A gamma knife radiation monitoring method and gamma knife radiation monitoring system

Abstract

The application discloses a gamma knife radiation monitoring method, comprising the following steps: collecting radiation dose data and environmental temperature by using a flexible radiation sensing array module; collecting three-dimensional space coordinates of a target point, and patient position change and respiratory motion data in real time by using a multi-modal data acquisition module; an AI dynamic compensation and analysis module first corrects the collected radiation dose data in real time according to the obtained radiation dose data and environmental temperature, and obtains original radiation dose data; then, the original radiation dose data is compensated point by point and dynamically according to the three-dimensional space coordinates of the target point, and the patient position change and respiratory motion data, and an accurate actual irradiation radiation dose distribution is output; then, it is judged whether the radiation dose data is abnormal in real time, and an abnormal trigger signal is output when the radiation dose data is abnormal; and a gamma knife treatment device performs corresponding work according to the abnormal trigger signal. The application also discloses a gamma knife radiation monitoring system. The application can be used for gamma knife radiation closed-loop monitoring.

Description

Technical Field

[0001] This invention relates to gamma knife radiation monitoring, and more specifically, to a gamma knife radiation monitoring method based on a flexible sensor array and AI dynamic compensation. Background Technology

[0002] Gamma knife treatment relies on multiple beams of gamma rays focused on the target, requiring extremely high precision control of radiation dose. Existing radiation monitoring technologies are mainly divided into three categories: one is fixed-point dose measurement based on ionization chambers and thermoluminescent dosimeters (TLDs); the second is static quality control monitoring based on phantoms; and the third is regional radiation monitoring based on detector arrays.

[0003] Static quality control monitoring equipment based on phantoms utilizes standardized physical phantoms to perform regular and stable performance testing on medical imaging equipment, ensuring that image quality meets clinical and regulatory requirements. Monitoring is conducted even when the equipment is unattended. Therefore, commonly used clinical monitoring protocols currently include point-to-point dose measurement based on ionization chambers and thermoluminescent dosimeters (TLDs) and area radiation monitoring based on detector arrays. Among these: Ionization chambers offer high monitoring accuracy but lack adaptability, failing to conform to the complex surface of the human body; thermoluminescent dosimeters (TLDs) require offline readings after treatment and cannot provide real-time feedback on dose deviations; regional radiation monitoring equipment based on detector arrays is mostly a rigid structure, easily affected by changes in patient position and respiratory movements, leading to distorted monitoring data and making it difficult to achieve closed-loop management of "monitoring-early warning-intervention".

[0004] Existing gamma knife radiation monitoring systems have the following drawbacks: 1. Insufficient adaptability: Rigid monitoring equipment cannot fit complex areas such as the head and neck. Even slight changes in the patient's position can cause the monitoring points to shift, resulting in decreased data accuracy. 2. Lack of real-time capability: Most dosimeters require offline data analysis, making it impossible to provide synchronous feedback on dose deviations during treatment and hindering timely intervention for abnormal radiation risks; 3. Lack of error compensation: The influence of respiratory motion and target displacement on dose distribution is not considered, resulting in a deviation between the monitoring data and the actual irradiated dose, which cannot accurately reflect the radiation exposure of normal tissues. 4. Insufficient closed-loop capability: Most existing monitoring devices only have data acquisition functions and cannot be linked with the Gamma Knife treatment system. When abnormal doses occur, manual intervention is required, which results in a delayed response and is prone to medical risks.

[0005] Clearly, existing gamma knife radiation monitoring methods suffer from insufficient adaptability, lack of real-time performance, lack of error compensation, and insufficient closed-loop capability. Summary of the Invention

[0006] To address the technical problems of insufficient adaptability, lack of real-time performance, lack of error compensation, and insufficient closed-loop capability in existing gamma knife radiation monitoring methods, this invention provides a gamma knife radiation monitoring method and a gamma knife radiation monitoring system.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is to design a gamma knife radiation monitoring method, comprising: Based on the location of the treatment target, the avoidance area of ​​the flexible radiation sensing array module is adjusted, and the flexible radiation sensing array module is attached to the non-treatment area of ​​the human body surface. The target point is located in three dimensions using the multimodal data acquisition module, and its initial coordinates are extracted. Radiation dose threshold and temperature compensation parameters are set in the AI ​​dynamic compensation and analysis module. The flexible radiation sensor array module is used to collect radiation dose data and ambient temperature; the multimodal data acquisition module is used to collect the three-dimensional spatial coordinates of the target point and the patient's positional changes and respiratory motion data in real time. The AI ​​dynamic compensation and analysis module acquires the radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target point, and patient position changes and respiratory motion data. The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real time based on the obtained radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data. Then, based on the three-dimensional spatial coordinates of the target point and the patient's body position changes and respiratory motion data, it quantifies the radiation dose error value under different body position deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Based on the radiation dose error value, it performs point-by-point dynamic compensation on the original radiation dose data, outputs an accurate actual radiation dose distribution, and generates a real-time radiation dose cloud map. Finally, based on the preset radiation dose threshold standard and the actual radiation dose distribution, it judges in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs. The Gamma Knife treatment device receives the abnormal trigger signal and performs the corresponding work according to the abnormal trigger signal.

[0008] The flexible radiation sensing array module includes a radiation protection film, a distributed sensing unit, and a medical pressure-sensitive adhesive color bonding layer arranged sequentially from top to bottom. The distributed sensing unit is sandwiched between the radiation protection film and the medical pressure-sensitive adhesive color bonding layer. The distributed sensing unit includes: A radiation sensing unit for acquiring the radiation dose data; the radiation sensing unit comprises multiple units arranged in a ring array spaced apart from each other. A temperature sensor is used to collect the ambient temperature; A signal transmission unit is connected to the radiation sensing unit and the temperature sensor and wirelessly transmits the radiation dose data and ambient temperature to the AI ​​dynamic compensation and analysis module. The flexible radiation sensing array module is provided with hollowed-out avoidance holes, which constitute the avoidance area, and the radiation sensing unit is arranged around the avoidance area. The flexible radiation sensing array module also includes multiple medical non-woven adhesive ear wings located at the edge of the medical pressure-sensitive adhesive colored bonding layer.

[0009] The multimodal data acquisition module includes: A cone-beam computed tomography unit acquires the three-dimensional spatial coordinates of the target point; An inertial measurement unit that collects data on the patient's positional changes and respiratory movements; The cone-beam computed tomography unit performs a 360° circumferential scan of the treatment area to acquire continuous tomographic image data. Based on voxel modeling and target feature point matching algorithms, it extracts the grayscale feature points and spatial coordinates of the target points in the tomographic image data, generates the three-dimensional spatial coordinates of the target points through three-dimensional reconstruction, and tracks the coordinate changes in real time, and performs real-time calibration with the isocentric coordinates of the Gamma Knife treatment device. The inertial measurement unit (IMU) captures patient positional changes and respiratory motion data through chest-head dual-site data fusion and respiratory waveform feature extraction algorithms. The patient positional changes include pitch, roll, yaw rotation, X-axis translation, Y-axis translation, and Z-axis translation. The IMU includes a three-axis accelerometer and a three-axis gyroscope. The capture of patient positional changes and respiratory motion data includes: The accelerometer collects acceleration signals, filters and denoises the acceleration signals, and extracts the angle and displacement parameters of body position changes; The gyroscope collects angular velocity signals, and the angular velocity signals collected by the gyroscope are integrated. Combined with auxiliary motion data of the chest surface, the waveform, amplitude and period of respiratory motion are fitted to accurately identify the minute displacement of the target point caused by respiratory motion.

[0010] The method, based on a pre-calibrated temperature-radiation dose response curve and a linear compensation algorithm, performs real-time correction on the collected radiation dose data to obtain the original radiation dose data, including: The temperature sensor was calibrated in advance within the range of 0 to 45 °C, a database of radiation dose measurement errors at different temperatures was established, and a temperature-radiation dose compensation formula was fitted. During treatment, the AI ​​dynamic compensation and analysis module calculates the error correction coefficient based on the real-time collected ambient temperature and substitutes it into the temperature-radiation dose compensation formula to dynamically correct the original radiation dose data.

[0011] The radiation dose threshold includes the target radiation dose threshold and the stray radiation safety threshold; The real-time determination of whether radiation dose data is abnormal, and the output of an abnormality trigger signal when abnormality occurs, includes: When the target radiation dose deviates from the target radiation dose threshold by 1% to 2%, it is judged as a level one anomaly; the anomaly triggering signal is an audible and visual warning signal and a collimation adjustment command; When the target radiation dose exceeds the target radiation dose threshold by 2% or the stray radiation exceeds the stray radiation safety threshold, it is judged as a level two anomaly; the anomaly trigger signal is a pause command and anomaly information and fault location suggestions sent to the medical terminal; The Gamma Knife treatment device finely adjusts the collimator angle according to the collimation adjustment command to correct the radiation dose distribution. The correction process is fed back to the AI ​​dynamic compensation and analysis module in real time until the target radiation dose is restored to the target radiation dose threshold range. The Gamma Knife treatment device stops working upon receiving a pause command; The Gamma Knife radiation monitoring method also includes automatically storing the radiation dose data, ambient temperature, raw radiation dose data, three-dimensional spatial coordinates of the target point, patient position changes and respiratory motion data, actual radiation dose distribution, radiation dose cloud map and abnormal trigger signals. The stored data is archived using an encryption algorithm, and a quality control report is generated and synchronized to the hospital's radiotherapy information system to complete data traceability and filing.

[0012] The present invention also provides a gamma knife radiation monitoring system, comprising: A flexible radiation sensing array module is attached to a non-treatment area of ​​the human body surface. The flexible radiation sensing array module collects radiation dose data and ambient temperature. The multimodal data acquisition module collects the three-dimensional spatial coordinates of the target point, as well as patient position changes and respiratory motion data. The AI ​​dynamic compensation and analysis module is connected to the flexible radiation sensing array module and the multimodal data acquisition module. The AI ​​dynamic compensation and analysis module acquires the radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target point, and patient position change and respiratory motion data. The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real time based on the obtained radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data. Then, based on the three-dimensional spatial coordinates of the target point and the patient's body position changes and respiratory motion data, it quantifies the radiation dose error value under different body position deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Based on the radiation dose error value, it performs point-by-point dynamic compensation on the original radiation dose data, outputs an accurate actual radiation dose distribution, and generates a real-time radiation dose cloud map. Finally, based on the preset radiation dose threshold standard and the actual radiation dose distribution, it judges in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs. The Gamma Knife treatment device is connected to the AI ​​dynamic compensation and analysis module and performs corresponding tasks based on the abnormal trigger signal.

[0013] The flexible radiation sensing array module includes a radiation protection film, a distributed sensing unit, and a medical pressure-sensitive adhesive color bonding layer arranged sequentially from top to bottom. The distributed sensing unit is sandwiched between the radiation protection film and the medical pressure-sensitive adhesive color bonding layer. The distributed sensing unit includes: A radiation sensing unit for acquiring the radiation dose data; the radiation sensing unit comprises multiple units arranged in a ring array spaced apart from each other. A temperature sensor is used to collect the ambient temperature; A signal transmission unit is connected to the radiation sensing unit and the temperature sensor and wirelessly transmits the radiation dose data and ambient temperature to the AI ​​dynamic compensation and analysis module. The flexible radiation sensing array module is provided with hollowed-out avoidance holes, which form an avoidance area to avoid the target position of treatment, and the radiation sensing unit is arranged around the avoidance area. The flexible radiation sensing array module also includes multiple medical non-woven adhesive ear wings located at the edge of the medical pressure-sensitive adhesive colored bonding layer.

[0014] The multimodal data acquisition module includes: A cone-beam computed tomography unit acquires the three-dimensional spatial coordinates of the target point; An inertial measurement unit that collects data on the patient's positional changes and respiratory movements; The cone-beam computed tomography unit performs a 360° circumferential scan of the treatment area to acquire continuous tomographic image data. Based on voxel modeling and target feature point matching algorithms, it extracts the grayscale feature points and spatial coordinates of the target points in the tomographic image data, generates the three-dimensional spatial coordinates of the target points through three-dimensional reconstruction, and tracks the coordinate changes in real time, and performs real-time calibration with the isocentric coordinates of the Gamma Knife treatment device. The inertial measurement unit (IMU) captures patient positional changes and respiratory motion data through chest-head dual-site data fusion and respiratory waveform feature extraction algorithms. The patient positional changes include pitch, roll, yaw rotation, X-axis translation, Y-axis translation, and Z-axis translation. The IMU includes a three-axis accelerometer and a three-axis gyroscope. The capture of patient positional changes and respiratory motion data includes: The accelerometer collects acceleration signals, filters and denoises the acceleration signals, and extracts the angle and displacement parameters of body position changes; The gyroscope collects angular velocity signals, and the angular velocity signals collected by the gyroscope are integrated. Combined with auxiliary motion data of the chest surface, the waveform, amplitude and period of respiratory motion are fitted to accurately identify the minute displacement of the target point caused by respiratory motion.

[0015] The method, based on a pre-calibrated temperature-radiation dose response curve and a linear compensation algorithm, performs real-time correction on the collected radiation dose data to obtain the original radiation dose data, including: The temperature sensor was calibrated in advance within the range of 0 to 45 °C, a database of radiation dose measurement errors at different temperatures was established, and a temperature-radiation dose compensation formula was fitted. During treatment, the AI ​​dynamic compensation and analysis module calculates the error correction coefficient based on the real-time collected ambient temperature and substitutes it into the temperature-radiation dose compensation formula to dynamically correct the original radiation dose data.

[0016] The radiation dose threshold includes the target radiation dose threshold and the stray radiation safety threshold; The real-time determination of whether radiation dose data is abnormal, and the output of an abnormality trigger signal when abnormality occurs, includes: When the target radiation dose deviates from the target radiation dose threshold by 1% to 2%, it is judged as a level one anomaly; the anomaly triggering signal is an audible and visual warning signal and a collimation adjustment command; When the target radiation dose exceeds the target radiation dose threshold by 2% or the stray radiation exceeds the stray radiation safety threshold, it is judged as a level two anomaly; the anomaly trigger signal is a pause command and anomaly information and fault location suggestions sent to the medical terminal; The Gamma Knife treatment device finely adjusts the collimator angle according to the collimation adjustment command to correct the radiation dose distribution. The correction process is fed back to the AI ​​dynamic compensation and analysis module in real time until the target radiation dose is restored to the target radiation dose threshold range. The Gamma Knife treatment device stops working upon receiving a pause command; The Gamma Knife radiation monitoring system also includes a memory that automatically stores the radiation dose data, ambient temperature, raw radiation dose data, three-dimensional spatial coordinates of the target point, patient position changes and respiratory motion data, actual radiation dose distribution, radiation dose cloud map and abnormal trigger signals. It uses an encryption algorithm to archive the stored data, generate a quality control report, and synchronize it to the hospital's radiotherapy information system to complete data traceability and filing.

[0017] This invention utilizes a flexible radiation sensing array module to collect radiation dose data and ambient temperature, a multimodal data acquisition module to collect real-time three-dimensional spatial coordinates of the target point and patient positional and respiratory motion data, and an AI dynamic compensation and analysis module to acquire the radiation dose data, ambient temperature, target point three-dimensional spatial coordinates, and patient positional and respiratory motion data. The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real-time based on the acquired radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and a linear compensation algorithm, to obtain the original radiation dose data; then, based on the target... The three-dimensional spatial coordinates of the points, along with the patient's positional changes and respiratory motion data, are used to quantify the radiation dose error values ​​under different positional deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Then, based on these radiation dose error values, the original radiation dose data is dynamically compensated point-by-point, outputting a precise distribution of the actual irradiated radiation dose and generating a real-time radiation dose cloud map. Next, based on a preset radiation dose threshold standard and the actual irradiated radiation dose distribution, the system determines in real-time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs. The Gamma Knife treatment device receives the abnormality trigger signal and performs corresponding operations based on it. This allows the flexible radiation sensor array module to adapt to the complex human body surface, improving monitoring stability under positional changes. The flexible radiation sensor array module collects radiation dose data and ambient temperature in real time, while the multimodal data acquisition module collects the three-dimensional spatial coordinates of the target point and patient positional and respiratory motion data in real time. The ambient temperature, radiation dose data, target point three-dimensional spatial coordinates, and patient positional and respiratory motion data are synchronously fed back to the AI ​​dynamic compensation and analysis module for analysis. In case of anomalies, an anomaly trigger signal is output to the Gamma Knife treatment device. The Gamma Knife treatment device performs corresponding actions based on the anomaly trigger signal. This not only achieves real-time compensation for dose errors caused by ambient temperature and dynamic compensation for dose errors caused by breathing and positional changes, but also constructs a closed loop of "monitoring-analysis-early warning-intervention," linking the Gamma Knife system for precise control. This achieves human-fitting design, real-time monitoring, intelligent correction, and closed-loop intervention, solving the problems of poor adaptability, insufficient real-time performance, large errors, and insufficient closed-loop capability of existing technologies, thus improving the safety and accuracy of Gamma Knife treatment. Attached Figure Description

[0018] The present invention will now be described in detail with reference to the embodiments and accompanying drawings, wherein: Figure 1 This is a flowchart illustrating the steps of the gamma knife radiation monitoring method of the present invention. Figure 2 This is a schematic diagram of the flexible radiation sensing array module of the present invention; Figure 3This is a cross-sectional schematic diagram of the flexible radiation sensing array module of the present invention; Figure 4 This is a schematic diagram of the gamma knife radiation monitoring system of the present invention. Detailed Implementation

[0019] The specific embodiments of the present invention are further described below with reference to the accompanying drawings: Please see also Figures 1 to 4 The gamma knife radiation monitoring method of the present invention includes: Step 1: Adjust the avoidance area of ​​the flexible radiation sensing array module according to the location of the treatment target, and attach the flexible radiation sensing array module to the non-treatment area of ​​the human body surface.

[0020] The flexible radiation sensing array module 1 has a flexible patch sandwich structure, including a radiation protection film 11, a distributed sensing unit 12 and a medical pressure-sensitive adhesive color bonding layer 13 arranged from top to bottom. The distributed sensing unit is sandwiched between the radiation protection film and the medical pressure-sensitive adhesive color bonding layer.

[0021] The radiation protection film 11 is made of polyimide with a thickness of 0.05mm. Its core function is to protect the sensor from direct gamma ray irradiation, which can cause device aging and performance degradation. It also prevents physical damage to the sensing unit caused by body secretions and external friction during treatment. The protective film is made of low radiation attenuation material with a gamma ray penetration rate of ≥98%, and does not affect radiation dose acquisition and treatment beam transmission.

[0022] The distributed sensing unit 12 includes a radiation sensing unit 121, a temperature sensor 122, and a signal transmission unit 123. Wherein: The radiation sensing unit 121 is used to collect the radiation dose data; the radiation sensing unit includes multiple units arranged in a ring array spaced apart from each other. The radiation sensing unit uses a polysiloxane composite scintillator material (with added YAG:Ce luminescent particles) as the core of the sensor functional layer. The composite scintillator material itself has excellent flexibility properties, eliminating the need for an additional flexible substrate; this material serves as the core detection layer of the radiation sensing unit, and is externally encapsulated by a polysiloxane flexible encapsulation layer to form a complete radiation sensing unit.

[0023] The distributed sensing unit 12 contains 8-12 radiation sensing units arranged in a distributed ring layout. Each unit is 5mm×5mm in size, and the overall array size is 20cm×15cm, thus fitting an adult's head. Each radiation sensing unit integrates a scintillator detector and a signal conversion chip, which can simultaneously collect the dose around the target point, the body surface scattering dose, and the equipment leakage radiation dose. The sampling frequency reaches 100Hz, the dose measurement range is 0.01~200 Gy, and the measurement accuracy is ±1.5%.

[0024] Temperature sensor 122 is used to collect the ambient temperature. The temperature sensor adopts a miniature NTC temperature sensor with a temperature measurement accuracy of ±0.1 ℃. The sampling frequency is synchronized with the radiation dose acquisition, both being 100Hz, to collect the operating ambient temperature of the radiation sensing unit in real time, eliminating the interference of ambient temperature on radiation dose measurement.

[0025] The signal transmission unit 123 is connected to the radiation sensing unit and the temperature sensor and wirelessly transmits the radiation dose data and ambient temperature to the AI ​​dynamic compensation and analysis module.

[0026] The signal transmission unit can use the TI CC2650 Bluetooth chip to achieve encrypted transmission with a transmission delay of ≤50ms. It synchronously packages and encrypts radiation dose data and ambient temperature, and transmits them to the AI ​​dynamic compensation and analysis module. The transmission process uses frequency hopping technology, which does not interfere with the normal operation of the Gamma Knife device.

[0027] The flexible radiation sensing array module is provided with a hollowed-out avoidance hole 124, which constitutes the avoidance area, and the radiation sensing unit is arranged around the avoidance area.

[0028] The medical pressure-sensitive adhesive colored bonding layer 13 is a re-adhesive, residue-free medical-grade structure. The flexible radiation sensing array module also includes multiple medical non-woven adhesive ear flaps 14 located at the edges of the medical pressure-sensitive adhesive colored bonding layer. The adhesive ear flaps 14 are primarily used to fix the flexible radiation sensing array module and prevent it from moving relative to the body surface.

[0029] Based on the target location, the avoidance area of ​​the flexible radiation sensing array module is adjusted so that the avoidance hole is located in the treatment area. The flexible radiation sensing array module is then attached to a non-treatment area on the body surface, such as 2-3 cm lateral to the target point. The flexible radiation sensing array module and the treatment site are adjacent and avoidant, which neither obstructs the treatment site nor affects the focusing and transmission of the gamma beam, while enabling comprehensive monitoring of stray radiation and scattered dose around the target point, completely avoiding interference from the monitoring equipment to the treatment. During attachment, initial fixation is achieved through a pressure-sensitive adhesive layer, followed by secondary fixation through adhesive ear wings in non-treatment areas of the head (such as the temporal side or occipital side), ensuring no relative displacement between the array and the body surface during treatment. The fixation structure can be flexibly adjusted according to the treatment site, leaving no skin residue after disassembly, and its biocompatibility meets the YY / T 0287-2017 standard.

[0030] The second step involves using the multimodal data acquisition module to complete the three-dimensional localization of the target point and extract its initial coordinates; and then setting the radiation dose threshold and temperature compensation parameters in the AI ​​dynamic compensation and analysis module.

[0031] In this specific embodiment, the multimodal data acquisition module 2 includes a cone-beam computed tomography (CBCT) unit 21 and an inertial measurement unit (IMU) 22. Wherein: The cone-beam computed tomography (CBCT) unit is used to acquire the three-dimensional spatial coordinates of the target point. The CBCT module uses a medical imaging component with a resolution of 0.5 mm and is equipped with voxel modeling and target feature point matching algorithms.

[0032] The inertial measurement unit is used to collect data on the patient's positional changes and respiratory movements.

[0033] The cone-beam computed tomography (CBCT) unit performs a 360° circumferential scan of the treatment area to acquire continuous tomographic image data. Based on voxel modeling and target feature point matching algorithms, it extracts the grayscale feature points and spatial coordinates of the target points in the tomographic image data, generates the three-dimensional spatial coordinates of the target points through three-dimensional reconstruction, and tracks the coordinate changes in real time, performing real-time calibration with the isocentric coordinates of the Gamma Knife treatment device.

[0034] By employing sub-millimeter-level layered scanning and three-dimensional reconstruction algorithms, compared to traditional CBCT which only achieves image localization, the cone-beam computed tomography unit of this application adds real-time target coordinate extraction and dynamic calibration functions, with a localization error ≤0.1mm.

[0035] The inertial measurement unit (IMU) captures patient positional changes and respiratory motion data through chest-head dual-site data fusion and respiratory waveform feature extraction algorithms. These positional changes include pitch, roll, yaw, X-axis translation, Y-axis translation, and Z-axis translation. The IMU includes a three-axis accelerometer and a three-axis gyroscope. The IMU uses an MPU6050 six-axis sensor with a sampling frequency of 50Hz, measuring the object's angular velocity and acceleration in three-dimensional space to calculate the object's attitude. Compared to traditional IMUs that only collect motion data from a single location, this unit improves motion capture accuracy through multi-site data fusion, achieving a motion data error ≤0.05mm.

[0036] The data captured on patient position changes and respiratory movements include: The accelerometer collects acceleration signals, filters and denoises the acceleration signals, and extracts the angle and displacement parameters of body position changes; The gyroscope collects angular velocity signals, and the angular velocity signals collected by the gyroscope are integrated. Combined with auxiliary motion data of the chest surface, the waveform, amplitude and period of respiratory motion are fitted to accurately identify the minute displacement of the target point caused by respiratory motion.

[0037] Before treatment begins, each module is initialized, and the target point is located in three dimensions using the CBCT unit. The initial coordinates of the target point are extracted, and the radiation dose threshold, temperature compensation parameters, and safety parameters are set.

[0038] The multimodal data acquisition module is also equipped with a wireless data transmission unit. The wireless data transmission unit can use the TICC2650 Bluetooth chip to achieve encrypted transmission with a transmission delay of ≤50ms. The three-dimensional spatial coordinates of the target point, as well as the patient's positional changes and respiratory motion data, are synchronously packaged and encrypted before being transmitted to the AI ​​dynamic compensation and analysis module. The transmission process uses frequency hopping technology to avoid interfering with the normal operation of the Gamma Knife device.

[0039] The third step involves using the flexible radiation sensor array module to collect radiation dose data and ambient temperature; and using the multimodal data acquisition module to collect the three-dimensional spatial coordinates of the target point and the patient's positional changes and respiratory motion data in real time.

[0040] The radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target, and patient position changes and respiratory motion data are synchronously packaged and encrypted, and then transmitted to the AI ​​dynamic compensation and analysis module. Frequency hopping technology is used in the transmission process to avoid interfering with the normal operation of the Gamma Knife device.

[0041] The fourth step involves the AI ​​dynamic compensation and analysis module performing dynamic compensation and analysis on radiation dose data, ambient temperature, the three-dimensional spatial coordinates of the target point, and patient position changes and respiratory motion data. It determines in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs.

[0042] The AI ​​dynamic compensation and analysis module acquires the radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target point, and patient position changes and respiratory motion data.

[0043] The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real time based on the obtained radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data. Then, based on the three-dimensional spatial coordinates of the target point and the patient's positional changes and respiratory motion data, it quantifies the radiation dose error value under different positional deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Based on the radiation dose error value, it performs point-by-point dynamic compensation on the original radiation dose data, outputting an accurate distribution of the actual irradiated radiation dose and generating a real-time radiation dose cloud map. Finally, based on the preset radiation dose threshold standard and the actual irradiated radiation dose distribution, it determines in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs.

[0044] The AI ​​dynamic compensation and analysis module uses an industrial-grade tablet PC equipped with an Intel Core i7 processor, running Windows 10, and includes AI algorithm programs, data management software, and a visual user interface.

[0045] The system employs Python to develop algorithms for temperature compensation, target feature point extraction, respiratory waveform feature extraction, and radiation dose error compensation. It integrates the Qt interface framework to support visual operation and parameter adjustment by medical staff.

[0046] The algorithm and model can be trained using AI models. Data from 1,000 cases of Gamma Knife treatment (including radiation dose, ambient temperature, imaging, body position, and respiratory movement data) were collected to construct a convolutional neural network (CNN) model. The model parameters were optimized using gradient descent, and the training error was controlled within 1%.

[0047] The data uses the AES-256 encryption algorithm to encrypt data transmission and storage. The data output format is compatible with the DICOM standard and can be directly connected to the hospital's radiotherapy information system (RIS).

[0048] Step 5: The Gamma Knife treatment device receives the abnormal trigger signal and performs the corresponding work according to the abnormal trigger signal.

[0049] In this specific embodiment, the real-time correction of the collected radiation dose data based on the pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data includes: The temperature sensor was calibrated in advance within the range of 0 to 45 °C, a database of radiation dose measurement errors at different temperatures was established, and a temperature-radiation dose compensation formula was fitted. During treatment, the AI ​​dynamic compensation and analysis module calculates the error correction coefficient based on the real-time collected ambient temperature and substitutes it into the temperature-radiation dose compensation formula to dynamically correct the original radiation dose data.

[0050] The radiation dose thresholds include the target radiation dose threshold and the stray radiation safety threshold.

[0051] The real-time determination of whether radiation dose data is abnormal, and the output of an abnormality trigger signal when abnormality occurs, includes: When the target radiation dose deviates from the target radiation dose threshold by 1% to 2%, it is judged as a level one anomaly; the anomaly triggering signal is an audible and visual warning signal and a collimation adjustment command; When the target radiation dose exceeds the target radiation dose threshold by 2% or the stray radiation exceeds the stray radiation safety threshold, it is judged as a level two anomaly; the anomaly trigger signal is a pause command and anomaly information and fault location suggestions sent to the medical terminal; The Gamma Knife treatment device finely adjusts the collimator angle according to the collimation adjustment command to correct the radiation dose distribution. The correction process is fed back to the AI ​​dynamic compensation and analysis module in real time until the target radiation dose is restored to the target radiation dose threshold range. The gamma knife treatment device stops working upon receiving a pause command.

[0052] Step 6: Data Archiving The Gamma Knife radiation monitoring method also includes automatically storing the radiation dose data, ambient temperature, raw radiation dose data, three-dimensional spatial coordinates of the target point, patient positional changes and respiratory motion data, actual radiation dose distribution, radiation dose cloud map, and abnormal trigger signals. An encryption algorithm is used to archive the stored data, generate a quality control report, and synchronize it to the hospital's radiotherapy information system to complete data traceability and filing. The data storage format conforms to the DICOM radiotherapy standard, is compatible with the hospital's radiotherapy information system (RIS), and supports encrypted data export and full-cycle traceability.

[0053] This invention utilizes a flexible radiation sensor array module to adapt to the complex human body surface, improving monitoring stability under postural changes. The flexible radiation sensor array module collects radiation dose data and ambient temperature in real time, while a multimodal data acquisition module collects the three-dimensional spatial coordinates of the target point and patient postural and respiratory motion data in real time. These data are synchronously fed back to an AI dynamic compensation and analysis module for analysis. In case of anomalies, an anomaly trigger signal is output to the Gamma Knife treatment device, which then performs corresponding actions based on the anomaly trigger signal. This not only achieves real-time compensation for dose errors caused by ambient temperature and dynamic compensation for dose errors caused by respiration and postural changes, but also constructs a closed loop of "monitoring-analysis-early warning-intervention," linking the Gamma Knife system for precise control. It achieves surface adhesion, real-time dose monitoring, dynamic intelligent correction, and closed-loop treatment intervention, overcoming the shortcomings of existing technologies such as poor adaptability, insufficient real-time performance, large measurement errors, and weak closed-loop control capabilities, significantly improving the safety and accuracy of Gamma Knife radiotherapy.

[0054] Fifty patients who underwent Gamma Knife treatment for the head were selected and divided into an experimental group (monitored using this system) and a control group (monitored using a traditional TLD + ionization chamber). Clinical trials validated the results, which showed that: The dose monitoring error under positional changes was ≤1.5% in the experimental group and ≥4% in the control group.

[0055] The abnormal dose response time in the experimental group was ≤0.1s, while the control group required manual intervention (response time ≥30s).

[0056] The treatment comfort score of patients in the experimental group was 20% higher than that of the control group, and there were no problems such as monitoring equipment interfering with treatment or skin irritation at the application site.

[0057] When the temperature changes by 5 to 10 ℃, the dose measurement accuracy of the experimental group remains at ±1.5%, while the dose error of the control group increases to more than 3%.

[0058] Please see also Figure 4 The gamma knife radiation monitoring system of this invention includes a flexible radiation sensor array module 1, a multimodal data acquisition module 2, an AI dynamic compensation and analysis module 3, and a gamma knife treatment device 4. Wherein: The flexible radiation sensing array module 1 is attached to the non-treatment area of ​​the human body surface, and the flexible radiation sensing array module collects radiation dose data and ambient temperature.

[0059] The flexible radiation sensing array module 1 has a flexible patch sandwich structure, including a radiation protection film 11, a distributed sensing unit 12 and a medical pressure-sensitive adhesive color bonding layer 13 arranged from top to bottom. The distributed sensing unit is sandwiched between the radiation protection film and the medical pressure-sensitive adhesive color bonding layer.

[0060] The radiation protection film 11 is made of polyimide with a thickness of 0.05mm. Its core function is to protect the sensor from direct gamma ray irradiation, which can cause device aging and performance degradation. It also prevents physical damage to the sensing unit caused by body secretions and external friction during treatment. The protective film is made of low radiation attenuation material with a gamma ray penetration rate of ≥98%, and does not affect radiation dose acquisition and treatment beam transmission.

[0061] The distributed sensing unit 12 includes a radiation sensing unit 121, a temperature sensor 122, and a signal transmission unit 123. Wherein: The radiation sensing unit 121 is used to collect the radiation dose data; the radiation sensing unit includes multiple units arranged in a ring array spaced apart from each other. The radiation sensing unit uses a polysiloxane composite scintillator material (with added YAG:Ce luminescent particles) as the core of the sensor functional layer. The composite scintillator material itself has excellent flexibility properties, eliminating the need for an additional flexible substrate; this material serves as the core detection layer of the radiation sensing unit, and is externally encapsulated by a polysiloxane flexible encapsulation layer to form a complete radiation sensing unit.

[0062] The distributed sensing unit 12 contains 8 to 12 radiation sensing units arranged in a distributed ring layout. Each unit measures 5mm × 5mm, and the overall array size is 20cm × 15cm, thus fitting an adult's head. Each radiation sensing unit integrates a scintillator detector and a signal conversion chip, enabling simultaneous acquisition of dose around the target point, surface scattering dose, and equipment leakage radiation dose. The sampling frequency reaches 100Hz, the dose measurement range is 0.01–200 Gy, and the measurement accuracy is ±1.5%.

[0063] Temperature sensor 122 is used to collect the ambient temperature. The temperature sensor adopts a miniature NTC temperature sensor with a temperature measurement accuracy of ±0.1℃. The sampling frequency is synchronized with the radiation dose acquisition, both being 100Hz, to collect the ambient temperature of the radiation sensing unit in real time, eliminating the interference of ambient temperature on radiation dose measurement.

[0064] The signal transmission unit 123 is connected to the radiation sensing unit and the temperature sensor and wirelessly transmits the radiation dose data and ambient temperature to the AI ​​dynamic compensation and analysis module.

[0065] The signal transmission unit can use the TI CC2650 Bluetooth chip to achieve encrypted transmission with a transmission delay of ≤50ms. It synchronously packages and encrypts radiation dose data and ambient temperature, and transmits them to the AI ​​dynamic compensation and analysis module. The transmission process uses frequency hopping technology, which does not interfere with the normal operation of the Gamma Knife device.

[0066] The flexible radiation sensing array module is provided with a hollowed-out avoidance hole 124, which constitutes the avoidance area, and the radiation sensing unit is arranged around the avoidance area.

[0067] The medical pressure-sensitive adhesive colored bonding layer 13 is a re-adhesive, residue-free medical-grade structure. The flexible radiation sensing array module also includes multiple medical non-woven adhesive ear wings 14 disposed at the edge of the medical pressure-sensitive adhesive colored bonding layer. In this specific embodiment, three medical non-woven adhesive ear wings 14 are provided at the edge of the flexible radiation sensing array module.

[0068] Based on the target location, the avoidance area of ​​the flexible radiation sensing array module is adjusted so that the avoidance hole is located in the treatment area. The flexible radiation sensing array module is then attached to a non-treatment area on the body surface, such as 2-3 cm lateral to the target point. The flexible radiation sensing array module and the treatment site are adjacent and avoidant, which neither obstructs the treatment site nor affects the focusing and transmission of the gamma beam, while enabling comprehensive monitoring of stray radiation and scattered dose around the target point, completely avoiding interference from the monitoring equipment to the treatment. During attachment, initial fixation is achieved through a pressure-sensitive adhesive layer, followed by secondary fixation through adhesive ear wings in non-treatment areas of the head (such as the temporal side or occipital side), ensuring no relative displacement between the array and the body surface during treatment. The fixation structure can be flexibly adjusted according to the treatment site, leaving no skin residue after disassembly, and its biocompatibility meets the YY / T 0287-2017 standard.

[0069] The multimodal data acquisition module 2 acquires the three-dimensional spatial coordinates of the target point as well as the patient's positional changes and respiratory motion data.

[0070] In this specific embodiment, the multimodal data acquisition module 2 includes a cone-beam computed tomography (CBCT) unit and an inertial measurement unit (IMU). Wherein: The cone-beam computed tomography (CBCT) unit is used to acquire the three-dimensional spatial coordinates of the target point. The CBCT module uses a medical imaging component with a resolution of 0.5 mm and is equipped with voxel modeling and target feature point matching algorithms.

[0071] The inertial measurement unit is used to collect data on the patient's positional changes and respiratory movements.

[0072] The cone-beam computed tomography (CBCT) unit performs a 360° circumferential scan of the treatment area to acquire continuous tomographic image data. Based on voxel modeling and target feature point matching algorithms, it extracts the grayscale feature points and spatial coordinates of the target points in the tomographic image data, generates the three-dimensional spatial coordinates of the target points through three-dimensional reconstruction, and tracks the coordinate changes in real time, performing real-time calibration with the isocentric coordinates of the Gamma Knife treatment device.

[0073] By employing sub-millimeter-level layered scanning and three-dimensional reconstruction algorithms, compared to traditional CBCT which only achieves image localization, the cone-beam computed tomography unit of this application adds real-time target coordinate extraction and dynamic calibration functions, with a localization error ≤0.1mm.

[0074] The inertial measurement unit (IMU) captures patient positional changes and respiratory motion data through chest-head dual-site data fusion and respiratory waveform feature extraction algorithms. These positional changes include pitch, roll, yaw, X-axis translation, Y-axis translation, and Z-axis translation. The IMU includes a three-axis accelerometer and a three-axis gyroscope. The IMU uses an MPU6050 six-axis sensor with a sampling frequency of 50Hz, measuring the object's angular velocity and acceleration in three-dimensional space to calculate the object's attitude. Compared to traditional IMUs that only collect motion data from a single location, this unit improves motion capture accuracy through multi-site data fusion, achieving a motion data error ≤0.05mm.

[0075] The data captured on patient position changes and respiratory movements include: The accelerometer collects acceleration signals, filters and denoises the acceleration signals, and extracts the angle and displacement parameters of body position changes; The gyroscope collects angular velocity signals, and the angular velocity signals collected by the gyroscope are integrated. Combined with auxiliary motion data of the chest surface, the waveform, amplitude and period of respiratory motion are fitted to accurately identify the minute displacement of the target point caused by respiratory motion.

[0076] Before treatment begins, each module is initialized, and the target point is located in three dimensions using the CBCT unit. The initial coordinates of the target point are extracted, and the radiation dose threshold, temperature compensation parameters, and safety parameters are set.

[0077] The multimodal data acquisition module is also equipped with a wireless data transmission unit. The wireless data transmission unit can use the TICC2650 Bluetooth chip to achieve encrypted transmission with a transmission delay of ≤50ms. The three-dimensional spatial coordinates of the target point, as well as the patient's positional changes and respiratory motion data, are synchronously packaged and encrypted before being transmitted to the AI ​​dynamic compensation and analysis module. The transmission process uses frequency hopping technology to avoid interfering with the normal operation of the Gamma Knife device.

[0078] The AI ​​dynamic compensation and analysis module 3 is connected to the flexible radiation sensing array module and the multimodal data acquisition module. The AI ​​dynamic compensation and analysis module acquires the radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target point, and patient position changes and respiratory motion data.

[0079] The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real time based on the obtained radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data. Then, based on the three-dimensional spatial coordinates of the target point and the patient's positional changes and respiratory motion data, it quantifies the radiation dose error value under different positional deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Based on the radiation dose error value, it performs point-by-point dynamic compensation on the original radiation dose data, outputting an accurate distribution of the actual irradiated radiation dose and generating a real-time radiation dose cloud map. Finally, based on the preset radiation dose threshold standard and the actual irradiated radiation dose distribution, it determines in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs.

[0080] The AI ​​dynamic compensation and analysis module uses an industrial-grade tablet PC equipped with an Intel Core i7 processor, running Windows 10, and includes AI algorithm programs, data management software, and a visual user interface.

[0081] The system employs Python to develop algorithms for temperature compensation, target feature point extraction, respiratory waveform feature extraction, and radiation dose error compensation. It integrates the Qt interface framework to support visual operation and parameter adjustment by medical staff.

[0082] The algorithm and model can be trained using AI models. Data from 1,000 cases of Gamma Knife treatment (including radiation dose, ambient temperature, imaging, body position, and respiratory movement data) were collected to construct a convolutional neural network (CNN) model. The model parameters were optimized using gradient descent, and the training error was controlled within 1%.

[0083] The data uses the AES-256 encryption algorithm to encrypt data transmission and storage. The data output format is compatible with the DICOM standard and can be directly connected to the hospital's radiotherapy information system (RIS).

[0084] The Gamma Knife treatment device 4 is connected to the AI ​​dynamic compensation and analysis module and performs corresponding tasks based on the abnormal trigger signal.

[0085] In this specific embodiment, the real-time correction of the collected radiation dose data based on the pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data includes: The temperature sensor was calibrated in advance within the range of 0 to 45 °C, a database of radiation dose measurement errors at different temperatures was established, and a temperature-radiation dose compensation formula was fitted. During treatment, the AI ​​dynamic compensation and analysis module calculates the error correction coefficient based on the real-time collected ambient temperature and substitutes it into the temperature-radiation dose compensation formula to dynamically correct the original radiation dose data.

[0086] The radiation dose thresholds include the target radiation dose threshold and the stray radiation safety threshold.

[0087] The real-time determination of whether radiation dose data is abnormal, and the output of an abnormality trigger signal when abnormality occurs, includes: When the target radiation dose deviates from the target radiation dose threshold by 1% to 2%, it is judged as a level one anomaly; the anomaly triggering signal is an audible and visual warning signal and a collimation adjustment command; When the target radiation dose exceeds the target radiation dose threshold by 2% or the stray radiation exceeds the stray radiation safety threshold, it is judged as a level two anomaly; the anomaly trigger signal is a pause command and anomaly information and fault location suggestions sent to the medical terminal; The Gamma Knife treatment device finely adjusts the collimator angle according to the collimation adjustment command to correct the radiation dose distribution. The correction process is fed back to the AI ​​dynamic compensation and analysis module in real time until the target radiation dose is restored to the target radiation dose threshold range. The gamma knife treatment device stops working upon receiving a pause command.

[0088] In this specific embodiment, the Gamma Knife radiation monitoring system further includes a memory 5. The memory 5 automatically stores the radiation dose data, ambient temperature, raw radiation dose data, three-dimensional spatial coordinates of the target point, patient position changes and respiratory motion data, actual radiation dose distribution, radiation dose cloud map, and abnormal trigger signals. It employs an encryption algorithm to archive the stored data, generates a quality control report, and synchronizes it to the hospital's radiotherapy information system to complete data traceability and filing. The data storage format conforms to the DICOM radiotherapy standard, is compatible with the hospital's radiotherapy information system (RIS), and supports encrypted data export and full-cycle traceability.

[0089] This invention utilizes a flexible radiation sensor array module to adapt to the complex surface of the human body, improving monitoring stability under postural changes. The flexible radiation sensor array module collects radiation dose data and ambient temperature in real time, while a multimodal data acquisition module collects the three-dimensional spatial coordinates of the target point and patient postural and respiratory motion data in real time. These data are then synchronously fed back to an AI dynamic compensation and analysis module for analysis. In case of anomalies, an anomaly trigger signal is output to the Gamma Knife treatment device. The Gamma Knife treatment device then performs corresponding actions based on the anomaly trigger signal. This not only achieves real-time compensation for dose errors caused by ambient temperature and dynamic compensation for dose errors caused by respiration and postural changes, but also constructs a closed loop of "monitoring-analysis-early warning-intervention," linking with the Gamma Knife system for precise control. This achieves a fit to the human body, real-time monitoring, intelligent correction, and closed-loop intervention, solving the problems of poor adaptability, insufficient real-time performance, large errors, and insufficient closed-loop capability in existing technologies, thus improving the safety and accuracy of Gamma Knife treatment.

[0090] This invention has the following characteristics: 1. Significantly improves monitoring adaptability and stability, solving the problem of complex body surface adhesion: This invention uses polysiloxane composite scintillator material as the sensor core. The material itself has flexible properties, eliminating the need for an additional substrate. The flexible radiation sensing array module is designed with a sandwich structure, with a bottom layer of medical pressure-sensitive adhesive colored bonding layer and edge adhesive ear wings for double fixation, leaving no adhesive residue and exhibiting excellent biocompatibility. The ring-shaped avoidance layout both conforms to complex body surfaces and does not obstruct the treatment site. Clinical testing shows that the dose monitoring error under positional changes is ≤1.5%, far superior to the control group (error ≥4%), significantly improving monitoring stability in complex treatment scenarios.

[0091] 2. Real-time monitoring and dynamic error compensation ensure accurate dose measurement: The system eliminates the interference of temperature on radiation dose measurement in real time through a temperature sensor and pre-calibration compensation algorithm; it dynamically compensates for radiation dose error by fusing CBCT target coordinates and IMU motion data based on a CNN model, with a model training error ≤1%; the sensor array has a sampling frequency of 100Hz, which can simultaneously collect multi-dimensional dose data. Compared with the shortcomings of traditional TLD dosimeters that are offline and have no error compensation, this system can output accurate actual radiation dose distribution, completely solving the problem of the disconnect between monitoring data and actual radiation dose.

[0092] 3. Constructing an intelligent closed-loop management system to improve treatment safety and response efficiency: This system links with the Gamma Knife treatment system through a closed-loop control and early warning module. Based on dual threshold standards (target dose ±2%, stray radiation conforming to GBZ121—2020 standards), it achieves graded early warning and intervention. Level 1 anomalies automatically adjust the collimator angle to correct the dose, while Level 2 anomalies immediately trigger beam emission pause. The anomaly response time is ≤0.1s. Compared to existing technologies that require manual intervention (response time ≥30s) and are inherently slow, this system can quickly avoid medical risks caused by abnormal radiation, upgrading from "passive monitoring" to "active management," comprehensively protecting the safety of patients and medical staff.

[0093] 4. Optimize clinical treatment experience while balancing practicality and compliance: The lightweight and flexible design of the flexible sensor array, combined with a medical-grade fixation structure, improves patient comfort during treatment. In clinical trials, patient comfort scores were 20% higher than traditional monitoring solutions. The data storage and traceability module adopts the AES-256 encryption algorithm, is compatible with hospital radiotherapy information systems and the DICOM standard, and enables full-cycle traceability of dose data, meeting the dual needs of clinical quality control and radiation protection supervision. No additional data management system is required, reducing the cost of clinical implementation.

[0094] 5. Wide compatibility and easy upgrades, with strong clinical application value: Adopting a modular design, it is compatible with mainstream Gamma Knife devices for the head and body. The avoidance area and size of the flexible sensor array can be flexibly adjusted according to the treatment site. The hardware selection consists of mature medical components, which are easy to assemble and maintain. At the same time, the AI ​​compensation model and threshold parameters can be upgraded and optimized through software algorithms without large-scale hardware modifications. It can not only meet the monitoring needs of existing Gamma Knife treatment, but also adapt to the future development of precision radiotherapy technology, with broad prospects for promotion and application.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring gamma knife radiation, characterized in that... include: Based on the location of the treatment target, the avoidance area of ​​the flexible radiation sensing array module is adjusted, and the flexible radiation sensing array module is attached to the non-treatment area of ​​the human body surface. The target point is located in three dimensions using the multimodal data acquisition module, and its initial coordinates are extracted. Radiation dose threshold and temperature compensation parameters are set in the AI ​​dynamic compensation and analysis module. The flexible radiation sensor array module is used to collect radiation dose data and ambient temperature; the multimodal data acquisition module is used to collect the three-dimensional spatial coordinates of the target point and the patient's positional changes and respiratory motion data in real time. The AI ​​dynamic compensation and analysis module acquires the radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target point, and patient position changes and respiratory motion data. The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real time based on the obtained radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data. Then, based on the three-dimensional spatial coordinates of the target point and the patient's body position changes and respiratory motion data, it quantifies the radiation dose error value under different body position deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Based on the radiation dose error value, it performs point-by-point dynamic compensation on the original radiation dose data, outputs an accurate actual radiation dose distribution, and generates a real-time radiation dose cloud map. Finally, based on the preset radiation dose threshold standard and the actual radiation dose distribution, it judges in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs. The Gamma Knife treatment device receives the abnormal trigger signal and performs the corresponding work according to the abnormal trigger signal.

2. The gamma knife radiation monitoring method according to claim 1, characterized in that: The flexible radiation sensing array module includes a radiation protection film, a distributed sensing unit, and a medical pressure-sensitive adhesive color bonding layer arranged sequentially from top to bottom. The distributed sensing unit is sandwiched between the radiation protection film and the medical pressure-sensitive adhesive color bonding layer. The distributed sensing unit includes: A radiation sensing unit for acquiring the radiation dose data; the radiation sensing unit comprises multiple units arranged in a ring array spaced apart from each other. A temperature sensor is used to collect the ambient temperature; A signal transmission unit is connected to the radiation sensing unit and the temperature sensor and wirelessly transmits the radiation dose data and ambient temperature to the AI ​​dynamic compensation and analysis module. The flexible radiation sensing array module is provided with hollowed-out avoidance holes, which constitute the avoidance area, and the radiation sensing unit is arranged around the avoidance area. The flexible radiation sensing array module also includes multiple medical non-woven adhesive ear wings located at the edge of the medical pressure-sensitive adhesive colored bonding layer.

3. The gamma knife radiation monitoring method according to claim 1, characterized in that: The multimodal data acquisition module includes: A cone-beam computed tomography unit acquires the three-dimensional spatial coordinates of the target point; An inertial measurement unit that collects data on the patient's positional changes and respiratory movements; The cone-beam computed tomography unit performs a 360° circumferential scan of the treatment area to acquire continuous tomographic image data. Based on voxel modeling and target feature point matching algorithms, it extracts the grayscale feature points and spatial coordinates of the target points in the tomographic image data, generates the three-dimensional spatial coordinates of the target points through three-dimensional reconstruction, and tracks the coordinate changes in real time, and performs real-time calibration with the isocentric coordinates of the Gamma Knife treatment device. The inertial measurement unit (IMU) captures patient positional changes and respiratory motion data through chest-head dual-site data fusion and respiratory waveform feature extraction algorithms. The patient positional changes include pitch, roll, yaw rotation, X-axis translation, Y-axis translation, and Z-axis translation. The IMU includes a three-axis accelerometer and a three-axis gyroscope. The capture of patient positional changes and respiratory motion data includes: The accelerometer collects acceleration signals, filters and denoises the acceleration signals, and extracts the angle and displacement parameters of body position changes; The gyroscope collects angular velocity signals, and the angular velocity signals collected by the gyroscope are integrated. Combined with auxiliary motion data of the chest surface, the waveform, amplitude and period of respiratory motion are fitted to accurately identify the minute displacement of the target point caused by respiratory motion.

4. The gamma knife radiation monitoring method according to claim 2, characterized in that: The method, based on a pre-calibrated temperature-radiation dose response curve and a linear compensation algorithm, performs real-time correction on the collected radiation dose data to obtain the original radiation dose data, including: The temperature sensor was calibrated in advance within the range of 0 to 45 °C, a database of radiation dose measurement errors at different temperatures was established, and a temperature-radiation dose compensation formula was fitted. During treatment, the AI ​​dynamic compensation and analysis module calculates the error correction coefficient based on the real-time collected ambient temperature and substitutes it into the temperature-radiation dose compensation formula to dynamically correct the original radiation dose data.

5. The gamma knife radiation monitoring method according to claim 1, characterized in that: The radiation dose threshold includes the target radiation dose threshold and the stray radiation safety threshold; The real-time determination of whether radiation dose data is abnormal, and the output of an abnormality trigger signal when abnormality occurs, includes: When the target radiation dose deviates from the target radiation dose threshold by 1% to 2%, it is judged as a level one anomaly; the anomaly triggering signal is an audible and visual warning signal and a collimation adjustment command; When the target radiation dose exceeds the target radiation dose threshold by 2% or the stray radiation exceeds the stray radiation safety threshold, it is judged as a level two anomaly; the anomaly trigger signal is a pause command and anomaly information and fault location suggestions sent to the medical terminal; The Gamma Knife treatment device finely adjusts the collimator angle according to the collimation adjustment command to correct the radiation dose distribution. The correction process is fed back to the AI ​​dynamic compensation and analysis module in real time until the target radiation dose is restored to the target radiation dose threshold range. The Gamma Knife treatment device stops working upon receiving a pause command; The Gamma Knife radiation monitoring method also includes automatically storing the radiation dose data, ambient temperature, raw radiation dose data, three-dimensional spatial coordinates of the target point, patient position changes and respiratory motion data, actual radiation dose distribution, radiation dose cloud map and abnormal trigger signals. The stored data is archived using an encryption algorithm, and a quality control report is generated and synchronized to the hospital's radiotherapy information system to complete data traceability and filing.

6. A gamma knife radiation monitoring system, characterized in that... include: A flexible radiation sensing array module is attached to a non-treatment area of ​​the human body surface. The flexible radiation sensing array module collects radiation dose data and ambient temperature. The multimodal data acquisition module collects the three-dimensional spatial coordinates of the target point, as well as patient position changes and respiratory motion data. The AI ​​dynamic compensation and analysis module is connected to the flexible radiation sensing array module and the multimodal data acquisition module. The AI ​​dynamic compensation and analysis module acquires the radiation dose data, ambient temperature, three-dimensional spatial coordinates of the target point, and patient position change and respiratory motion data. The AI ​​dynamic compensation and analysis module first corrects the acquired radiation dose data in real time based on the obtained radiation dose data and ambient temperature, using a pre-calibrated temperature-radiation dose response curve and linear compensation algorithm to obtain the original radiation dose data. Then, based on the three-dimensional spatial coordinates of the target point and the patient's body position changes and respiratory motion data, it quantifies the radiation dose error value under different body position deviations and respiratory motion amplitudes based on motion characteristics and a radiation dose error model. Based on the radiation dose error value, it performs point-by-point dynamic compensation on the original radiation dose data, outputs an accurate actual radiation dose distribution, and generates a real-time radiation dose cloud map. Finally, based on the preset radiation dose threshold standard and the actual radiation dose distribution, it judges in real time whether the radiation dose data is abnormal and outputs an abnormality trigger signal when an abnormality occurs. The Gamma Knife treatment device is connected to the AI ​​dynamic compensation and analysis module and performs corresponding tasks based on the abnormal trigger signal.

7. The gamma knife radiation monitoring system according to claim 6, characterized in that: The flexible radiation sensing array module includes a radiation protection film, a distributed sensing unit, and a medical pressure-sensitive adhesive color bonding layer arranged sequentially from top to bottom. The distributed sensing unit is sandwiched between the radiation protection film and the medical pressure-sensitive adhesive color bonding layer. The distributed sensing unit includes: A radiation sensing unit for acquiring the radiation dose data; the radiation sensing unit comprises multiple units arranged in a ring array spaced apart from each other. A temperature sensor is used to collect the ambient temperature; A signal transmission unit is connected to the radiation sensing unit and the temperature sensor and wirelessly transmits the radiation dose data and ambient temperature to the AI ​​dynamic compensation and analysis module. The flexible radiation sensing array module is provided with hollowed-out avoidance holes, which form an avoidance area to avoid the target position of treatment, and the radiation sensing unit is arranged around the avoidance area. The flexible radiation sensing array module also includes multiple medical non-woven adhesive ear wings located at the edge of the medical pressure-sensitive adhesive colored bonding layer.

8. The gamma knife radiation monitoring system according to claim 6, characterized in that: The multimodal data acquisition module includes: A cone-beam computed tomography unit acquires the three-dimensional spatial coordinates of the target point; An inertial measurement unit that collects data on the patient's positional changes and respiratory movements; The cone-beam computed tomography unit performs a 360° circumferential scan of the treatment area to acquire continuous tomographic image data. Based on voxel modeling and target feature point matching algorithms, it extracts the grayscale feature points and spatial coordinates of the target points in the tomographic image data, generates the three-dimensional spatial coordinates of the target points through three-dimensional reconstruction, and tracks the coordinate changes in real time, and performs real-time calibration with the isocentric coordinates of the Gamma Knife treatment device. The inertial measurement unit (IMU) captures patient positional changes and respiratory motion data through chest-head dual-site data fusion and respiratory waveform feature extraction algorithms. The patient positional changes include pitch, roll, yaw rotation, X-axis translation, Y-axis translation, and Z-axis translation. The IMU includes a three-axis accelerometer and a three-axis gyroscope. The capture of patient positional changes and respiratory motion data includes: The accelerometer collects acceleration signals, filters and denoises the acceleration signals, and extracts the angle and displacement parameters of body position changes; The gyroscope collects angular velocity signals, and the angular velocity signals collected by the gyroscope are integrated. Combined with auxiliary motion data of the chest surface, the waveform, amplitude and period of respiratory motion are fitted to accurately identify the minute displacement of the target point caused by respiratory motion.

9. The gamma knife radiation monitoring system according to claim 7, characterized in that: The method, based on a pre-calibrated temperature-radiation dose response curve and a linear compensation algorithm, performs real-time correction on the collected radiation dose data to obtain the original radiation dose data, including: The temperature sensor was calibrated in advance within the range of 0 to 45 °C, a database of radiation dose measurement errors at different temperatures was established, and a temperature-radiation dose compensation formula was fitted. During treatment, the AI ​​dynamic compensation and analysis module calculates the error correction coefficient based on the real-time collected ambient temperature and substitutes it into the temperature-radiation dose compensation formula to dynamically correct the original radiation dose data.

10. The gamma knife radiation monitoring system according to claim 6, characterized in that: The radiation dose threshold includes the target radiation dose threshold and the stray radiation safety threshold; The real-time determination of whether radiation dose data is abnormal, and the output of an abnormality trigger signal when abnormality occurs, includes: When the target radiation dose deviates from the target radiation dose threshold by 1% to 2%, it is judged as a level one anomaly; the anomaly triggering signal is an audible and visual warning signal and a collimation adjustment command; When the target radiation dose exceeds the target radiation dose threshold by 2% or the stray radiation exceeds the stray radiation safety threshold, it is judged as a level two anomaly; the anomaly trigger signal is a pause command and anomaly information and fault location suggestions sent to the medical terminal; The Gamma Knife treatment device finely adjusts the collimator angle according to the collimation adjustment command to correct the radiation dose distribution. The correction process is fed back to the AI ​​dynamic compensation and analysis module in real time until the target radiation dose is restored to the target radiation dose threshold range. The Gamma Knife treatment device stops working upon receiving a pause command; The Gamma Knife radiation monitoring system also includes a memory that automatically stores the radiation dose data, ambient temperature, raw radiation dose data, three-dimensional spatial coordinates of the target point, patient position changes and respiratory motion data, actual radiation dose distribution, radiation dose cloud map and abnormal trigger signals. It uses an encryption algorithm to archive the stored data, generate a quality control report, and synchronize it to the hospital's radiotherapy information system to complete data traceability and filing.

Images