Risk management method and system for breast interventional ultrasonic operation

Abstract

The invention relates to the technical field of medical care informatics, in particular to a risk management method and system for a mammary gland interventional ultrasonic surgery, and the method comprises the steps: judging whether a tissue elasticity initial risk exists or not through an ultrasonic image and an elasticity image which are synchronized in the surgery; judging whether the initial risk of energy deposition exists or not according to the average temperature change rate of the ultrasonic image energy acting area in the operation and the gray scale gradient value of the anatomical boundary; judging whether a physiological compensation initial risk exists or not according to the relevance between the preoperative heart rate variability signal and the mammary gland capillary blood flow signal; determining similar cases in a historical surgical case database by using the initial risk type and the high-risk area calibration result; and determining the real-time risk probability of risk upgrading in the current operation process by using the similar cases, and carrying out risk alarm. According to the technical scheme, the hysteresis problem of traditional single-threshold early warning is solved, accurate risk intervention opportunity judgment is provided for doctors, and the risk of complications in the operation is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of healthcare informatics, in particular to a risk management method and system for breast interventional ultrasound surgery. BACKGROUND

[0002] Breast interventional ultrasound surgery is a minimally invasive technique for precise diagnosis and treatment of tumors guided by real-time ultrasound images. This technique uses a high-frequency ultrasound probe to locate lesions and combines with instruments such as puncture needles or ablation electrodes to complete tissue sampling or tumor ablation under real-time image monitoring. During the operation, the doctor can dynamically observe the instrument movement trajectory, energy action range and local tissue reaction through the ultrasound image, and its advantages are visualization, small trauma and precise targeting of millimeter-level lesions, providing important intraoperative decision support for clinical practice.

[0003] The existing risk management of breast interventional ultrasound surgery is insufficient, mainly due to the lack of multi-dimensional real-time data fusion and dynamic risk coupling evaluation capability, which cannot identify early malignant events and does not build a dynamic closed-loop feedback mechanism of tissue elasticity deformation-energy deposition-physiological compensation, resulting in risk warning relying on lagging single indicators, which cannot accurately assess the evolution path of intraoperative sudden risks, and ultimately causes risk decision lag and intraoperative safety control blind area. SUMMARY

[0004] In order to solve the technical problems of the existing risk management of breast interventional ultrasound surgery, which is insufficient and difficult to accurately assess the evolution path of intraoperative sudden risks, and is prone to cause risk decision lag and intraoperative safety control blind area, the purpose of the present application is to provide a risk management method and system for breast interventional ultrasound surgery, and the technical solution adopted is as follows:

[0005] The present application provides a risk management method for breast interventional ultrasound surgery, which comprises:

[0006] Using the gray scale distribution histogram and the elastic signal distribution histogram corresponding to the synchronous ultrasound image and the elastic image during the operation respectively, it is judged whether there is an initial risk of tissue elasticity;

[0007] Using the average temperature change rate of the energy action zone of the ablation instrument in the ultrasound image during the operation and the gray scale gradient value of the anatomical boundary therein, it is judged whether there is an initial risk of energy deposition;

[0008] Using the correlation between the heart rate variability signal before the operation and the breast microvascular blood flow signal in the operation area, it is judged whether there is an initial risk of physiological compensation;

[0009] Using the superposition degree between the occurrence position of the initial risk type and the high-risk area calibration result, and combining the initial risk type to determine the similar cases in the historical operation case database;

[0010] Determine the initial risk probability of the current surgery process risk escalation by using similar cases, and obtain a real-time risk probability, and perform risk warning based on the real-time risk probability.

[0011] Further, the use of intraoperative synchronous ultrasound images and elastic imaging corresponding gray scale distribution histogram and elastic signal distribution histogram respectively, to determine whether there is an initial risk of tissue elasticity, comprising:

[0012] The gray scale distribution histogram of the ultrasound image and the elastic signal distribution histogram of the elastic imaging are clustered to obtain each gray scale interval and each elastic signal interval;

[0013] Traverse all combinations of gray scale intervals and elastic signal intervals, and use the distribution probability of the gray scale value of the gray scale interval and the elastic signal of the elastic signal interval in the combination to obtain the gray signal entropy, the elastic signal entropy and the joint entropy;

[0014] Determine the mutual information entropy of the combination using the gray signal entropy, the elastic signal entropy and the joint entropy, and construct the Young's modulus distribution matrix using the combination corresponding to the maximum mutual information entropy;

[0015] Determine the abnormal coefficient of the elastic gradient abnormal area based on the Young's modulus distribution matrix, and use the abnormal coefficient to determine whether there is an initial risk of tissue elasticity in the current surgery process.

[0016] Further, the use of intraoperative synchronous ultrasound images and elastic imaging corresponding gray scale distribution histogram and elastic signal distribution histogram respectively, to determine whether there is an initial risk of tissue elasticity, comprising:

[0017] Calculate the Young's modulus difference of adjacent pixel points based on the Young's modulus distribution matrix to obtain the elastic gradient of the pixel points;

[0018] The pixel points with an elastic gradient greater than a preset multiple standard deviation of the elastic gradient of the same type of tissue before the surgery are regarded as abnormal elastic gradient pixels;

[0019] Aggregate the continuous abnormal elastic gradient pixels into an elastic gradient abnormal area, and use the first area ratio of the elastic gradient abnormal area relative to the lesion edge area and the second area ratio relative to the normal tissue area to obtain the abnormal coefficient thereof;

[0020] If the abnormal coefficient is less than 1, it is determined that there is an initial risk of tissue elasticity in the current surgery process, otherwise it is determined that there is no initial risk of tissue elasticity in the current surgery process.

[0021] Further, the use of intraoperative ultrasound images and elastic imaging corresponding gray scale distribution histogram and elastic signal distribution histogram respectively, to determine whether there is an initial risk of tissue elasticity, comprising:

[0022] Determine the average temperature change rate of the energy action zone by using the temperature change rate of each pixel point in the energy action zone of the ablation instrument in the intraoperative ultrasound image;

[0023] In the case of continuous increase of the average temperature change rate of the energy action zone, obtain the gray scale gradient values of the boundary pixels of the anatomical boundary of the energy action zone in the horizontal and vertical directions respectively;

[0024] Determine whether there is an initial risk of energy deposition in the current surgical process by using the average temperature change rate and the respective gray scale gradient values.

[0025] Further, the determination of whether there is an initial risk of energy deposition in the current surgical process by using the average temperature change rate and the respective gray scale gradient values comprises:

[0026] Obtain the edge strength of the boundary pixels by using the respective gray scale gradient values, and obtain the definition score of the energy action zone by using the edge strength of all boundary pixels;

[0027] If the average temperature change rate of the energy action zone continuously increases and the definition score first decreases, it is determined that there is an initial risk of energy deposition in the current surgical process.

[0028] Further, the determination of whether there is an initial risk of physiological compensation in the current surgical process by using the correlation between the heart rate variability signal before surgery and the breast microvascular blood flow signal in the surgical area comprises:

[0029] Determine the Pearson correlation coefficient between the high-frequency component of the heart rate variability signal before surgery extracted by high-pass filtering and the breast microvascular blood flow signal in the surgical area, and obtain the correlation degree change curve by using the Pearson correlation coefficients in multiple preset analysis periods;

[0030] If the correlation degree change curve is continuously negatively inclined in the preset number of analysis periods and the blood flow rhythm peak value is continuously delayed, it is determined that there is an initial risk of physiological compensation in the current surgical process.

[0031] Further, the determination of similar cases in the historical surgical case database by combining the initial risk type comprises:

[0032] Retrieve the three-dimensional anatomical model of the breast reconstructed by CT before surgery, and obtain the high-risk area demarcation result based on the anatomical structure characteristics in the three-dimensional anatomical model;

[0033] Determine the spatial overlap between the occurrence position of the initial risk type and the high-risk area demarcation result by using the spatial overlap algorithm, and determine the risk occurrence time of the initial risk type;

[0034] Determine similar cases in the historical surgery case database by using the initial risk type, the superposition degree and the risk occurrence time.

[0035] Further, the method comprises the following steps of:

[0036] Determine the initial risk probability of the current surgery process risk escalation by using the similar cases, and obtain a real-time risk probability, and perform risk warning based on the real-time risk probability.

[0037] Determine a plurality of surgery factors occurring from the initial risk type to the occurrence of the display accident, and compare the actual values of the surgery factors with the corresponding standard reference values to determine the risk scene;

[0038] Determine the similar case number ratio of the target initial risk state to other initial risk states in the target risk scene, and take the similar case number ratio as the element value of the transition matrix in the target risk scene;

[0039] Obtain the real-time risk probability by using the initial risk probability and the element value of the transition matrix in the target risk scene, and perform risk warning based on the real-time risk probability.

[0040] Further, the method comprises the following steps of:

[0041] Obtain the transition probability of the target initial risk state to other initial risk states by using the initial risk probability and the element value of the transition matrix in the target risk scene;

[0042] Obtain the updated initial risk probability by using all the transition probabilities in the transition matrix in the target risk scene, obtain the real-time risk probability, and compare the preset safety threshold with the real-time risk probability to perform risk warning.

[0043] The present application also provides a risk management system for breast interventional ultrasound surgery, which is used to implement the risk management method for breast interventional ultrasound surgery as described in any one of the above.

[0044] The risk analysis module is used to determine whether there is an initial risk of tissue elasticity by using the gray scale distribution histogram and the elastic signal distribution histogram corresponding to the ultrasound image and the elastic image during the surgery respectively, determine whether there is an initial risk of energy deposition by using the average temperature change rate of the energy action area of the ablation instrument in the ultrasound image during the surgery and the gray scale gradient value of the anatomical boundary therein, and determine whether there is an initial risk of physiological compensation by using the correlation between the heart rate variability signal before the surgery and the breast microvascular blood flow signal in the surgery region.

[0045] The risk warning module is used for determining similar cases in the historical surgery case database by using the superposition degree between the occurrence position of the initial risk type and the high-risk area calibration result in combination with the initial risk type; determining the initial risk probability of the risk upgrade of the current surgery process by using the similar cases and obtaining a real-time risk probability; and performing risk warning based on the real-time risk probability.

[0046] The present application has the following beneficial effects:

[0047] The present application combines ultrasound image gray mapping and elastic imaging mutual information entropy analysis to construct an intraoperative stable Young's modulus estimation model, effectively identifies abnormal diffusion of tissue elastic gradient, and breaks through the bottleneck of traditional elastic imaging due to operation interference error; synchronously integrates tissue deformation, energy deposition and physiological compensation dynamic data, realizes cross verification of early risks such as carbonization precursor and microvessel injury, further realizes real-time fusion of operation parameters, anatomical condition superposition degree and historical case evolution law based on state transition matrix, predicts risk upgrade path through multi-scene probability iteration, solves the hysteresis problem of traditional single threshold early warning, provides precise risk intervention opportunity judgment for doctors, and reduces intraoperative complication risk. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 A step flowchart of a breast interventional ultrasound surgery risk management method provided by an embodiment of the present application;

[0049] Figure 2 A detailed flowchart of step S1 in the breast interventional ultrasound surgery risk management method provided by an embodiment of the present application;

[0050] Figure 3 A detailed flowchart of step S2 in the breast interventional ultrasound surgery risk management method provided by an embodiment of the present application;

[0051] Figure 4 A detailed flowchart of step S3 in the breast interventional ultrasound surgery risk management method provided by an embodiment of the present application;

[0052] Figure 5 A detailed flowchart of step S4 in the breast interventional ultrasound surgery risk management method provided by an embodiment of the present application;

[0053] Figure 6 A detailed flowchart of step S5 in the breast interventional ultrasound surgery risk management method provided by an embodiment of the present application;

[0054] Figure 7 A structural schematic diagram of a hardware running environment of a breast interventional ultrasound surgery risk management device related to the embodiment scheme of the present application;

[0055] Figure 8 The framework structure diagram of the breast interventional ultrasound surgery risk management system involved in the embodiment of the present application. DETAILED DESCRIPTION

[0056] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined object, the following describes in detail the specific embodiments, structure, features and effects of the breast interventional ultrasound surgery risk management method according to the present application, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0057] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0058] The specific scheme of the breast interventional ultrasound surgery risk management method provided by the present application is specifically described below in combination with the accompanying drawings.

[0059] Embodiment one:

[0060] For the breast interventional ultrasound surgery risk management method provided by the present application, please refer to Figure 1 , which shows the breast interventional ultrasound surgery risk management method step flowchart provided by one embodiment of the present application.

[0061] The breast interventional ultrasound surgery risk management method includes:

[0062] Step S1, using the gray scale distribution histogram and the elastic signal distribution histogram corresponding to the intraoperative synchronous ultrasound image and the elastic imaging respectively, to determine whether there is an initial risk of tissue elasticity;

[0063] In this embodiment, the full amount of historical data of 500 or more breast interventional ultrasound surgeries can be first screened, and two types of invalid cases can be excluded: one type is the case without any risk event record throughout the process, and the other type is the case with more than 30% missing of key monitoring items (such as instrument coordinates, energy parameters, ultrasound image data). Finally, the valid cases containing risk related information are retained as the data basis for developing each of the following embodiments.

[0064] Specifically, please refer to Figure 2 , the step S1 includes:

[0065] Step S11, respectively, the gray scale distribution histogram of the ultrasound image and the elastic signal distribution histogram of the elastic imaging are clustered to obtain each gray scale interval and each elastic signal interval;

[0066] Step S12, all combinations of the gray scale interval and the elastic signal interval are traversed, and the gray scale signal entropy, the elastic signal entropy and the joint entropy are obtained by using the distribution probability of the gray scale value of the gray scale interval and the elastic signal of the elastic signal interval in the combination respectively;

[0067] Step S13, the mutual information entropy of the combination is determined by using the gray scale signal entropy, the elastic signal entropy and the joint entropy, and the Young's modulus distribution matrix is constructed by using the combination corresponding to the maximum mutual information entropy;

[0068] Step S14, the abnormal coefficient of the elastic gradient abnormal region is determined based on the Young's modulus distribution matrix, and whether there is an initial risk of tissue elasticity in the current surgical process is judged by using the abnormal coefficient.

[0069] More specifically, the step S14 comprises:

[0070] The Young's modulus difference of adjacent pixel points is calculated based on the Young's modulus distribution matrix to obtain the elastic gradient of the pixel points;

[0071] The pixel points with the elastic gradient greater than the preset multiple standard deviation of the elastic gradient of the same type of tissue before the operation are regarded as abnormal elastic gradient pixels;

[0072] The continuous abnormal elastic gradient pixels are aggregated into an elastic gradient abnormal region, and the abnormal coefficient thereof is obtained by using the first area ratio of the elastic gradient abnormal region with respect to the lesion edge region and the second area ratio of the elastic gradient abnormal region with respect to the normal tissue region;

[0073] If the abnormal coefficient is less than 1, it is determined that there is an initial risk of tissue elasticity in the current surgical process, otherwise it is determined that there is no initial risk of tissue elasticity in the current surgical process.

[0074] In this embodiment, the ultrasound elastic imaging data (i.e. elastic imaging) under the static scanning of the breast surgery region in each case data is obtained, the anatomical boundaries of fat, breast gland, lesion and surrounding normal tissue are captured by edge segmentation, and the number of tissue regions m is obtained;

[0075] The Young's modulus of each tissue before the operation maintains the baseline level, there is no elastic gradient abnormal region, and the risk is in a latent state; therefore, more than 5 typical sampling points (avoiding blood vessels and calcification points) are selected for each type of tissue, the average Young's modulus value in each type of tissue region is calculated by repeatedly measuring and eliminating the maximum value and the minimum value, and the baseline Young's modulus of different tissue types is obtained.

[0076] Then when the puncture, ablation operation starts, the instrument will exert pressure or energy on the tissue, the local tissue Young's modulus deviates from the baseline, if sporadic abnormal pixels represent the risk may begin to emerge at this time;

[0077] But when the operation begins, it will destroy the stable tissue environment of ultrasound elastography, making the information readability of ultrasound elastography greatly reduced, and a large number of false abnormalities appear; while the ultrasound image is relatively stable, the embodiment simultaneously utilizes the ultrasound image and the ultrasound elastography to map the stable ultrasound image gray value to the elastic signal estimation, so that the elastic abnormality of the tissue can be stably found during the operation;

[0078] Synchronous ultrasound image and elastic imaging data after the operation are acquired, both of which are subjected to probe offset calibration to avoid unnecessary error problems;

[0079] Then the gray distribution histogram of the ultrasound image is acquired; the elastic signal (Young's modulus) distribution histogram of the elastic imaging is acquired; and according to the number of tissue regions in the breast operation area, all gray levels of the gray distribution histogram are subjected to k-means clustering to obtain m gray intervals, and the elastic signal distribution histogram is also divided into m elastic signal intervals by k-means clustering;

[0080] All combinations of the gray intervals and the elastic signal intervals are traversed, that is, any one gray interval is made to correspond to any one elastic signal interval;

[0081] Then the distribution probability of all gray values in each gray interval in the ultrasound image is counted, and then the information entropy (gray signal entropy) in the gray interval is calculated; similarly, the distribution probability of all elastic signals in each elastic signal interval in the real-time elastic imaging is obtained, and then the information entropy (elastic signal entropy) in the elastic signal interval is calculated;

[0082] Then for all "gray interval-elastic signal interval" combinations, the joint probability of all gray values in the gray interval and all elastic signal values (Young's modulus) in the elastic signal interval is calculated, and then the joint (information) entropy of each "gray interval-elastic signal interval" combination is calculated, which reflects the uncertainty of the joint occurrence of the gray interval and the elastic signal interval;

[0083] Then the mutual information entropy of each "gray interval-elastic signal interval" combination = joint entropy - (gray signal entropy + elastic signal entropy) is calculated, which is used to represent the correlation degree between the above-mentioned gray value and the elastic signal;

[0084] Each combination of the gray interval-elastic signal interval can output a mutual information entropy, the elastic signal is fixed as a variable, and the combination corresponding to the maximum mutual information entropy in all combination modes of each gray interval is selected as the target elastic signal interval corresponding to the gray interval.

[0085] All gray scale interval-target elasticity signal interval combinations can construct a Young's modulus distribution matrix;

[0086] The purpose of constructing the matrix is to convert the gray value in the stable ultrasound image into the Young's modulus that can reflect the elasticity of the tissue, so as to avoid the problem that the elasticity imaging information is not reliable during the operation;

[0087] Based on the real-time Young's modulus distribution matrix, the Young's modulus difference value of adjacent pixel points, i.e. the elasticity gradient, is calculated, and the pixel points whose difference value exceeds twice (preset multiple, which can be adjusted specifically) of the standard deviation of the elasticity of the same type of tissue (such as the lesion edge region) before the operation are called abnormal elasticity gradient pixels (simply referred to as "abnormal pixels");

[0088] During the operation, the abnormal pixels form a continuous abnormal region through the elasticity gradient diffusion, and the region growing algorithm (prior art) is used to aggregate the spatially continuous abnormal pixels into a complete elasticity gradient abnormal region (simply referred to as "abnormal region"), and the tissue type to which the abnormal region belongs is marked by comparing the anatomical boundary before the operation, such as lesion edge region abnormality, fat-gland boundary region abnormality, etc.

[0089] The first area ratio S1 of the abnormal region in the lesion edge region in the continuous frames is calculated, and the second area ratio S2 of the abnormal region in the normal tissue region in the continuous frames is calculated;

[0090] The abnormal coefficient = S1 / S2, i.e. the ratio of the first area ratio to the second area ratio.

[0091] If the abnormal coefficient is greater than or equal to 1, it means that the area ratio of the abnormal region in the lesion edge region is higher, and the area ratio in the normal tissue region is lower, so it is determined that there is no initial risk of tissue elasticity in the current operation process, but it is prompted that the risk is more likely to come from the lesion infiltration;

[0092] If the abnormal coefficient is less than 1, it means that the area ratio of the abnormal region in the normal tissue region is higher, so there is an initial risk of tissue elasticity and it is prompted that the risk is more likely to come from the tissue damage caused by the operation.

[0093] Step S2, using the average temperature change rate of the energy action zone of the ablation instrument in the intraoperative ultrasound image and the gray scale gradient value of the anatomical boundary therein, to determine whether there is an initial risk of energy deposition;

[0094] Specifically, please refer to Figure 3 , the step S2 comprises:

[0095] Step S21, using the temperature change rate of each pixel point in the energy action zone of the ablation instrument in the intraoperative ultrasound image to determine the average temperature change rate of the energy action zone;

[0096] Step S22, in the case of a sustained increase in the average temperature change rate of the energy action zone, obtaining the gray scale gradient values of the boundary pixel points of the energy action zone in the anatomical boundary in the horizontal and vertical directions, respectively;

[0097] Step S23, using the average temperature change rate and the respective gray scale gradient values, determining whether there is an initial risk of energy deposition in the current surgical process.

[0098] More specifically, the step S23 includes:

[0099] Using the respective gray scale gradient values to obtain the edge strength of the boundary pixel points, and using the edge strength of all the boundary pixel points to obtain the definition score of the energy action zone;

[0100] If the average temperature change rate of the energy action zone continues to increase and the definition score first decreases, it is determined that there is an initial risk of energy deposition in the current surgical process.

[0101] In this embodiment, the position of the electrode tip of the ablation instrument is identified by ultrasonic imaging. A circular energy action zone is set according to the natural diffusion range of the energy of the ablation instrument with the tip as the center. The energy action zone is marked on the ultrasonic image by an image superposition algorithm (prior art), and the anatomical structure of the energy action zone, including the gland duct, microvessel, etc., is recorded synchronously, which is called sensitive structure.

[0102] Then the acquisition frequency of the ultrasonic thermal imaging module is adjusted to align with the energy output timing of the ablation instrument, ensuring that a thermal imaging image is acquired every 0.5 seconds, synchronized with the energy output state.

[0103] The temperature change rate of the energy action zone is calculated. Based on the synchronized thermal imaging image, the temperature values of the same pixel points in two consecutive images are calculated by difference, and then divided by the interval time between the two frames to obtain the temperature change rate of the pixel points. Finally, the average temperature change rate of the entire energy action zone is calculated by regional averaging algorithm.

[0104] When the temperature change rate of the energy action zone further increases, the gray value rises sharply and the definition decreases significantly, which may indicate a precursor of carbonization. Carbonized tissue forms a blind area in ultrasonic imaging, on the one hand, and increases the risk of needle channel adhesion, on the other hand, which increases the risk.

[0105] Therefore, the gray value and definition of the ultrasonic image are monitored synchronously for the region with a sustained increase in the temperature change rate.

[0106] The pixel points of the anatomical boundary (such as the lesion edge and the blood vessel contour) in the image are extracted by the Canny edge detection algorithm, the gray scale gradient values of each boundary pixel point in the horizontal and vertical directions are calculated, the square root of the sum of the squares of the two is taken as the edge intensity of the point, and the arithmetic mean of the edge intensities of all boundary pixel points is taken to obtain the definition score, and the higher the value represents the clearer the boundary.

[0107] If the definition score is lower than before the temperature rise, it can be further confirmed that there is a carbonization precursor, that is, when the definition score first decreases, it can be determined that there is an initial risk of energy deposition in the current surgical process;

[0108] Further, when the carbonization area overlaps with the microvessel, it represents that the thermal injury may cause the risk of microvessel rupture, and the image blind area may cause the doctor to be unable to accurately control the instrument, further increasing the risk of energy deposition and tissue damage;

[0109] The carbonization precursor recognition result is compared with the anatomical structure annotation result of the energy action area to determine whether the carbonization area contains sensitive structures such as microvessels and gland ducts;

[0110] If the above sensitive structures are contained, the overlapping area of the carbonization area and the sensitive structure is further calculated: the contours of the carbonization area and the sensitive structure (microvessel, gland duct) are binarized, the total number of pixels in the overlapping area is counted, and then the pixel number is converted into the actual anatomical overlapping area according to the inherent pixel-actual size conversion ratio of the ultrasound image;

[0111] If the temperature change rate continues to rise (the temperature change slope of the continuous frame is positive) and the overlapping area expands (the overlapping area change slope of the continuous frame is positive), it indicates that the damage risk is increasing;

[0112] Step S3, using the correlation between the heart rate variability signal before the operation and the breast microvessel blood flow signal in the operation area to determine whether there is an initial risk of physiological compensation;

[0113] Specifically, please refer to Figure 4 , the step S3 comprises:

[0114] Step S31, determine the Pearson correlation coefficient between the high-frequency component of the heart rate variability signal before the operation extracted by the high-pass filter and the breast microvessel blood flow signal in the operation area, and obtain the correlation degree change curve by using the Pearson correlation coefficients in multiple preset analysis periods;

[0115] Step S32, if the correlation degree change curve is continuously negatively inclined in a preset number of analysis periods and the blood flow rhythm peak is continuously delayed, it is determined that there is an initial risk of physiological compensation in the current surgical process.

[0116] In this embodiment, the heart rate variability (HRV) signal and the breast microvascular blood flow signal in the stable state of the patient 10 minutes before the operation are collected, the high frequency component (high pass filtering extraction, generally 0.15 Hz to 0.40 Hz) of the HRV is calculated through a sliding window with a length of 1 minute to reflect the vagus nerve activity, and the Pearson correlation coefficient between the high frequency component of the HRV and the microvascular blood flow pulsation cycle in the sliding window is calculated as the correlation (degree) between the two, the above calculation is repeated 5 times to take the stable value as the basic correlation reference of physiological compensation.

[0117] With 5 minutes as an analysis period, the correlation degree is calculated repeatedly 3 times (1st, 3rd, and 5th minutes) in each period, 3 period (multiple preset analysis periods, which can be adjusted specifically) correlation degree values are obtained, a correlation degree change curve is generated through linear fitting, and the slope of the curve is observed. If the slope changes from positive to negative, it means that the correlation degree changes from stable to decline, and the slope remains negative and the absolute value continuously increases for 2 consecutive (preset number, which can be adjusted specifically) periods, which indicates that the neuro-vascular compensation coordination starts to be unbalanced.

[0118] For the period with a declining correlation degree, the time domain curve of the microvascular blood flow velocity in the corresponding period is extracted, the blood flow velocity fluctuation rhythm is identified through a peak detection algorithm, and the rhythm should be consistent with the HRV fluctuation in normal compensation. If the blood flow rhythm peak appears 5 times in succession with a delay compared with the HRV fluctuation, it further verifies the destruction of the synchronization of neuro-vascular compensation regulation.

[0119] Summary: If the negative slope of the correlation degree change curve for 2 consecutive periods and the peak delay of the blood flow rhythm for 5 consecutive times are met at the same time, it is determined that there is an initial risk event of abnormal physiological compensation, i.e., an initial risk of physiological compensation.

[0120] Step S4, using the superposition degree between the occurrence position of the initial risk type and the high risk area demarcation result, combining the initial risk type to determine similar cases in the historical operation case database;

[0121] Combined with the monitoring results of the previous three types of initial risks (tissue elasticity initial risk, energy deposition initial risk, and physiological compensation initial risk), the intraoperative initial risks in all cases are identified, and the summary is as follows:

[0122] a. If the elastic abnormal area in the tissue elasticity monitoring spreads across the tissue boundary, i.e., the abnormal coefficient is less than 1, it is determined as a tissue elasticity initial risk;

[0123] b. If the temperature change rate continuously increases and the clarity first decreases in the energy deposition monitoring, it represents a precursor of carbonization, and it is determined as an energy deposition initial risk;

[0124] c. If the correlation degree change curve is negatively inclined for 2 consecutive periods and the blood flow rhythm peak value is delayed for 5 consecutive times in the physiological compensation monitoring, it is determined that there is an initial risk of physiological compensation;

[0125] For each type of initial risk event, record the occurrence time, the anatomical location and the associated operation type to form an initial risk profile, which provides starting point data for subsequent risk accumulation analysis.

[0126] Specifically, referring to Figure 5 , the step S4 comprises:

[0127] Step S41, retrieve the breast three-dimensional anatomical model reconstructed by preoperative CT, and obtain high-risk area labeling results based on the anatomical structure characteristics in the model;

[0128] Step S42, determine the superposition degree between the occurrence position of the initial risk type and the high-risk area labeling results by using a spatial superposition degree algorithm, and determine the risk occurrence time of the initial risk type;

[0129] Step S43, determine similar cases in the historical surgery case database by using the initial risk type, the superposition degree and the risk occurrence time.

[0130] In this embodiment, the breast three-dimensional anatomical model reconstructed by preoperative CT (Computed Tomography) is retrieved, and high-risk area labeling results are obtained based on the anatomical structure characteristics, such as the lung tissue adjacent area defined based on the natural adjacency relationship between the lower boundary of the breast and the lung tissue, the large blood vessel dense area defined based on the blood vessel distribution labeled by preoperative ultrasound, the nerve running area defined based on the matching of anatomical atlas and preoperative image, etc., to provide spatial basis for subsequent judgment of whether the initial risk is superimposed with high-risk conditions;

[0131] Based on the a and b risk occurrence positions in the initial risk profile and the high-risk area labeling results, the spatial superposition degree algorithm is used to calculate the superposition degree (prior art): first, the initial risk occurrence area and the high-risk area are converted into polygons in the same coordinate system, and then the ratio of the overlapping area of the two to the area of the initial risk area is calculated, which is the superposition degree;

[0132] The higher the superposition degree, the higher the risk accumulation intensity;

[0133] For example, when the needle deviation initial risk occurrence area has a high superposition degree with the lung tissue adjacent area (assuming 40%), it indicates that the initial risk has been partially superimposed with high-risk conditions, and the risk is relatively high;

[0134] According to the identified risk type, if it is identified as a, the superposition degree of a risk is obtained, and if it is identified as b, the superposition degree of b risk is obtained; at the same time, the a and b risk occurrence times after the start of the operation are obtained;

[0135] If c is, only the c risk occurrence time after the operation starts is obtained;

[0136] Different initial risk events can be extracted for different operation cases;

[0137] Dynamic deduction of risk upgrade probability is performed to perform risk warning:

[0138] During the current operation process, after triggering any of the above risk events, the historical similar operation case database is called to screen similar cases close to the current "initial risk type, superposition degree, and risk occurrence time" (all of which can be converted into vectors to obtain the vector similarity, such as cosine similarity, prior art), and the proportion of initial risk upgrade to explicit accidents (bleeding, pneumothorax, and tissue necrosis) in the case is counted.

[0139] Step S5, using similar cases to determine the initial risk probability of risk upgrade in the current operation process and obtaining a real-time risk probability, and performing risk warning based on the real-time risk probability.

[0140] Specifically, please refer to Figure 6 , the step S5 comprises:

[0141] Step S51, determining the proportion of the similar cases from the initial risk type upgrade to the explicit accidents as the initial risk probability of risk upgrade in the current operation process and the corresponding initial risk state;

[0142] Step S52, determining a plurality of operation factors from the initial risk type occurrence to the appearance of the explicit accident, and comparing the actual value of the operation factor with the corresponding standard reference value to determine the risk scene;

[0143] Step S53, determining the similar case number ratio from the target initial risk state to other initial risk states in the target risk scene, and taking the similar case number ratio as the element value of the transition matrix in the target risk scene;

[0144] Step S54, obtaining the real-time risk probability by using the initial risk probability and the element value of the transition matrix in the target risk scene, and performing risk warning based on the real-time risk probability.

[0145] More specifically, the step S54 comprises:

[0146] Using the initial risk probability and the element value of the transition matrix in the target risk scene to obtain the transition probability from the target initial risk state to other initial risk states;

[0147] The initial risk probability is updated by using all transition probabilities corresponding to the target risk scenario in the transition matrix to obtain a real-time risk probability, and the real-time risk probability is compared with a preset safety threshold to perform risk alarm.

[0148] Based on the above various embodiments, after the proportion of the initial risk upgrading to the explicit accident (bleeding, pneumothorax, tissue necrosis) in the statistical cases is obtained, the proportion is taken as the initial risk probability of the current surgery process risk upgrading;

[0149] Then, based on the cases, a plurality of surgery factors (the more, the better) between the occurrence of the initial risk and the occurrence of the explicit accident are sorted, such as surgery operation, surgery instrument parameter, patient physiological data, and the like, and the state transition of the initial probability is performed:

[0150] First, five (initial) risk states are defined, and the initial risk probability interval 0-100% is divided into a risk level every 20%: S0 basic risk (0-20%); S1 low risk upgrading (20%-40%); S2 medium risk upgrading (40%-60%); S3 high risk upgrading (60%-80%); and S4 explicit accident (80%-100%);

[0151] According to the initial risk probability of the current surgery process, the initial risk level to which the current surgery process belongs is determined, and is recorded as ;

[0152] Then, according to the abnormal degree of the factors (i.e., the difference degree of the actual values of the surgery factors compared with the standard reference values of the factors), three risk scenarios are divided: normal, mild abnormality, and significant abnormality; and the abnormal degree of different factors directly changes the risk state transition rule;

[0153] Then, the similar case number ratio of the risk state (as the target initial risk state) to the risk state (as other initial risk states) under each risk scenario (normal, mild abnormality, and significant abnormality) is respectively counted, the ratio is taken as the value of the corresponding position element in the (state) transition matrix, and three risk scenario exclusive transition matrices are formed;

[0154] In the matrix, the horizontal axis and the vertical axis are S0, S1, S2, S3, and S4; the elements in the matrix exist (the element in the i-th row and the j-th column), and also exist (the element in the j-th row and the j-th column), which can be expressed in one of the two forms; that is, there is risk upgrading, and there is risk degradation;

[0155] Then, taking the initial risk probability as the starting point, the factor data is collected and matched with the corresponding abnormal scenario every period (10s), and the transition matrix of the scenario is called;

[0156] The initial risk can be transformed from the original risk state to any other risk state;

[0157] The initial risk probability is multiplied by the transition probability of each row in the transition matrix That is, the transition probability of the initial risk to any other risk state, all transition probabilities are accumulated, and the initial risk probability is updated to obtain the real-time risk probability;

[0158] Then, in the next cycle (10s), the state transition matrix is updated according to the real-time collected factor data, and the updated state transition matrix and risk probability are used to continue iterative updating, and the process is repeated to realize real-time prediction of surgical risk;

[0159] When the predicted real-time risk probability exceeds the preset safety threshold, such as more than 50% (medium risk continuous stage, which can be adjusted specifically), risk warning is performed.

[0160] The present application combines ultrasound image gray mapping and elastic imaging mutual information entropy analysis to construct an intraoperative stable Young's modulus estimation model, effectively identifies abnormal diffusion of tissue elastic gradient, and breaks through the bottleneck of traditional elastic imaging due to operation interference error; Synchronously integrate tissue deformation, energy deposition and physiological compensation dynamic data, realize cross verification of early risks such as carbonation precursor and microvascular injury; Further based on the state transition matrix, real-time fusion of surgical operation parameters, anatomical condition superposition degree and historical case evolution law, through multi-scene probability iterative prediction of risk upgrading path, solve the hysteresis problem of traditional single threshold early warning, provide precise risk intervention opportunity judgment for doctors, and reduce the risk of intraoperative complications.

[0161] Embodiment two:

[0162] The present application also provides a breast interventional ultrasound surgery risk management device. The device can be a computer, a server or a combination of multiple data analysis and calculation devices.

[0163] As shown in Figure 7 , the breast interventional ultrasound surgery risk management device of the present application is a hardware running environment structure diagram. Figure 7

[0164] As shown in Figure 7 ​As shown, the breast interventional ultrasound surgery risk management device can include a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. The communication bus 1002 is used to realize the connection communication between the components. The user interface 1003 can include a display, an input unit such as a control panel, and the optional user interface 1003 can also include a standard wired interface, a wireless interface. The network interface 1004 can optionally include a standard wired interface, a wireless interface (such as a WIFI interface). The memory 1005 can be a high-speed RAM memory or a stable memory (non-volatile memory) such as a disk memory. The memory 1005 can also be an independent storage device from the aforementioned processor 1001. The memory 1005 as a computer storage medium can include the breast interventional ultrasound surgery risk management program.

[0165] Those skilled in the art can understand that the hardware structure shown in the foregoing embodiments is not a limitation on the device, and the device can include more or fewer components than those shown in the drawings, or combine certain components, or different component arrangements. Figure 7

[0166] With reference to the foregoing description of the breast interventional ultrasound surgery risk management method, the breast interventional ultrasound surgery risk management device can be used to implement the breast interventional ultrasound surgery risk management method. Figure 7 Figure 7 The memory 1005 as a computer readable storage medium can include an operating system, a user interface module, a network communication module, and a breast interventional ultrasound surgery risk management program.

[0167] In the foregoing description of the breast interventional ultrasound surgery risk management method, the network communication module is mainly used to connect to a server and can communicate data with the server; and the processor 1001 can call the breast interventional ultrasound surgery risk management program stored in the memory 1005 and execute the steps in the foregoing various embodiments. Figure 7 Based on the hardware structure of the breast interventional ultrasound surgery risk management device described above, the various embodiments of the breast interventional ultrasound surgery risk management method of the present application are implemented.

[0168] In addition, the present application also provides a breast interventional ultrasound surgery risk management system, please refer to

[0169] The breast interventional ultrasound surgery risk management system includes: Figure 8

[0170] ​​​The risk analysis module A10 is configured to determine whether there is an initial risk of tissue elasticity by using the gray scale distribution histogram and the elastic signal distribution histogram corresponding to the intraoperative synchronous ultrasound image and the elastography respectively; determine whether there is an initial risk of energy deposition by using the average temperature change rate of the energy action zone of the ablation device in the intraoperative ultrasound image and the gray scale gradient value of the anatomical boundary therein; and determine whether there is an initial risk of physiological compensation by using the correlation between the heart rate variability signal before the operation and the breast microvascular blood flow signal in the operation region.

[0171] The risk warning module A20 is configured to determine similar cases in the historical operation case database by using the superposition degree between the occurrence position of the initial risk type and the high-risk area calibration result in combination with the initial risk type; determine the initial risk probability of the risk escalation in the current operation process and obtain a real-time risk probability by using the similar cases; and perform risk warning based on the real-time risk probability.

[0172] Further, the risk analysis module A10 is further configured to:

[0173] cluster the gray scale distribution histogram of the ultrasound image and the elastic signal distribution histogram of the elastography to obtain respective gray scale intervals and respective elastic signal intervals;

[0174] traverse all combinations of the gray scale intervals and the elastic signal intervals, and obtain the gray scale signal entropy, the elastic signal entropy, and the joint entropy by using the respective distribution probabilities of the gray scale values of the gray scale intervals and the elastic signal values of the elastic signal intervals in the combinations respectively;

[0175] determine the mutual information entropy of the combinations by using the gray scale signal entropy, the elastic signal entropy, and the joint entropy, and construct a Young's modulus distribution matrix by using the combination corresponding to the maximum mutual information entropy;

[0176] determine the abnormal coefficient of the elastic gradient abnormal region based on the Young's modulus distribution matrix, and determine whether there is an initial risk of tissue elasticity in the current operation process by using the abnormal coefficient.

[0177] Further, the risk analysis module A10 is further configured to:

[0178] calculate the Young's modulus difference of adjacent pixel points to obtain the elastic gradient of the pixel points based on the Young's modulus distribution matrix;

[0179] regard the pixel points with the elastic gradient greater than the preset multiple standard deviation of the elastic gradient of the same type of tissue before the operation as abnormal elastic gradient pixels;

[0180] aggregate the continuous abnormal elastic gradient pixels into an elastic gradient abnormal region, and obtain the abnormal coefficient of the elastic gradient abnormal region by using the first area ratio of the elastic gradient abnormal region with respect to the lesion edge region and the second area ratio of the elastic gradient abnormal region with respect to the normal tissue region respectively;

[0181] If the abnormality coefficient is less than 1, it is determined that there is an initial risk of tissue elasticity in the current surgical process, otherwise it is determined that there is no initial risk of tissue elasticity in the current surgical process.

[0182] Further, the risk analysis module A10 is further configured to:

[0183] Determine the average temperature change rate of the energy action zone by using the temperature change rate of each pixel point in the energy action zone of the ablation instrument in the intraoperative ultrasound image.

[0184] In the case that the average temperature change rate of the energy action zone continues to rise, obtain the respective gray scale gradient values of the boundary pixels of the anatomical boundary of the energy action zone in the horizontal and vertical directions.

[0185] Determine whether there is an initial risk of energy deposition in the current surgical process by using the average temperature change rate and the respective gray scale gradient values.

[0186] Further, the risk analysis module A10 is further configured to:

[0187] Obtain the edge strength of the boundary pixels by using the respective gray scale gradient values, and obtain the definition score of the energy action zone by using the edge strength of all boundary pixels.

[0188] If the average temperature change rate of the energy action zone continues to rise and the definition score first decreases, it is determined that there is an initial risk of energy deposition in the current surgical process.

[0189] Further, the risk analysis module A10 is further configured to:

[0190] Determine the Pearson correlation coefficient between the high-frequency component of the preoperative heart rate variability signal extracted by high-pass filtering and the breast microvascular blood flow signal in the surgical area, and obtain the correlation degree change curve by using the Pearson correlation coefficients in a plurality of preset analysis periods.

[0191] If the correlation degree change curve is continuously negatively inclined in a preset number of analysis periods and the blood flow rhythm peak value is continuously delayed, it is determined that there is an initial risk of physiological compensation in the current surgical process.

[0192] Further, the risk alarm module A20 is further configured to:

[0193] Retrieve the breast three-dimensional anatomical model reconstructed by preoperative CT, and obtain the high-risk area labeling result based on the labeling of the high-risk condition area according to the anatomical structure characteristics.

[0194] Determine the superposition degree between the occurrence position of the initial risk type and the high-risk area labeling result by using the spatial superposition degree algorithm, and determine the risk occurrence time of the initial risk type.

[0195] Determine similar cases in the historical surgery case database by using the initial risk type, the superposition degree, and the risk occurrence time.

[0196] Further, the risk alarm module A20 is further used for:

[0197] Determine the proportion of similar cases upgraded from the initial risk type to the explicit accident as the initial risk probability and the corresponding initial risk state of the current surgery process risk upgrade.

[0198] Determine a plurality of surgery factors from the initial risk type to the occurrence of the explicit accident, and compare the actual value of the surgery factor with the corresponding standard reference value to determine the risk scene;

[0199] Determine the proportion of similar cases from the target initial risk state to other initial risk states in the target risk scene as the element value of the transition matrix in the target risk scene.

[0200] Obtain the real-time risk probability by using the initial risk probability and the element value of the transition matrix in the target risk scene, and perform risk alarm based on the real-time risk probability.

[0201] Further, the risk alarm module A20 is further used for:

[0202] Obtain the transition probability from the target initial risk state to other initial risk states by using the initial risk probability and the element value of the transition matrix in the target risk scene.

[0203] Obtain the real-time risk probability by using the updated initial risk probability obtained by using all transition probabilities in the transition matrix in the target risk scene, and compare the preset safety threshold with the real-time risk probability to perform risk alarm.

[0204] The specific implementation of the breast interventional ultrasound surgery risk management system is basically the same as that of the breast interventional ultrasound surgery risk management method, and will not be repeated here.

[0205] In addition, the present application also provides a computer readable storage medium. The breast interventional ultrasound surgery risk management program is stored on the computer readable storage medium of the present application, wherein the breast interventional ultrasound surgery risk management program is executed by the processor to realize the steps of the breast interventional ultrasound surgery risk management method as described above.

[0206] The method realized by the breast interventional ultrasound surgery risk management program when executed can refer to each embodiment of the breast interventional ultrasound surgery risk management method of the present application, and will not be repeated here.

[0207] It is to be understood that the foregoing description is that of only one implementation of the application. Various modifications can be made to the implementation described and illustrated herein, without departing from the spirit and scope of the application, as will be apparent to those skilled in the art from this disclosure. For example, the order of steps can be different from that described and illustrated herein. Further, other steps can be provided, or steps can be eliminated, from the described implementation.

[0208] Various embodiments in this disclosure are described in progressive manner, and the same or similar parts between various embodiments can be referred to each other, and each embodiment focuses on the difference from other embodiments.

[0209] Those skilled in the art will appreciate that embodiments of the application can be devised for a method, a system, or a computer program product. Accordingly, the present application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage media, etc.) embodying computer readable program code.

[0210] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are only by way of example and are not limiting of the scope of the application. Numerous other embodiments can be devised which do not depart from the spirit and scope of the present application.

Claims

1. A risk management method for breast interventional ultrasound surgery, characterized in that, The method includes: By using the grayscale distribution histograms and elastic signal distribution histograms corresponding to the ultrasound images and elastography images simultaneously during surgery, respectively, the existence of initial risk to tissue elasticity can be determined. By using the average temperature change rate of the energy action zone of the ablation device and the gray-scale gradient value of the anatomical boundary in the intraoperative ultrasound imaging, the presence of initial risk of energy deposition can be determined. By utilizing the correlation between preoperative heart rate variability signals and breast microvascular blood flow signals in the surgical area, it can be determined whether there is an initial risk of physiological compensation. By utilizing the overlap between the location of the initial risk type and the high-risk area identification results, similar cases in the historical surgical case database can be identified in combination with the initial risk type. By using similar cases, the initial risk probability of escalation in the current surgical procedure is determined and the real-time risk probability is obtained. Risk alerts are then issued based on the real-time risk probability.

2. The risk management method for breast interventional ultrasound surgery according to claim 1, characterized in that, The method of using the grayscale distribution histograms and elastic signal distribution histograms corresponding to synchronous ultrasound images and elastography during surgery, respectively, to determine whether there is an initial risk of tissue elasticity loss includes: Clustering was performed on the gray-level distribution histogram of ultrasound images and the elastic signal distribution histogram of elastography to obtain each gray-level interval and each elastic signal interval. By traversing all combinations of grayscale intervals and elastic signal intervals, the grayscale signal entropy, elastic signal entropy, and joint entropy are obtained by using the respective distribution probabilities of the grayscale values ​​in the grayscale intervals and the elastic signals in the elastic signal intervals in the combinations. The mutual information entropy of the combination is determined by using the gray signal entropy, elastic signal entropy and joint entropy, and the Young's modulus distribution matrix is ​​constructed by using the combination corresponding to the largest mutual information entropy; The abnormal coefficients of the elastic gradient abnormal region are determined based on the Young's modulus distribution matrix, and the abnormal coefficients are used to determine whether there is an initial risk of tissue elasticity during the current surgical procedure.

3. The risk management method for breast interventional ultrasound surgery according to claim 2, characterized in that, The method of determining the abnormal coefficients of the elastic gradient abnormality region based on the Young's modulus distribution matrix, and using the abnormal coefficients to determine whether there is an initial risk of tissue elasticity during the current surgical procedure, includes: The elastic gradient of a pixel is obtained by calculating the difference in Young's modulus between adjacent pixels based on the Young's modulus distribution matrix. Pixels whose elastic gradient is greater than a preset multiple of the standard deviation of the elastic gradient of similar tissues before surgery are identified as abnormal elastic gradient pixels. Continuous abnormal elastic gradient pixels are aggregated into an abnormal elastic gradient region. The abnormal coefficient is obtained by using the first area ratio of the abnormal elastic gradient region relative to the lesion edge region and the second area ratio of the normal tissue region. If the abnormality coefficient is less than 1, it is determined that there is an initial risk of tissue elasticity during the current surgical procedure; otherwise, it is determined that there is no initial risk of tissue elasticity during the current surgical procedure.

4. The risk management method for breast interventional ultrasound surgery according to claim 1, characterized in that, The method of determining the existence of initial risk of energy deposition by utilizing the average temperature change rate of the energy-acting area of ​​the ablation device in intraoperative ultrasound imaging and the grayscale gradient value of the anatomical boundary therein includes: The average temperature change rate of the energy-affected area was determined by using the temperature change rate of each pixel in the energy-affected area of ​​the ablation instrument in the ultrasound image during surgery. As the average temperature change rate of the energy-affected region continues to increase, the grayscale gradient values ​​of the boundary pixels of the anatomical boundary of the energy-affected region in the horizontal and vertical directions are obtained. By using the average temperature change rate and their respective grayscale gradient values, it is determined whether there is an initial risk of energy deposition during the current surgical procedure.

5. The risk management method for breast interventional ultrasound surgery according to claim 4, characterized in that, The method of determining whether there is an initial risk of energy deposition during the current surgical procedure by using the average temperature change rate and their respective grayscale gradient values ​​includes: The edge intensity of the boundary pixels is obtained by using their respective grayscale gradient values, and the sharpness score of the energy zone is obtained by using the edge intensity of all boundary pixels. If the average temperature change rate of the energy-affected area continues to increase and the clarity score shows a decrease for the first time, it is determined that there is an initial risk of energy deposition during the current surgical procedure.

6. The risk management method for breast interventional ultrasound surgery according to claim 1, characterized in that, The method of using the correlation between preoperative heart rate variability signals and breast microvascular blood flow signals in the surgical area to determine whether there is an initial risk of physiological compensation includes: The Pearson correlation coefficient between the high-frequency components of the preoperative heart rate variability signal extracted by high-pass filtering and the microvascular blood flow signal of the breast in the surgical area was determined, and the correlation change curve was obtained by using the Pearson correlation coefficients in multiple preset analysis periods. If the correlation curve tilts negatively continuously within a preset number of analysis periods and the peak of blood flow rhythm appears continuously with a delay, it is determined that there is an initial risk of physiological compensation during the current surgical procedure.

7. The risk management method for breast interventional ultrasound surgery according to claim 1, characterized in that, The method of utilizing the overlap between the location of the initial risk type and the high-risk area identification results, combined with the initial risk type, to determine similar cases in the historical surgical case database includes: The three-dimensional anatomical model of the breast reconstructed from the preoperative CT scan was retrieved, and high-risk condition areas were marked based on the anatomical structural characteristics to obtain the high-risk area labeling results; The spatial overlap algorithm is used to determine the degree of overlap between the location of the initial risk type and the high-risk area calibration results, thereby determining the risk occurrence time of the initial risk type. By utilizing the initial risk type, overlap, and time of risk occurrence, similar cases in the historical surgical case database are identified.

8. The risk management method for breast interventional ultrasound surgery according to claim 1, characterized in that, The process of determining the initial risk probability of escalation in the current surgical procedure using similar cases and obtaining the real-time risk probability, and then issuing a risk alert based on the real-time risk probability, includes: The proportion of similar cases that escalate from the initial risk type to an explicit accident is determined as the initial risk probability and corresponding initial risk status of the current surgical procedure risk escalation; Identify multiple surgical factors between the occurrence of the initial risk type and the appearance of the accident, and compare the actual values ​​of the surgical factors with the corresponding standard reference values ​​to determine the risk scenario in which they occur; Determine the percentage of similar cases that transform from the initial risk state to other initial risk states under the target risk scenario, and use the percentage of similar cases as the element value of the transition matrix under the target risk scenario; The real-time risk probability is obtained by using the initial risk probability and the element values ​​of the transition matrix under the target risk scenario, and risk alarms are generated based on the real-time risk probability.

9. The risk management method for breast interventional ultrasound surgery according to claim 8, characterized in that, The process of obtaining the real-time risk probability using the initial risk probability and the element values ​​of the transition matrix under the target risk scenario, and then issuing a risk alarm based on the real-time risk probability, includes: The transition probability of the target initial risk state transforming into other initial risk states is obtained by using the initial risk probability and the element values ​​of the transition matrix under the target risk scenario; The initial risk probability is updated and the real-time risk probability is obtained by using all the corresponding transition probabilities in the transition matrix under the target risk scenario. The preset safety threshold is compared with the real-time risk probability to generate a risk alarm.

10. A risk management system for interventional ultrasound surgery of the breast, characterized in that, The system is used to implement the risk management method for breast interventional ultrasound surgery as described in any one of claims 1 to 9; the system includes: The risk analysis module is used to determine whether there is an initial risk of tissue elasticity using the gray-level distribution histogram and elastic signal distribution histogram corresponding to the ultrasound images and elastography images during surgery, respectively; to determine whether there is an initial risk of energy deposition using the average temperature change rate of the energy action zone of the ablation device in the ultrasound images during surgery and the gray-level gradient value of the anatomical boundary therein; and to determine whether there is an initial risk of physiological compensation using the correlation between the preoperative heart rate variability signal and the blood flow signal of the breast microvessels in the surgical area. The risk alarm module is used to determine similar cases in the historical surgical case database by combining the overlap between the location of the initial risk type and the high-risk area calibration results with the initial risk type; it uses similar cases to determine the initial risk probability of the current surgical procedure risk escalation and obtains the real-time risk probability, and then performs risk alarms based on the real-time risk probability.

Images