix, each row of the matrix corresponds to an ionization chamber unit, and they are arranged in order of index value. Based on the index value of the non-radiative unit in all ionization chamber units, its corresponding row in the spatial distribution basis matrix can be determined.

[0044] In one specific implementation of this invention, the process of obtaining the deduction matrix is ​​expressed by the following formula: ;in, For the derivation matrix; This is the transpose of a portion of the observation matrix; This is a partial observation matrix; It is the identity matrix; Preset regularization parameters; This is the regularization compensation matrix; To correct the inverse matrix.

[0045] Preset regularization parameters in the embodiments of the present invention The value range is set to to The value can be adjusted according to the specific implementation environment. In the embodiments of the present invention, the preferred value is... This parameter, acting as a regularization compensation term, provides a small support for the main diagonal of the matrix during inversion operations. By... Within the aforementioned range, it can be ensured that the corrected inverse matrix remains invertible even under extreme conditions where the linear correlation of the edge reference signal is high (tending towards singularity), thus avoiding computational deadlock or numerical runaway. Implementers can fine-tune within this range based on the specific sensitivity range of the detector's ionization chamber and the ambient background noise level to achieve the optimal balance between reconstruction accuracy and numerical stability.

[0046] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the original total current vector includes: The original current values ​​of all ionization chamber units at each sampling time are arranged sequentially according to their ionization chamber unit indices to determine the original total current vector. By vectorizing the current values ​​of all detector units at each sampling time according to their index order, bidirectional alignment of the entire array of detector units in both time and space is achieved. This ensures that the original observation data at each sampling time strictly corresponds to the subsequently reconstructed full array interference estimation vector in terms of dimension and physical location. This vector, as the original mixed observation source containing radiation dose signal and mechanical deformation noise, establishes the physical benchmark for interference stripping operations, ensuring point-to-point consistency during subtraction correction. This effectively prevents dose distribution distortion or artifacts caused by data misalignment, ensuring the authenticity of the final output dose response signal in its spatiotemporal distribution.

[0047] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the edge reference signal vector includes: The average of the original current values ​​of each non-radiative region element at a predetermined second number of sampling times before each sampling time is used as the historical current reference value of each non-radiative region element at each sampling time. The corresponding edge reference signal value is determined based on the difference between the original current value of each non-radiative region element at each sampling time and the corresponding historical current reference value. The edge reference signal values ​​of all non-radiative region elements at each sampling time are arranged sequentially according to the ionization chamber element index to determine the edge reference signal vector.

[0048] This processing step utilizes a moving average filtering algorithm to establish a dynamic DC bias reference for each non-radiative region cell. Because the detector's output signal often contains slowly drifting low-frequency components due to ambient temperature fluctuations and electronic leakage current during long-term scanning, the system effectively deconstructs the original current value into two parts: a static background and dynamic fluctuations. This eliminates baseline deviations caused by individual sensor differences and ambient temperature drift, preventing these non-deformation factors from being incorrectly projected and amplified into the full-array interference model by the extrapolation matrix. The resulting edge reference signal vector ensures that the physical signal of the input model is always a pure dynamic component relative to zero bias, significantly enhancing the algorithm's sensitivity to dynamic mechanical interference and providing an accurate reference for achieving high-fidelity, drift-free radiation dose detection.

[0049] It should be noted that the preset second quantity in this embodiment of the invention is set to 500 to 1500 (corresponding to a 500 Hz sampling frequency of 1 to 3 seconds), which can be adjusted according to the specific implementation environment. A preferred setting is 1000, aiming to balance the ability to suppress temperature drift with the degree of preservation of rapid mechanical disturbances. For sampling moments with a number of previous sampling moments lower than the preset second quantity, the average of the original current values ​​of that sampling moment and all previous sampling moments is used as the historical current reference value for analysis to avoid situations where calculation is impossible. Through this dynamic reference tracking mechanism, the system does not need to manually calibrate the zero point before each measurement, which will not be further elaborated here.

[0050] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the full array interference estimation vector includes: The inference matrix and the edge reference signal vector are multiplied to determine the characteristic space coefficient vector; the spatial distribution basis matrix and the characteristic space coefficient vector are multiplied to determine the full array interference estimation vector.

[0051] By linearly multiplying the derivation matrix with the edge reference signal vector, the physical meaning lies in completing a mapping from the physical observation space to the mathematical characteristic space. In this process, the derivation matrix acts as a specific spatiotemporal decoder, transforming the discrete, local edge reference signal vector into a characteristic space coefficient vector describing the weight proportions of each typical deformation mode (such as overall bending and torsion) of the detector circuit board at the current moment. Subsequently, a basis vector linear superposition operation is performed on these coefficient vectors using the spatially distributed basis matrix. The purpose is to reconstruct the complete dynamic interference field at all ionization chamber unit locations of the entire array using pre-calibrated spatial linkage rules. The functional role of this reconstruction process is to achieve "holographic completion" of the mechanical interference signal, filling the information gap in the central radiation field region of the array where background noise cannot be directly observed. The full array interference estimation vector, as the final generated noise distribution model, is naturally constrained within a smooth deformation range that conforms to the physical and mechanical properties of the plate due to the fact that its reconstruction process is limited by the low-frequency, low-rank spatial distribution basis matrix. This achieves the essential decoupling of mechanical interference (low-frequency spatial distribution) and true dose distribution (high-frequency spatial gradient) in terms of physical logic, ensuring that subsequent stripping operations can accurately locate background noise without accidentally damaging key dose features such as the Bragg peak.

[0052] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the dose response signal vector includes: The dose response signal vector is determined by subtracting the full array interference estimation vector from the original total current vector. Through vector subtraction, this embodiment of the invention can accurately isolate the interference components at each ionization chamber unit location. While filtering out the spatially gradual mechanical background noise, it perfectly preserves the dose gradient details with high-frequency spatial characteristics in the original total current vector, achieving point-by-point decoupling of the radiation dose response charge and the mechanical deformation induced current in the physical dimension. This process restores the pure ionizing radiation response from the mixed signal, solving the problem of Bragg peak clipping or penumbra blurring caused by traditional smoothing filtering, and ensuring that the final output dose response signal vector has extremely high fidelity.

[0053] Preferably, in some possible implementations of the embodiments of the present invention, the process of detecting radiation dose distribution based on the dose response signal vector includes: At each sampling time, the ionizing radiation dose value of each ionization chamber unit is determined based on the dose response signal; radiation dose distribution is then detected based on the ionizing radiation dose value. After acquiring the clean dose response signal vector, a quantitative traceability process from electrical quantities to physical dose values ​​is further performed. Using a pre-calibrated dose scaling factor, the current value of each ionization chamber unit is converted into a standard ionizing radiation dose value, achieving the physical quantity conversion required for clinical quality control. Subsequently, the ionizing radiation dose value at each sampling time is precisely spatiotemporally correlated with the real-time spatial coordinates of the three-dimensional water tank scanning system, thereby reconstructing the radiation dose distribution curve (such as PDD or OAR curves) that varies with depth or radial direction. This process transforms discrete detection data into an intuitive dose distribution model, resulting in higher accuracy in radiation dose distribution detection.

[0054] In summary, a high-precision detection method for ionization chambers based on gas detectors determines the spatial distribution basis matrix by analyzing the temporal distribution correlation of the original current values ​​between different ionization chamber units, effectively extracting the spatial linkage characteristics of the detector array during movement. It utilizes the local mapping characteristics of the non-radiative zone units in the basis matrix and introduces a regularization term to determine the derivation matrix, constructing a stable and reliable inverse mapping relationship from local to global. Simultaneously, it determines the edge reference signal vector based on the historical relative deviation of the original current of the non-radiative zone units, coupling it with the spatial distribution basis matrix and the derivation matrix to determine the full array interference estimation vector, achieving accurate simulation and restoration of the dynamic mechanical interference characteristics of the entire array. Finally, it uses the relative deviation between the original total current vector and the full array interference estimation vector to determine the dose response signal vector, accurately extracting the dynamic mechanical interference current that fluctuates in real time with movement from the complex original mixed signal. This effectively avoids the erroneous damage to the real signal caused by traditional smoothing filtering algorithms, making the obtained dose response signal purer and more realistic, thus improving the accuracy of radiation dose distribution detection based on the dose response signal vector.

[0055] This invention also provides a high-precision detection system for an ionization chamber based on a gas detector; please refer to [link to relevant documentation]. Figure 2 The diagram shows a structural diagram of a high-precision detection system for an ionization chamber based on a gas detector according to an embodiment of the present invention. The system includes: a data acquisition and preprocessing module 201, a first determination module 202, a second determination module 203, and a radiation detection module 204.

[0056] The data acquisition and preprocessing module 201 is used to acquire the raw current value of each ionization chamber unit on the detector array in motion trajectory at each sampling time through the three-dimensional water tank scanning system; wherein, all ionization chamber units include at least two non-radiative zone units; The first determining module 202 is used to determine the spatial distribution basis matrix based on the temporal distribution correlation of the original current values ​​between different ionization chamber units; and to determine the derivation matrix by introducing a regularization term to perform inverse mapping calculation based on the local mapping characteristics of each non-radiation zone unit in the spatial distribution basis matrix. The second determining module 203 is used to determine the original total current vector based on the overall magnitude of the original current values ​​of each ionization chamber unit at each sampling time; to determine the edge reference signal vector based on the historical relative deviation of the original current corresponding to each non-radiation zone unit at each sampling time; and to determine the full array interference estimation vector by coupling the spatial distribution basis matrix, the derivation matrix and the edge reference signal vector. The radiation detection module 204 is used to determine the dose response signal vector at each sampling time based on the relative deviation between the original total current vector and the full array interference estimation vector; and to detect the radiation dose distribution based on the dose response signal vector.

[0057] It should be noted that the system provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the high-precision detection system for ionization chambers based on gas detectors and the high-precision detection method for ionization chambers based on gas detectors provided in the above embodiments belong to the same concept. The specific implementation process is detailed in the method embodiments and will not be repeated here.

[0058] This invention also provides a computer device; please refer to [link / reference]. Figure 3 The diagram illustrates a computer device structure according to an embodiment of the present invention. The computer device includes a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute any of the aforementioned high-precision detection methods for ionization chambers based on gas detectors.

[0059] This invention also provides a computer program product that, when run on a computer device, enables the computer device to execute any of the aforementioned high-precision detection methods for ionization chambers based on gas detectors.

[0060] This invention also provides a computer-readable storage medium storing computer program code. When the computer program code is run on a computer device, the computer device can execute any of the aforementioned high-precision detection methods for ionization chambers based on gas detectors.

[0061] In the embodiments provided by the present invention, it should be understood that the computer device, computer program product and computer-readable storage medium provided are all used to execute the corresponding methods provided above, and therefore the beneficial effects they can achieve can be referred to the beneficial effects of the methods provided above, which will not be repeated here.

[0062] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0063] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A high-precision detection method based on a gas detector ionization chamber, characterized in that, The method includes: The original current value of each ionization chamber unit on the detector array in motion trajectory at each sampling time is obtained by a three-dimensional water tank scanning system; wherein, each ionization chamber unit includes at least two non-radiative zone units; Based on the temporal distribution correlation of the original current values ​​among different ionization chamber units, the spatial distribution basis matrix is ​​determined; based on the local mapping characteristics of each non-radiative region unit in the spatial distribution basis matrix, a regularization term is introduced to perform inverse mapping calculation to determine the derivation matrix. Based on the overall magnitude of the original current values ​​of each ionization chamber unit at each sampling time, the original total current vector is determined; based on the historical relative deviation of the original current corresponding to each non-radiation zone unit at each sampling time, the edge reference signal vector is determined; based on the coupling of the spatial distribution basis matrix, the inference matrix, and the edge reference signal vector, the full array interference estimation vector is determined. Based on the relative deviation between the original total current vector and the full array interference estimation vector, the dose response signal vector at each sampling time is determined; and the radiation dose distribution is detected based on the dose response signal vector.

2. The high-precision detection method of an ionization chamber based on a gas detector according to claim 1, characterized in that, The process of obtaining the spatial distribution basis matrix includes: Arrange the original current values ​​of each ionization chamber unit at all sampling times in chronological order to determine the original current time sequence. Construct a spatial correlation matrix, where the rows of the spatial correlation matrix represent each ionization chamber unit, the columns of the spatial correlation matrix represent each ionization chamber unit, and each element in the spatial correlation matrix is ​​the Pearson correlation coefficient between the original current time series of the ionization chamber unit in the corresponding row and the original current time series of the corresponding column. The spatial correlation matrix is ​​decomposed into eigenvalues ​​to obtain the eigenvectors corresponding to each eigenvalue; all eigenvalues ​​are arranged in descending order to determine the corresponding eigenvalue sequence; the eigenvectors corresponding to the first preset number of eigenvalues ​​in the eigenvalue sequence are arranged in columns to construct the spatial distribution basis matrix.

3. The high-precision detection method of an ionization chamber based on a gas detector according to claim 1, characterized in that, The process of obtaining the deduction matrix includes: Arrange the row vectors of the corresponding rows of all non-radiative region units in the spatial distribution basis matrix by row to construct a partial observation matrix; The transpose of the partial observation matrix is ​​multiplied with the partial observation matrix to determine the partial observation reference matrix; the regularization compensation matrix is ​​determined based on the product of the preset regularization parameter and the identity matrix; the dimension of the identity matrix is ​​the same as the dimension of the partial observation reference matrix. The partial observation reference matrix and the regularization compensation matrix are added together and then inverted to determine the corrected inverse matrix; the corrected inverse matrix is ​​multiplied by the transpose of the partial observation matrix to determine the derivation matrix.

4. The high-precision detection method of an ionization chamber based on a gas detector according to claim 1, characterized in that, The process of obtaining the original total current vector includes: The original current values ​​of all ionization chamber units at each sampling time are arranged sequentially according to the ionization chamber unit index to determine the original total current vector.

5. The high-precision detection method for an ionization chamber based on a gas detector according to claim 1, characterized in that, The process of obtaining the edge reference signal vector includes: The average of the original current values ​​of each non-radiative region element at a predetermined second number of sampling times before each sampling time is used as the historical current reference value of each non-radiative region element at each sampling time. The corresponding edge reference signal value is determined based on the difference between the original current value of each non-radiative region element at each sampling time and the corresponding historical current reference value. The edge reference signal values ​​of all non-radiative region elements at each sampling time are arranged sequentially according to the ionization chamber element index to determine the edge reference signal vector.

6. The method of claim 1, wherein the ionization chamber is a gas detector. The process of obtaining the full array interference estimation vector includes: ​ The inference matrix is ​​multiplied by the edge reference signal vector to determine the feature space coefficient vector; the spatial distribution basis matrix is ​​multiplied by the feature space coefficient vector to determine the full array interference estimation vector.

7. The high-precision detection method for an ionization chamber based on a gas detector according to claim 1, characterized in that, The process of obtaining the dose response signal vector includes: The dose response signal vector is determined by subtracting the full array interference estimation vector from the original total current vector.

8. The high-precision detection method for an ionization chamber based on a gas detector according to claim 1, characterized in that, The process of detecting radiation dose distribution based on the dose response signal vector includes: At each sampling time, the ionizing radiation dose value of each ionization chamber unit is determined based on the dose response signal; the radiation dose distribution is detected based on the ionizing radiation dose value.

9. A high-precision detection method for an ionization chamber based on a gas detector according to claim 3, characterized in that, The preset regularization parameter is set as follows: .

10. A high-precision detection system for an ionization chamber based on a gas detector, characterized in that, The system includes: The data acquisition and preprocessing module is used to acquire the raw current values ​​of each ionization chamber unit on the detector array in motion trajectory at each sampling time through a three-dimensional water tank scanning system; wherein, each ionization chamber unit includes at least two non-radiative zone units; The first determining module is used to determine the spatial distribution basis matrix based on the temporal distribution correlation of the original current values ​​between different ionization chamber units; and to determine the derivation matrix by introducing a regularization term and performing inverse mapping calculation based on the local mapping characteristics of each non-radiation zone unit in the spatial distribution basis matrix. The second determining module is used to determine the original total current vector based on the overall magnitude of the original current values ​​of each ionization chamber unit at each sampling time; to determine the edge reference signal vector based on the historical relative deviation of the original current corresponding to each non-radiation zone unit at each sampling time; and to determine the full array interference estimation vector by coupling the spatial distribution basis matrix, the inference matrix and the edge reference signal vector. The radiation detection module is used to determine the dose response signal vector at each sampling time based on the relative deviation between the original total current vector and the full array interference estimation vector; and to perform radiation dose distribution detection based on the dose response signal vector.

Images