CN-119581007-B - Method for evaluating benign and malignant tumor incised margin in operation

CN119581007BCN 119581007 BCN119581007 BCN 119581007BCN-119581007-B

Abstract

The application relates to the technical field of medical treatment, in particular to an intraoperative tumor incisal edge benign and malignant assessment method which comprises the steps of obtaining optical coherence tomography data of a historical incisal edge tissue, extracting an A-line reflectivity curve based on the optical coherence tomography data of the historical incisal edge tissue, calculating and extracting target characteristics based on the A-line reflectivity curve, wherein the target characteristics comprise depth characteristics, slope characteristics, distance characteristics between local maxima, intensity rapid fluctuation characteristics, frequency characteristics and optical attenuation coefficient characteristics, training a preset classifier based on benign and malignant labels of the target characteristics and the target characteristics corresponding to the A-line reflectivity curve to obtain a trained classifier, and inputting the target characteristics extracted based on the optical coherence tomography data of the incisal edge tissue to be tested into the trained classifier to obtain an assessment result. Therefore, the accuracy and the instantaneity of evaluating the benign and malignant tumor cutting edge can be greatly improved.

Inventors

FENG CUIXIA
PAN HEFU
YANG GUOYING
GUAN CHENXI
Wan Chaochen
ZHAO ZHONGMIN
WANG HAITAO

Assignees

北京心联光电科技有限公司

Dates

Publication Date: 20260512
Application Date: 20250208

Claims (7)

1. An intraoperative tumor incisional benign and malignant evaluation method, which is characterized by comprising the following steps: Acquiring optical coherence tomography data of historical incisal edge tissue; Extracting an A-line reflectivity curve based on the optical coherence tomography data of the historical incisal edge tissue, wherein the A-line reflectivity curve is a curve of image intensity changing along with the depth of the A-line; Calculating and extracting target features based on the A-line reflectivity curve, wherein the target features comprise depth features, slope features, distance features between local maxima, intensity rapid fluctuation features, frequency features and optical attenuation coefficient features, and the depth features comprise penetration depth of an effective signal interval; Training a preset classifier based on the target feature and a benign and malignant label of the A-line reflectivity curve corresponding to the target feature to obtain a trained classifier; Inputting target features extracted based on optical coherence tomography data of the incisal edge tissue to be detected into the trained classifier to obtain an evaluation result; the optical attenuation coefficient type characteristics comprise attenuation coefficients and attenuation coefficient amplitudes, wherein the process of extracting the optical attenuation coefficient type characteristics comprises the steps of introducing an axial confocal diffusion function and additive substrate noise of an OCT system into an optical attenuation model based on beer's law to obtain a tissue optical attenuation model based on the OCT system; the tissue optical attenuation model based on the OCT system is as follows: Wherein I (z) is the light intensity value at depth z, z cf is the depth position of the focal plane from the tissue boundary, z R is the Rayleigh length, A 1 is the OCT signal amplitude, μ t is the attenuation coefficient, A 2 is the additive basal noise; the rapid fluctuation characteristic of the intensity comprises a standard deviation and a peak value of the rapid fluctuation of the intensity, and the process for extracting the rapid fluctuation characteristic of the intensity comprises the following steps: performing smoothing treatment on the current A-line reflectivity curve to obtain a current A-line reflectivity smoothing curve; Determining a difference value of rapid fluctuation of intensity between the current A-line reflectivity curve and the current A-line reflectivity smooth curve; calculating the standard deviation of the difference value to be used as the standard deviation of the rapid fluctuation of the intensity of the current A-line reflectivity curve; and generating a difference curve based on the difference, and taking the local maximum value with the largest absolute value in the difference curve as the intensity rapid fluctuation peak value of the current A-line reflectivity curve.
2. The method of claim 1, wherein extracting the penetration depth of the effective signal interval comprises: performing smoothing treatment on the current A-line reflectivity curve to obtain a current A-line reflectivity smoothing curve; determining an effective signal interval of a current A-line reflectivity smooth curve; and taking the difference between the depths of the starting point and the end point of the effective signal interval as the penetration depth of the effective signal interval of the current A-line reflectivity curve.
3. The method of claim 2, wherein determining the effective signal interval of the current a-line reflectance smoothing curve comprises: Taking a depth position, corresponding to the target value, of the image intensity in the current A-line reflectivity smooth curve as a starting point of an effective signal interval; And selecting a part behind the starting point of the effective signal interval of the current A-line reflectivity smooth curve through a preset sliding window, and taking the starting point depth position in the current window as the end point of the effective signal interval when the linear fitting slope of the signal in the window tends to 0 or the signal in the window is completely 0.
4. The method of claim 1, wherein the slope characteristics include an average slope of the effective signal interval, an overall slope of the effective signal interval, and an overall intercept of the effective signal interval, and the step of extracting the slope characteristics includes: Sliding and selecting an effective signal interval of a current A-line reflectivity smooth curve through a preset sliding window to obtain a plurality of data subsets; Calculating the linear fitting slope corresponding to each data subset respectively, and taking the average value of the linear fitting slopes of all the data subsets as the average slope of the effective signal interval of the current A-line reflectivity curve; And calculating the linear fitting slope and intercept of the effective signal section of the current A-line reflectivity smooth curve, taking the slope as the integral slope of the effective signal section of the current A-line reflectivity curve, and taking the intercept as the integral intercept of the effective signal section of the current A-line reflectivity curve.
5. The method for evaluating the benign and malignant tumor margin according to claim 1, wherein the distance features between the local maxima comprise a distance mean and a distance standard deviation, and the process for extracting the distance features between the local maxima comprises the steps of: Determining a local maximum value of the image intensity in an effective signal interval of the current A-line reflectivity curve; Calculating the difference value of the depth corresponding to each adjacent local maximum value; taking the average value of all the differences as the distance average value of the current A-line reflectivity curve, and taking the standard deviation of all the differences as the distance standard deviation of the current A-line reflectivity curve.
6. The method for evaluating the benign and malignant tumor margin in operation according to claim 1, wherein the frequency class features comprise window frequency spectrum weighted average, window frequency spectrum main peak area, window frequency spectrum main peak number and window frequency spectrum standard deviation, and the process for extracting the frequency class features comprises the following steps: selecting a current A-line reflectivity curve through sliding of a preset sliding window to obtain multiple groups of A-line reflectivity curve information; for each set of a-line reflectivity curve information, converting it from the time domain to the frequency domain by fast fourier transform to determine the frequency spectrum of each frequency point in the set, and calculating the frequency spectrum weighted average of the current set of a-line reflectivity curve information by the following formula: Wherein k is the label of each frequency point, N is the total frequency component number, F k is the frequency of k frequency points, F (F k ) is the frequency spectrum corresponding to F k frequency; taking the average value of the frequency spectrum weighted averages of all groups of A-line reflectivity curve information as the window frequency spectrum weighted average of the current A-line reflectivity curve; Determining target frequency points with frequency spectrums larger than a first preset threshold value in the group according to the information of each group of A-line reflectivity curves, and accumulating frequency spectrums corresponding to all the target frequency points in the group to obtain the frequency spectrum main peak area of the current group of A-line reflectivity curve information; Determining the number of frequency points with the frequency spectrum larger than a second preset threshold value in the group according to the information of each group of A-line reflectivity curves to obtain the number of main frequency spectrum peaks of the information of the current group of A-line reflectivity curves; And calculating standard deviation of frequency spectrums in each group of A-line reflectivity curve information to obtain the standard deviation of the frequency spectrums of the current group of A-line reflectivity curve information, and taking the average value of the standard deviations of the frequency spectrums of all groups of A-line reflectivity curve information as the standard deviation of the window frequency spectrums of the current A-line reflectivity curve.
7. The method of claim 1, wherein the frequency class features further comprise an overall frequency spectrum weighted average, an overall frequency spectrum main peak area, an overall frequency spectrum main peak number, and an overall frequency spectrum standard deviation, and the extracting the frequency class features comprises: All information of the current A-line reflectivity curve is converted from a time domain to a frequency domain through fast Fourier transformation to determine a frequency spectrum of each frequency point, and the overall frequency spectrum weighted average of the current A-line reflectivity curve is calculated through the following formula: Wherein k is the label of each frequency point, N is the total frequency component number, F k is the frequency of k frequency points, F (F k ) is the frequency spectrum corresponding to F k frequency; Determining target frequency points with frequency spectrums larger than a first preset threshold value in all information of the current A-line reflectivity curve, and accumulating frequency spectrums corresponding to the target frequency points to obtain the total frequency spectrum main peak area of the current A-line reflectivity curve; Determining the number of frequency points with the frequency spectrum larger than a second preset threshold value in all information of the current A-line reflectivity curve to obtain the number of main peaks of the whole frequency spectrum of the current A-line reflectivity curve; And calculating standard deviation of the frequency spectrum in all information of the current A-line reflectivity curve to obtain the integral frequency spectrum standard deviation of the current A-line reflectivity curve.

Description

Method for evaluating benign and malignant tumor incised margin in operation Technical Field The invention relates to the technical field of medical treatment, in particular to an intraoperative tumor incisional edge benign and malignant evaluation method. Background With the increasing expansion of tumor patient populations, surgical treatment has become a major therapeutic approach. In this process, the accuracy of the intraoperative incisal margin assessment is critical to avoid the patient undergoing secondary surgery, especially the key link to ensure the success of surgery and patient recovery. The negative incisal margin in the operation can be realized, so that the tumor can be thoroughly cleared, the damage to healthy tissues can be effectively reduced, and the postoperative life quality and recovery speed of a patient can be obviously improved. Therefore, the accuracy of the intraoperative incisal margin assessment is enhanced, and the method has extremely important significance for the treatment of tumor patients. Optical coherence tomography (Optical Coherence Tomography, OCT) is a leading-edge non-invasive imaging technique that exploits subtle changes in the light scattering properties of different tissues to reveal the three-dimensional microstructural details of the tissue in a label-free, real-time manner. OCT has unique penetration capability, can penetrate deep about 2mm below the tissue surface while maintaining spatial resolution up to 10-20 microns, thereby generating an image quality similar to the histopathological resolution scale (histopathological resolution can reach 0.25 microns (40X), 0.5 microns (20X), OCT resolution of 10-20 microns, OCT resolution is currently based on OCT (full field OCT resolution of 1 micron). OCT-based tumor assessment techniques have been widely studied and have shown to have a high degree of relevance to histopathological examination. Currently, this technique is mainly used to train different specialists to identify suspicious and non-suspicious regions at the boundary of a specimen in oncology. However, since the OCT scan data acquired intraoperatively is enormous and requires a determination of suspicious regions in real time, relying solely on visual assessment by a clinician is time consuming and laborious. Therefore, an algorithm capable of automatically identifying and marking the suspicious regions is developed, so that the evaluation efficiency can be effectively improved, and the evaluation accuracy can be ensured. The application of such an algorithm will play a key role in the real-time guidance of tumor surgery. Disclosure of Invention Therefore, the invention aims to provide an evaluation method for benign and malignant tumor margin in operation, so as to solve the problems of poor evaluation accuracy and poor real-time performance of the conventional evaluation method for benign and malignant tumor margin in operation. In order to achieve the above purpose, the invention adopts the following technical scheme: the embodiment of the application provides a method for evaluating benign and malignant tumor incisional edge in surgery, which comprises the following steps: Acquiring optical coherence tomography data of historical incisal edge tissue; Extracting an A-line reflectivity curve based on the optical coherence tomography data of the historical incisal edge tissue, wherein the A-line reflectivity curve is a curve of image intensity changing along with the depth of the A-line; Calculating and extracting target characteristics based on the A-line reflectivity curve, wherein the target characteristics comprise depth characteristics, slope characteristics, distance characteristics between local maxima, intensity rapid fluctuation characteristics, frequency characteristics and optical attenuation coefficient characteristics; Training a preset classifier based on the target feature and a benign and malignant label of the A-line reflectivity curve corresponding to the target feature to obtain a trained classifier; And inputting the target characteristics extracted based on the optical coherence tomography data of the incisal edge tissue to be detected into the trained classifier to obtain an evaluation result. Further, in some embodiments of the present application, the depth feature includes a penetration depth of an effective signal interval, and the process of extracting the penetration depth of the effective signal interval includes: performing smoothing treatment on the current A-line reflectivity curve to obtain a current A-line reflectivity smoothing curve; determining an effective signal interval of a current A-line reflectivity smooth curve; and taking the difference between the depths of the starting point and the end point of the effective signal interval as the penetration depth of the effective signal interval of the current A-line reflectivity curve. Further, in some embodiments of the present application, the determining the eff