CN-122017943-A - Training method and device applied to detection result correction model of radiotherapy

CN122017943ACN 122017943 ACN122017943 ACN 122017943ACN-122017943-A

Abstract

The specification discloses a training method and device of a detection result correction model applied to radiotherapy, and relates to the technical field of radiotherapy. The parameters of the detector and the irradiation dose value of each detector after each irradiation can be collected in the process that the plurality of detectors are irradiated by the radiation light for n times respectively. Correction coefficient data is determined based on the plurality of irradiation dose values. And carrying out data preprocessing on the plurality of detector parameters and the dose measured values to determine a training sample. And inputting the correction coefficient data and the training sample into a detection result correction model, and training the model until the loss function of the detection result correction model converges. Therefore, the long-term and cumulative degradation effect caused by the ionizing radiation damage of the detector body after the detector is irradiated for many times is considered, and the training of the detection result correction model is realized.

Inventors

LI HUA
LIU ZHAOXING
LIU XIN
SHI ZHONGYAN
LIANG RUNCHENG
JIN HAIJING
ZHAO RI
LIU LIYE
CHEN FAGUO
YANG BIAO
GUO RONG
ZHANG JING

Assignees

中国辐射防护研究院

Dates

Publication Date: 20260512
Application Date: 20251231

Claims (12)

1. A training method for a detection result correction model applied to radiotherapy, comprising: Collecting detector parameters and radiation dose values after each irradiation in the process of respectively irradiating a plurality of detectors for n times by using the radiation light, wherein the detector parameters comprise breakdown voltage and/or series resistance and/or quenching resistance and/or parasitic resistance and/or pixel capacitive reactance and/or discharge time constant and/or gain and/or dark counting rate and/or photon detection efficiency, and the radiation dose values comprise standard dose values and dose measurement values collected by the detectors; Determining correction coefficient data according to the plurality of irradiation dose values; Performing data preprocessing on the plurality of detector parameters and the dose measurement values to determine a training sample; and inputting the correction coefficient data and the training sample into a detection result correction model, and training the detection result correction model until the loss function of the detection result correction model converges.
2. The method of claim 1, wherein before each of the plurality of detectors is irradiated n times, the method further comprises: for any one detector, determining the maximum radiation dose corresponding to the detector according to the use scene of the detector; And determining the single radiation dose when the detector is irradiated for n times according to the maximum radiation dose corresponding to the detector.
3. The method of training a correction model for detection results of radiation therapy according to claim 1, wherein said determining correction coefficient data based on said plurality of radiation dose values comprises: For any irradiation dose value, determining a correction coefficient value corresponding to the irradiation dose value through the following formula; Wherein, the The correction coefficient value corresponding to the ith detector after being irradiated for the nth time; the corresponding standard dose value of the ith detector after being irradiated by the nth time; dose measurements acquired for the ith detector after being illuminated for the nth time; and determining correction coefficient data according to the plurality of correction coefficient values.
4. The method of claim 1, wherein the step of performing data preprocessing on the plurality of detector parameters and the dose measurement values to determine a training sample comprises: and establishing a correlation between the detector parameters and the measured dose value by a least square method, wherein the correlation is as follows: Wherein, the Is the breakdown voltage; is the series resistance; is the quenching resistor; Is the parasitic resistance; capacitive reactance for the pixel; the discharge time constant is the discharge time constant, G is the gain, DCR is the dark count rate, PDE is the photon detection efficiency, D is a dose measurement value; Solving the related relationship through a target algorithm to determine a feature vector matrix; A training sample is determined based on the feature vector, the plurality of detector parameters, and the dose measurement.
5. The method for training a correction model for detection results for radiation therapy according to claim 4, wherein the number of detectors is i; d is an i×n matrix, and the data at the same position correspond to each other; Solving the correlation relationship through a target algorithm to determine a training sample, wherein the method comprises the following steps: By a feature extraction method Performing dimension reduction, and solving eigenvectors of a covariance matrix of a sample space by adopting a principal component analysis method, wherein the sample space is The s-th row and t-th column elements of the covariance matrix are defined as follows: Wherein, the In order to be a sample space, Any sample is flattened i multiplied by 10 dimension vector; Is covariance matrix; a value representing the f sample in the s-th dimension; Is the average of the samples in the s-th dimension; and carrying out eigenvalue decomposition on the covariance matrix: Wherein, the The system is a diagonal matrix, wherein each diagonal element is a characteristic value of a covariance matrix M and is different from the covariance matrix M, and omega is a characteristic vector matrix; A training sample is determined based on the feature vector, the plurality of detector parameters, and the dose measurement.
6. The method for training the modified model of the detection result for radiation therapy according to claim 5, Can be determined by the following formula: Wherein, the A value representing the f sample in the s-th dimension.
7. A detection result correction method applied to radiation therapy, characterized by using the detection result correction model applied to radiation therapy trained by any one of claims 1 to 6, comprising: Obtaining a dose measured value acquired by a detector, and breakdown voltage, series resistance, quenching resistance, parasitic resistance, pixel capacitance, discharge time constant, gain, dark count rate and photon detection efficiency; And inputting the measured dose value, breakdown voltage, series resistance, quenching resistance, parasitic resistance, pixel capacitive reactance, discharge time constant, gain, dark count rate and photon detection efficiency into the detection result correction model applied to radiotherapy, and determining the corrected dose value output by the detection result correction model applied to radiotherapy.
8. A training device for a detection result correction model for radiation therapy, comprising: The system comprises an acquisition unit, a radiation dose value acquisition unit and a radiation dose detection unit, wherein the acquisition unit is used for acquiring detector parameters and radiation dose values after each irradiation in the process of respectively irradiating a plurality of detectors for n times by using radiation light, the detector parameters comprise breakdown voltage and/or series resistance and/or quenching resistance and/or parasitic resistance and/or pixel capacitive reactance and/or discharge time constant and/or gain and/or dark count rate and/or photon detection efficiency, and the radiation dose values comprise standard dose values and dose measurement values acquired by the detectors; a determining unit for determining correction coefficient data according to the plurality of irradiation dose values; the processing unit is used for carrying out data preprocessing on the plurality of detector parameters and the dose measured values and determining training samples; And the training unit is used for inputting the correction coefficient data and the feature vector into a detection result correction model, and training the detection result correction model until the loss function of the detection result correction model converges.
9. A detection result correction device applied to radiation therapy, characterized by comprising: The acquisition unit is used for acquiring the dose measured value acquired by the detector, and breakdown voltage, series resistance, quenching resistance, parasitic resistance, pixel capacitance, discharge time constant, gain, dark count rate and photon detection efficiency; The correction unit is used for inputting the measured dose value, breakdown voltage and/or series resistance and/or quenching resistance and/or parasitic resistance and/or pixel capacitive reactance and/or discharge time constant and/or gain and/or dark counting rate and/or photon detection efficiency into the detection result correction model applied to radiotherapy, and determining a corrected dose value output by the detection result correction model applied to radiotherapy, wherein the detection result correction model is a detection result correction model applied to radiotherapy and trained by any one of claims 1-6.
10. A computer readable storage medium, characterized in that the storage medium stores a computer program which, when executed by a processor, implements the method of any of the preceding claims 1-7.
11. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of any of the preceding claims 1-7 when executing the program.
12. A computer program product, characterized in that instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform the method of any of claims 1-7.

Description

Training method and device applied to detection result correction model of radiotherapy Technical Field The present disclosure relates to the field of radiotherapy, and in particular, to a training method and apparatus for a correction model of a detection result applied to radiotherapy. Background In the modern radiotherapy process, the accuracy of dose measurement has important significance for ensuring the treatment effect and the safety of patients. In order to achieve high accuracy dose verification and treatment quality control, a Detector such as an X-ray flat panel imaging Detector (FLAT PANEL Detector, FPD), an ionization chamber array, a diode array, etc. is generally used for dose verification. However, in long-term, high-dose clinical application environments, the detector itself acts as a radiation-sensitive device, and its electronics structure and materials are inevitably subject to cumulative damage from ionizing radiation. Thereby affecting the accuracy of the detector response to the radiation signal and, in turn, causing systematic drift or error accumulation in the dose measurement. At present, in order to compensate the deviation in the measurement result of the detector, the common method mainly focuses on the influence correction of the ambient temperature and the humidity, but ignores the long-term and accumulated degradation effect caused by the ionizing radiation damage of the detector body, so that the requirement of the dose measurement precision in the high-dose and long-term operation environment is difficult to meet. Disclosure of Invention The present disclosure provides a training method and apparatus for a correction model of a detection result applied to radiotherapy, so as to at least partially solve the above-mentioned problems in the prior art. The technical scheme adopted in the specification is as follows: the specification provides a training method of a detection result correction model applied to radiotherapy, comprising the following steps: Collecting detector parameters and radiation dose values after each irradiation in the process of respectively irradiating a plurality of detectors for n times by using the radiation light, wherein the detector parameters comprise breakdown voltage and/or series resistance and/or quenching resistance and/or parasitic resistance and/or pixel capacitive reactance and/or discharge time constant and/or gain and/or dark counting rate and/or photon detection efficiency, and the radiation dose values comprise standard dose values and dose measurement values collected by the detectors; Determining correction coefficient data according to the plurality of irradiation dose values; Performing data preprocessing on the plurality of detector parameters and the dose measurement values to determine a training sample; and inputting the correction coefficient data and the training sample into a detection result correction model, and training the detection result correction model until the loss function of the detection result correction model converges. Preferably, before the irradiating the plurality of detectors n times respectively, the method further comprises: for any one detector, determining the maximum radiation dose corresponding to the detector according to the use scene of the detector; And determining the single radiation dose when the detector is irradiated for n times according to the maximum radiation dose corresponding to the detector. Preferably, the determining correction coefficient data according to the plurality of irradiation dose values includes: For any irradiation dose value, determining a correction coefficient value corresponding to the irradiation dose value through the following formula; Wherein, the The correction coefficient value corresponding to the ith detector after being irradiated for the nth time; the corresponding standard dose value of the ith detector after being irradiated by the nth time; dose measurements acquired for the ith detector after being illuminated for the nth time; and determining correction coefficient data according to the plurality of correction coefficient values. Preferably, the data preprocessing is performed on a plurality of detector parameters and dose measurements to determine training samples, including: and establishing a correlation between the detector parameters and the measured dose value by a least square method, wherein the correlation is as follows: Wherein, the Is the breakdown voltage; is the series resistance; is the quenching resistor; Is the parasitic resistance; capacitive reactance for the pixel; the discharge time constant is the discharge time constant, G is the gain, DCR is the dark count rate, PDE is the photon detection efficiency, D is a dose measurement value; Solving the related relationship through a target algorithm to determine a feature vector matrix; A training sample is determined based on the feature vector, the plurality of detector parameter