Search

CN-122005059-A - Radio frequency thermotherapy intelligent control method based on perturbation frequency sweep and multi-mode sensing

CN122005059ACN 122005059 ACN122005059 ACN 122005059ACN-122005059-A

Abstract

The invention relates to the technical field of radio frequency hyperthermia control, and particularly discloses a radio frequency hyperthermia intelligent control method based on micro-disturbance frequency sweep and multi-mode perception, which comprises the steps of applying micro-disturbance radio frequency scanning signals to a treatment target area, and processing feedback data to dynamically reconstruct a real-time dielectric characteristic map of a tissue; the method comprises the steps of acquiring real-time temperature distribution data of a magnetic resonance, fusing the map with the real-time temperature distribution data acquired by the magnetic resonance, calculating instantaneous specific absorption rate distribution and predicting thermal field evolution trend, constructing an optimization cost function based on the information, generating a dynamic adjustment instruction for radio frequency power, frequency, phase and multi-antenna array focal point position after solving, evaluating control efficiency through a closed loop verification mechanism and adaptively optimizing system parameters.

Inventors

  • Ge Luping
  • Wan Jingran

Assignees

  • 南京从景生物技术有限公司

Dates

Publication Date
20260512
Application Date
20260325

Claims (9)

  1. 1. The intelligent control method for the radio frequency thermotherapy based on the perturbation frequency sweep and the multi-mode perception is characterized by comprising the following steps: The method comprises the steps of S1, applying micro-disturbance radio frequency scanning signals to a treatment target area of a patient, receiving and processing feedback data of the radio frequency scanning signals in real time, solving an inverse problem, dynamically reconstructing dielectric constants and conductivity distribution of tissues in the treatment area on a plurality of frequency points, and generating a real-time dielectric characteristic map; s2, acquiring real-time temperature distribution data of a treatment area, performing space-time registration and fusion on the real-time dielectric characteristic map and the real-time temperature distribution data, calculating instantaneous specific absorption rate distribution reflecting the current energy deposition state, and predicting a thermal field evolution trend at a future moment; S3, based on instantaneous specific absorption rate distribution and predicted thermal field evolution trend, constructing a real-time optimization cost function aiming at improving the heating quality of a target area and inhibiting normal tissue damage, solving the real-time optimization cost function, and generating a dynamic adjustment instruction for radio frequency power, frequency, phase and multi-antenna array focus position; S4, after the dynamic adjustment instruction generated in the step S3 is executed, the steps S1 and S2 are executed again, new real-time dielectric characteristic patterns and real-time temperature distribution data are acquired again, and the control efficiency is verified by comparing differences of the dielectric characteristic patterns before and after adjustment and analyzing the coincidence degree of the real-time temperature distribution and the evolution trend of the thermal field.
  2. 2. The intelligent control method for radio frequency thermotherapy based on perturbation frequency sweep and multi-mode sensing according to claim 1, wherein the step S1 specifically comprises: Applying a perturbation radio frequency scanning signal covering a specific frequency range to a treatment target area, wherein the specific frequency range is divided into a plurality of discrete scanning sub-frequency bands in advance according to the type of target tissue; sequentially switching to each scanning sub-band, transmitting micro-disturbance radio frequency scanning signals, and synchronously receiving original echo signals which correspond to each sub-band and contain amplitude and phase information; preprocessing an original echo signal, wherein the preprocessing comprises the steps of carrying out dynamic weighted average on the signal and carrying out frequency domain transformation to extract frequency spectrum characteristic vectors corresponding to each sub-band; Carrying out spatial spectral cluster analysis on the spectral feature vectors under all the scanning sub-bands, and automatically dividing a treatment area into a plurality of tissue subregions with similar dielectric characteristics according to the response consistency of different spatial positions on a plurality of frequency bands; Based on the division result of the tissue subregions, a unified dielectric parameter initial value is given to each subregion, and an initial dielectric distribution field with space smoothness constraint is constructed according to the initial dielectric parameter initial value, so that a real-time dielectric characteristic map is generated.
  3. 3. The intelligent control method for radio frequency hyperthermia based on perturbation frequency sweep and multi-mode sensing according to claim 2, wherein the dividing process of the tissue subregion is as follows: Constructing a multiband response vector of each spatial position point, wherein the multiband response vector is formed by combining amplitude and phase components of a spectral feature vector of the spatial position under all scanning sub-bands according to the frequency band sequence; calculating similarity measurement between multiband response vectors of any two spatial position points, wherein the similarity measurement adopts a weighted composite calculation mode based on vector angle cosine and Euclidean distance, and a spatial similarity matrix is constructed based on the similarity measurement of all position point pairs; Spectral cluster analysis is carried out on the spatial similarity matrix, feature vectors of the spatial similarity matrix are obtained through feature decomposition, clustering division is carried out on the feature vectors, and spatial points with consistent spectral features are merged into the same tissue subregion.
  4. 4. The intelligent control method for radio frequency hyperthermia based on perturbation frequency sweep and multi-mode sensing according to claim 1, wherein the step S2 specifically comprises: carrying out consistency calibration on the real-time temperature distribution data, and identifying and eliminating transient temperature fluctuation caused by blood perfusion by analyzing the temperature change modes of adjacent time points to form a calibrated stable temperature field; performing superposition analysis on the spatial variation characteristics of the stable temperature field and the conductivity distribution in the real-time dielectric characteristic map, and performing inversion calculation to obtain instantaneous specific absorption rate distribution by establishing a corresponding relation between heat accumulation and electromagnetic energy absorption; Based on the spatial pattern of the instantaneous specific absorption rate distribution, the current state of a stable temperature field is combined, and the temperature propagation path and the heat distribution change in a specified time period in the future are estimated through recursive simulation of a heat diffusion process, so that a thermal field evolution trend is generated.
  5. 5. The intelligent control method for radio frequency hyperthermia based on perturbation frequency sweep and multi-mode sensing according to claim 4, wherein the specific acquisition process of the instantaneous specific absorption rate distribution is as follows: carrying out spatial position alignment on three-dimensional spatial gradient distribution of a stable temperature field and conductivity distribution in the real-time dielectric characteristic map, and identifying a key region with obvious temperature change and consistent conductivity characteristics; Establishing a quantitative conversion relation between heat accumulation and electromagnetic energy absorption of each key area, and calculating to obtain the heat generation rate of each area by analyzing the corresponding rule of temperature rise and conductivity in unit time; And according to the heat generation rate and the corresponding conductivity distribution, reversely deriving the electromagnetic energy absorption intensity of each space point by the energy conservation principle to form the instantaneous specific absorption rate distribution.
  6. 6. The intelligent control method for radio frequency hyperthermia based on perturbation frequency sweep and multi-mode sensing according to claim 4, wherein the generating thermal field evolution trend comprises: identifying heat aggregation centers in the instantaneous specific absorption rate distribution, analyzing the spatial distribution characteristics of each aggregation center and the heat transfer relation between each aggregation center and surrounding tissues, and determining the dominant heat transfer direction; based on the current state of the stable temperature field, a heat flow path network is established, and the heat exchange intensity relation between adjacent areas is established by analyzing the temperature gradient and heat conduction characteristics between tissues; Performing recursive calculation of heat transfer along a heat flow path network, simulating a diffusion process of heat from a collection center to surrounding tissues, and updating predicted temperature values of all areas; and carrying out space-time integration on predicted temperature values of a plurality of continuous time points to generate a dynamic trend graph showing the heat propagation path and the temperature distribution change as a thermal field evolution trend.
  7. 7. The intelligent control method for radio frequency hyperthermia based on perturbation frequency sweep and multi-mode sensing according to claim 1, wherein the step S3 specifically comprises: Extracting central focusing power and edge attenuation degree of energy distribution in the target area from the instantaneous specific absorption rate distribution, and extracting energy leakage intensity from the normal tissue area; taking the central focal power and the temperature uniformity as positive optimization indexes, taking the edge attenuation degree, the energy leakage intensity and the temperature rising rate as negative inhibition indexes, and establishing a comprehensive evaluation index system; setting a safety threshold range of the radio frequency parameters according to the negative inhibition index by adopting a layered screening process, and then adjusting the radio frequency parameter combination in the safety threshold range to enable the positive optimization index to reach an optimal state; And selecting the combination of the radio frequency power, the frequency, the phase and the focal position of the multi-antenna array, which lead the positive optimization index to be optimal and the negative inhibition index to meet the safety requirement, through multiple rounds of parameter adjustment and effect verification, and outputting the combination as a dynamic adjustment instruction.
  8. 8. The intelligent control method for radio frequency thermotherapy based on perturbation frequency sweep and multi-mode sensing according to claim 7, wherein the specific establishment process of the comprehensive evaluation index system is as follows: Quantifying the forward optimization index, wherein the central focal power is obtained by calculating the ratio of the peak value to the average value of the energy distribution in the target area, and the temperature uniformity is obtained by calculating the inverse standard deviation of the temperature distribution in the target area; Carrying out quantization treatment on the negative inhibition index, wherein the edge attenuation degree is obtained by calculating the attenuation gradient of the boundary energy value and the center energy value of the target area, the energy leakage intensity is obtained by counting the number of energy points exceeding a set threshold value in normal tissues, and the temperature rise rate is obtained by calculating the temperature change quantity of the normal tissues in unit time; according to the characteristics of the current treatment stage, dynamic weights are distributed to all quantized indexes, wherein the central focal power is preferentially improved in the initial treatment stage, the temperature uniformity is preferentially optimized in the treatment stable stage, and the negative inhibition indexes are preferentially controlled in the whole treatment process; and carrying out weighted fusion on each quantization index and the corresponding weight to form a single comprehensive evaluation value, and establishing a comprehensive evaluation index system.
  9. 9. The intelligent control method for radio frequency hyperthermia based on perturbation frequency sweep and multi-mode sensing according to claim 1, wherein the step S4 specifically comprises: after the dynamic adjustment instruction is executed, the perturbation radio frequency scanning and the temperature data acquisition are re-executed, and new real-time dielectric characteristic map and real-time temperature distribution data are obtained; Comparing the spatial structural similarity of the new dielectric characteristic map with the dielectric characteristic map before adjustment, and evaluating the improvement degree of the energy focusing effect by calculating the change of the tissue boundary definition and the dielectric parameter distribution consistency; performing time sequence comparison on the new real-time temperature distribution data and the predicted thermal field evolution trend, analyzing the deviation degree of an actual temperature change path and a predicted path, and evaluating the accuracy of thermal field prediction; And dynamically correcting the sampling frequency of medium characteristic reconstruction and the time step of thermal field prediction in the subsequent control process according to the energy focusing effect improvement degree and the evaluation result of thermal field prediction accuracy.

Description

Radio frequency thermotherapy intelligent control method based on perturbation frequency sweep and multi-mode sensing Technical Field The invention relates to the technical field of radio frequency hyperthermia control, in particular to an intelligent radio frequency hyperthermia control method based on perturbation frequency sweep and multi-mode perception. Background The radio frequency thermotherapy is used as an important means for tumor treatment, and the core of the curative effect is to realize the accurate and controllable deposition of energy in the target tissue. In the prior art, a mode of combining medical image guidance with temperature monitoring is adopted, and feedback adjustment is performed by monitoring the surface temperature of a body or a tumor surface. However, such methods rely solely on a hysteresis parameter of temperature and do not sense the dynamic changes in tissue electromagnetic properties during treatment. When the tissue is dehydrated, charred or blood flow changes due to heating, the dielectric properties thereof will develop non-linearly, resulting in a deviation of the actual energy deposition from the expected one, with the risk of overheating damage to the surrounding normal tissue due to insufficient treatment of the target area. The most important technical problems solved by the invention are thermal field mismatch and focus drift caused by the fact that the existing radio frequency thermotherapy control method cannot sense the dynamic change of tissue dielectric characteristics in real time. The traditional system carries out energy delivery based on a static tissue model, when the electromagnetic characteristics of tissues change in the treatment process, the energy deposition focus can drift from a preset target area to normal tissues, and the system can carry out lag adjustment only after overheat occurs. The invention reconstructs the dielectric characteristic map in real time through the micro-disturbance frequency sweep and combines the multi-mode temperature sensing to realize the feedforward prediction and the feedback correction of the energy deposition process and solve the control misalignment problem caused by the dynamic change of the tissue. Disclosure of Invention The invention aims to provide an intelligent control method for radio frequency thermotherapy based on micro-disturbance frequency sweep and multi-mode perception, so as to solve the problems in the background. The aim of the invention can be achieved by the following technical scheme: the intelligent control method for the radio frequency thermotherapy based on the perturbation frequency sweep and the multi-mode perception comprises the following steps: The method comprises the steps of S1, applying micro-disturbance radio frequency scanning signals to a treatment target area of a patient, receiving and processing feedback data of the radio frequency scanning signals in real time, solving an inverse problem, dynamically reconstructing dielectric constants and conductivity distribution of tissues in the treatment area on a plurality of frequency points, and generating a real-time dielectric characteristic map; s2, acquiring real-time temperature distribution data of a treatment area, performing space-time registration and fusion on the real-time dielectric characteristic map and the real-time temperature distribution data, calculating instantaneous specific absorption rate distribution reflecting the current energy deposition state, and predicting a thermal field evolution trend at a future moment; S3, based on instantaneous specific absorption rate distribution and predicted thermal field evolution trend, constructing a real-time optimization cost function aiming at improving the heating quality of a target area and inhibiting normal tissue damage, solving the real-time optimization cost function, and generating a dynamic adjustment instruction for radio frequency power, frequency, phase and multi-antenna array focus position; S4, after the dynamic adjustment instruction generated in the step S3 is executed, the steps S1 and S2 are executed again, new real-time dielectric characteristic patterns and real-time temperature distribution data are acquired again, and the control efficiency is verified by comparing differences of the dielectric characteristic patterns before and after adjustment and analyzing the coincidence degree of the real-time temperature distribution and the evolution trend of the thermal field. As a further scheme of the invention, the S1 specifically comprises the following steps: Applying a perturbation radio frequency scanning signal covering a specific frequency range to a treatment target area, wherein the specific frequency range is divided into a plurality of discrete scanning sub-frequency bands in advance according to the type of target tissue; sequentially switching to each scanning sub-band, transmitting micro-disturbance radio frequency scanning signals, and s