Ablation parameter determination method and device for high-frequency pulse electric field ablation

Abstract

The invention discloses an ablation parameter determination method and device for high-frequency pulsed electric field ablation, and the method comprises the steps: obtaining collected medical image data containing depth information, and determining a corresponding lesion box containing the three-dimensional information of a lesion according to the medical image data; analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of an electrode for performing ablation on the lesion; taking a plurality of equidistant points on the boundary of the electrode contour as starting points, respectively emitting a preset number of rays to the periphery, and obtaining the number of intersection points of the rays and the focus box to determine whether the electrode is attached to the focus; when the electrode is attached to the focus, corresponding ablation parameters are determined according to the contour and depth of the focus, and the ablation parameters comprise the pulse voltage, the pulse width, the pulse string number and the discharge frequency. The problem that in the prior art, ablation parameters are not accurately determined is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of pulse electric field ablation, and in particular to a method and device for determining ablation parameters of high-frequency pulse electric field ablation. Background Art

[0002] Pulsed electric field ablation technology generates a high-voltage pulsed electric field with a pulse width of milliseconds, microseconds or even nanoseconds, which releases extremely high energy in a short period of time, which can cause a large number of irreversible micropores in the cell membrane and even in the cell organelles such as the endoplasmic reticulum, mitochondria, and nucleus, thereby causing apoptosis of diseased cells and achieving the desired treatment purpose.

[0003] The selection of ablation parameters is closely related to the treatment effect, patient safety and recovery after treatment. Therefore, how to scientifically and accurately determine the ablation parameters of high-frequency pulsed electric field ablation is the key to the development of this technology.

[0004] In the prior art, when determining ablation parameters, the ablation parameters set based on experience are often used as the ablation parameters for each ablation, which cannot adapt well to different lesions. Especially when a second ablation is required, it is more dependent on the doctor's personal experience to determine the ablation parameters required for the second ablation. As a result, the accuracy of the ablation parameters is low, making the treatment effect unstable and difficult to achieve the expected effect. Summary of the invention

[0005] In view of this, an object of the present invention is to provide a method and device for determining ablation parameters of high-frequency pulse electric field ablation, aiming to solve the problem of inaccurate ablation parameter determination in the prior art.

[0006] The embodiment of the present invention is implemented as follows: A method for determining ablation parameters of high-frequency pulse electric field ablation, the method comprising: Acquire the collected medical image data containing depth information, and determine a corresponding lesion box containing three-dimensional information of the lesion according to the medical image data; Analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion; Taking multiple equally spaced points on the boundary of the electrode outline as starting points, a preset number of rays are emitted to the surroundings respectively, and the number of intersections between the rays and the lesion box is obtained to determine whether the electrode fits the lesion; When the electrode is in contact with the lesion, corresponding ablation parameters are determined according to the contour and depth of the lesion, wherein the ablation parameters include pulse voltage, pulse width, number of pulse trains, and number of discharges.

[0007] Furthermore, in the above-mentioned method for determining ablation parameters of high-frequency pulse electric field ablation, the step of acquiring the collected medical image data containing depth information and determining the corresponding lesion box containing three-dimensional information of the lesion according to the medical image data comprises: Inputting the medical image data into a pre-trained lesion localization model to obtain a corresponding lesion box containing three-dimensional information of the lesion; Wherein, the training process of the lesion localization model is: Collecting a preset number of historical medical image data, each historical medical image data corresponds to a lesion range framed by a preset frame and annotated with a lesion type to obtain a training data set; The training data set is divided into a training set, a validation set and a test set according to a preset ratio, and then the training data set is input into a preset neural network for training until the model parameters tend to be stable to obtain the lesion localization model.

[0008] Furthermore, in the above-mentioned method for determining ablation parameters of high-frequency pulse electric field ablation, the step of analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion comprises: Inputting the lesion box into a pre-trained lesion segmentation model to obtain the corresponding lesion contour and depth and the contour of the electrode used to ablate the lesion; Wherein, the training process of the lesion segmentation model is: Collecting a preset number of historical lesion boxes and the contours and depths of the corresponding lesions and the contours of electrodes used to ablate the lesions to obtain a training data set; The training data set is divided into a training set, a validation set and a test set according to a preset ratio, and then the training data set is input into a preset neural network for training until the model parameters tend to be stable to obtain the lesion segmentation model.

[0009] Furthermore, in the above-mentioned method for determining ablation parameters of high-frequency pulse electric field ablation, the step of obtaining the number of intersections between the ray and the lesion box to determine whether the electrode fits the lesion comprises: When the number of intersections between the ray and the lesion box is an even number, it is determined that the electrode is not attached to the lesion; When the number of intersections between the ray and the lesion box is an odd number, it is determined that the electrode is in contact with the lesion.

[0010] Furthermore, in the above-mentioned method for determining ablation parameters of high-frequency pulse electric field ablation, the step of determining the corresponding ablation parameters according to the contour and depth of the lesion comprises: Inputting the contour and depth of the lesion into a preset first ablation parameter determination model to obtain corresponding ablation parameters; The training process of the first ablation parameter determination model is as follows: Collecting historical ablation parameters and corresponding ablation depths and ablation lengths as a training data set, and inputting the training data set into a naive Bayesian network for training to obtain the first ablation parameter determination model; The naive Bayesian network is based on the Bayesian probability formula and adjusts the optimal relationship between the ablation depth, ablation length and ablation parameters through multiple trainings; Among them, the mathematical expression of the Bayesian formula is: ； in, P(A|B) Indicates in the event B Events under the conditions A The probability of occurrence, P(B|A) Indicates in the event A Events under the conditions B The probability of occurrence, P(A) It is an event A The prior probability of independent occurrence, P(B) It is an event B The edge probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}.

[0011] Furthermore, in the above-mentioned method for determining ablation parameters of high-frequency pulse electric field ablation, after the step of determining the corresponding ablation parameters according to the contour and depth of the lesion when the electrode is attached to the lesion, the method further comprises: After completing an ablation discharge using the ablation parameters, acquiring impedance data, contour, and depth of the lesion within a preset time period after the ablation discharge ends; The impedance data and the contour and depth of the lesion after the ablation discharge are input into a pre-trained second ablation parameter determination model to obtain corresponding secondary ablation parameters.

[0012] Furthermore, in the above-mentioned method for determining ablation parameters of high-frequency pulse electric field ablation, the training process of the second ablation parameter determination model is as follows: Collecting historical impedance data, secondary ablation parameters, and corresponding ablation depth and ablation length as a training data set, and inputting the training data set into a Markov network for training to obtain the second ablation parameter determination model; The Markov network is based on the Bayesian probability formula and adjusts the optimal relationship between the ablation depth, ablation length and secondary ablation parameters through multiple trainings; Among them, the mathematical expression of the Bayesian formula is: ； in, P(A|B, I) Indicates in the event B Events under the conditions A, I The probability of occurrence, P(B|A, I) Indicates in the event A Events under the conditions B, I The probability of occurrence, P(A|I) Indicates in the event A The probability of event I occurring under the condition that P(B|I) Indicates in the event B Events under the conditions I The probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}, I The impedance data.

[0013] Another object of the present invention is to provide a device for determining ablation parameters of high-frequency pulse electric field ablation, the device comprising: An acquisition module, used to acquire the collected medical image data containing depth information, and determine a corresponding lesion box containing three-dimensional information of the lesion according to the medical image data; A segmentation module, used for analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used for ablating the lesion; A collision module is used to take a plurality of equally spaced points on the boundary of the electrode outline as starting points, respectively emit a preset number of rays to the surroundings, obtain the number of intersections between the rays and the lesion box, so as to determine whether the electrode fits the lesion; A determination module, used for determining corresponding ablation parameters according to the contour and depth of the lesion when the electrode is in contact with the lesion; or used to determine corresponding ablation parameters according to the contour and depth of the lesion when the electrode is attached to the lesion, and used to obtain impedance data, contour and depth of the lesion after the corresponding ablation parameters complete one ablation discharge, and determine corresponding secondary ablation parameters based on the impedance data and the contour and depth of the lesion; The ablation parameters and the secondary ablation parameters both include pulse voltage, pulse width, number of pulse trains and number of discharges.

[0014] Another object of the present invention is to provide a readable storage medium having a computer program stored thereon, wherein the program implements the steps of the above method when executed by a processor.

[0015] Another object of the present invention is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the program.

[0016] Compared with the prior art, the present invention can determine the lesion box containing the three-dimensional information of the lesion by acquiring medical image data, so that the contour and depth of the lesion and the electrode contour can be determined according to the potential information about the length, width and depth of the lesion inside the lesion box, so as to use the contour and depth of the lesion to determine each ablation parameter, and can accurately provide corresponding ablation parameters when facing different lesions. The problem of inaccuracy in determining ablation parameters in the prior art is solved. And the collision algorithm is used to detect whether the electrode and the lesion are fitted, which can further improve the accuracy of lesion elimination and avoid thermal damage.

[0017] In addition, the present invention has at least the following beneficial effects: 1. It can automatically identify and locate the lesion and automatically calculate the size of the lesion; 2. It can automatically identify the position of the catheter with integrated electrodes and determine whether the catheter is in place; 3. Establish a feedback method that can recommend the next ablation parameters based on the change in impedance after discharge. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of a method for determining ablation parameters of high-frequency pulse electric field ablation provided by the first embodiment of the present invention; Figure 2 It is a structural block diagram of an ablation parameter determination device for high-frequency pulse electric field ablation in the third embodiment of the present invention.

[0019] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0020] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0021] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0023] Embodiment 1 See also Figure 1 , shown is a method for determining ablation parameters of high-frequency pulse electric field ablation in the first embodiment of the present invention, the method includes steps S10 to S13.

[0024] Step S10, acquiring collected medical image data containing depth information, and determining a corresponding lesion box containing three-dimensional information of the lesion according to the medical image data.

[0025] Among them, the medical image data can be CT image data, which is data collected by the CT device in the DICOM format. This format is the native data output by the CT device, which contains a grayscale image and a corresponding depth value. The medical image data contains information on the range and type of lesions, and the corresponding lesion box containing three-dimensional information of the lesion can be obtained by analyzing the medical image data.

[0026] Specifically, the medical image data is input into a pre-trained lesion localization model to obtain a corresponding lesion box containing three-dimensional information of the lesion, wherein the medical image data containing depth information and pictures obtained by CT scanning is input into the lesion localization model, and the model can derive a specific lesion box containing three-dimensional information of the lesion.

[0027] Exemplarily, a preset number of historical medical image data are collected, and each historical medical image data corresponds to a preset frame to frame the lesion range and mark the lesion type to obtain a training data set. For example, 8,000 CT imaging data are sorted, and each CT data corresponds to a frame to frame the lesion range and lesion type information. The frame is a three-dimensional cube and the corresponding lesion type. The CT data, the 8 vertex coordinates of the three-dimensional cube, and the lesion type are taken as a set of data, and 30,000 sets of data are used as training data sets and input into the preset neural network for training using an AI training server. Until the training stop conditions are met, such as the model parameters tend to be stable or the preset number of training times is reached.

[0028] Furthermore, the training set, validation set and test set can be divided into preset proportions of 80%, 15% and 5% for training, validation and testing respectively. The preset neural network can use YOLO, specifically, YOLOv8. The trained model has mastered the inherent logic of using CT data to locate lesions. Therefore, the input CT data can output data containing 8 coordinates of a three-dimensional cube and the type of lesion. Among them, the three-dimensional cube composed of 8 coordinates is the lesion box. The lesion box contains the length, width and depth information of the lesion.

[0029] In addition, in some preferred embodiments of the present invention, in order to further obtain more accurate lesion box data, after obtaining the above-mentioned three-dimensional cube, the cube can be subjected to intersection and union operations to filter out repeated data in the range and obtain the final lesion box as output data.

[0030] Step S11, analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion.

[0031] Among them, in order to determine the ablation parameters during ablation discharge, it is necessary to obtain the contour and depth data of the corresponding lesion, so that the corresponding ablation parameters can be accurately set according to the contour and depth data of the lesion, and the contour of the electrode used to ablate the lesion (hereinafter collectively referred to as "electrode contour") can be obtained. The position information of the electrode can also be determined according to the electrode contour, that is, the appropriate ablation position can be determined, so that the above information can be displayed on the display screen of the ablation device. The operator of the device can confirm the insertion of the catheter integrated with the electrode into the specified position according to the electrode contour and position information. In specific implementation, it can also be implemented in conjunction with the position sensor on the catheter. The reason is that the catheter cannot actually move in a straight line. The sensor can accurately determine whether it touches other tissues, thereby adjusting the position of the catheter on the insertion path in real time until the catheter moves to the correct position.

[0032] Specifically, a lesion segmentation model can be used to output the corresponding lesion contour and depth and electrode contour according to the input lesion box, wherein UNET is selected as the neural network for training the lesion segmentation model, which is mainly used to mark the lesion contour, and the training data set is obtained using the historical lesion box and the corresponding lesion contour and depth, electrode contour. For example, the lesion box data can be obtained, the lesion contour and electrode contour in the lesion box can be calibrated, and the lesion box and the corresponding data after the annotated lesion contour and electrode contour are composed of the training data set. The training set, validation set and test set can be divided into preset ratios of 80%, 15% and 5%, which are used for training, validation and testing respectively. The lesion segmentation model has mastered the internal logic of the segmentation of the lesion contour and the electrode contour, so that the lesion box is input into the lesion segmentation model for reasoning, which can analyze the potential data of the length, width and depth of the lesion contained in the lesion box, so as to obtain the required lesion contour and depth and electrode contour.

[0033] Step S12, starting from a plurality of equally spaced points on the boundary of the electrode outline, respectively emitting a preset number of rays to the surroundings, obtaining the number of intersections between the rays and the lesion box to determine whether the electrode fits the lesion.

[0034] Among them, after determining the two major factors affecting the ablation parameters, namely the lesion contour and the electrode contour, it is necessary to detect the degree of fit between the electrode and the lesion, so as to ensure that pulse ablation is performed when the electrode and the lesion are in fit. Specifically, in this embodiment, a collision algorithm is used to determine whether the electrode and the lesion are in fit. In the specific implementation, a plurality of equally spaced points on the boundary of the electrode contour are used as starting points, and a preset number of rays are emitted to the surrounding areas. For example, 50 points are selected and 50 rays are emitted. When the number of intersections between the rays and the lesion box (i.e., the number of collisions) is an even number, it is determined that the electrode and the lesion are not in fit; when the number of intersections between the rays and the lesion box is an odd number, it is determined that the electrode and the lesion are in fit. Step S13, when the electrode is in contact with the lesion, corresponding ablation parameters are determined according to the contour and depth of the lesion, wherein the ablation parameters include pulse voltage, pulse width, number of pulse trains, and number of discharges.

[0035] Among them, after the contour (such as length) and depth information of the lesion are known, the corresponding ablation parameters can be accurately determined according to the contour and depth information of the lesion. Specifically, the ablation parameters include pulse voltage, pulse width, number of pulse trains and number of discharges. These parameters can be pushed to the operator of the ablation device, such as a doctor, so that the doctor can operate the ablation device according to these parameters. In addition, in an embodiment of the present invention, the type of lesion can also be determined based on the analysis of medical image data. After the type of lesion is determined, electrode catheters of different types or functions can be selected to adapt to the current ablation operation.

[0036] Specifically, a model is trained by means of model training, which can determine the corresponding ablation parameters according to the contour and depth of the lesion, wherein animals can be used to collect training data sets. For example, when mice or pigs are used as research subjects, the rectum of mice or the gallbladder, pancreas, liver and digestive tract of pigs can be selected as test subjects. For example, discharge ablation tests are performed on mice, and the corresponding ablation parameters and ablation effects are recorded. The ablation effect can be evaluated by comparing the difference between the ablation length and ablation depth after the current ablation and the ablation length and ablation depth after the previous ablation. That is, when the difference between the two is within the corresponding preset range at the same time, it can be judged that the expected ablation effect is achieved. The ablation length and ablation depth in are the length and depth of the lesion after ablation, and the ablation length and ablation depth during the first ablation are the initial length and initial depth of the lesion. The ablation parameters and the corresponding ablation length and ablation depth form a training data set and are input into the naive Bayesian network for training to obtain the first ablation parameter determination model. It should be noted that in actual training, the ablation parameters, ablation depth and ablation length form a training data set, and the ablation length and ablation depth are determined by the ablation parameters. However, in actual prediction, it is necessary to determine the ablation parameters through the contour and depth information of the lesion, namely the ablation length and ablation depth, so the naive Bayesian network is used. The naive Bayesian network can adjust the optimal relationship between the ablation length and ablation depth of the discharge ablation and the ablation parameters through multiple training based on the Bayesian probability formula, so that the corresponding ablation parameters can be accurately output under the premise of the known ablation length and ablation depth of the lesion.

[0037] More specifically, the mathematical expression of the Bayesian formula is: ； in, P(A|B) Indicates in the event B Events under the conditions A The probability of occurrence, P(B|A) Indicates in the event A Events under the conditions B The probability of occurrence, P(A) It is an event A The prior probability of independent occurrence, P(B) It is an event B The edge probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}.

[0038] In summary, the method for determining ablation parameters of high-frequency pulse electric field ablation in the above-mentioned embodiment of the present invention can determine the lesion box containing three-dimensional information of the lesion by acquiring medical image data, so that the contour and depth of the lesion and the electrode contour can be determined according to the potential information about the length, width and depth of the lesion inside the lesion box, so as to use the contour and depth of the lesion to determine the determination of the ablation parameters each time, and can accurately provide corresponding ablation parameters when facing different lesions. The problem of inaccuracy in determining ablation parameters in the prior art is solved. And the use of a collision algorithm to detect whether the electrode and the lesion are in contact can further improve the accuracy of lesion elimination and avoid thermal damage.

[0039] Embodiment 2 This embodiment also proposes a method for determining ablation parameters of high-frequency pulse electric field ablation. The difference between the method for determining ablation parameters of high-frequency pulse electric field ablation in this embodiment and the method for determining ablation parameters of high-frequency pulse electric field ablation in the first embodiment is that: After step S13, the following steps are also included: After completing an ablation discharge using the ablation parameters, acquiring impedance data, contour, and depth of the lesion within a preset time period after the ablation discharge ends; Specifically, the current impedance data is obtained through an external impedance analyzer, and the impedance data and the contour and depth of the lesion after the end of this ablation discharge are input into the pre-trained second ablation parameter determination model to obtain the corresponding secondary ablation parameters. It should be noted that the secondary ablation refers to the need to ablate the lesion again after the last ablation discharge, including the second ablation, the third ablation and the Nth ablation. In addition, it should be noted that before performing the secondary ablation, the current temperature data can also be obtained to determine whether to continue the ablation based on the data. For example, when the temperature is too high, the ablation is stopped directly, and when the temperature is normal, the secondary ablation is continued to avoid the occurrence of thermal damage. In addition, since the impedance of the lesion will undergo relatively drastic changes at the end of discharge, multiple impedance values ​​may appear within the preset time. In order to ensure the accuracy of the secondary ablation parameter determination, multiple impedance data with similar values ​​can be screened out within the preset time period to obtain the average value as the impedance data of the lesion, or the one with a larger number of identical values ​​can be screened out as the impedance data of the lesion. For example, the preset time period can be selected within 5S or 10S after the end of discharge. The specific range can be selected according to the actual situation and is not limited here.

[0040] Among them, an ablation parameter adjustment feedback mechanism is established. After an ablation discharge is performed, the effect of this ablation discharge can be evaluated, that is, whether the corresponding ablation length and ablation depth are reached under the acquired impedance data. Therefore, if the discharge effect does not meet expectations, that is, the ablation length and ablation depth of the lesion after the ablation discharge do not reach the corresponding ablation length and ablation depth, that is, the difference between the ablation length and ablation depth of the lesion after the ablation discharge and the ablation length and ablation depth of the lesion after the previous ablation discharge are not simultaneously within the corresponding preset range, the ablation parameters of this time can be improved to obtain secondary ablation parameters, which can improve the therapeutic effect of secondary ablation on the lesion.

[0041] It should be noted that in this embodiment, the impedance data after discharge is used to assist the ablation depth and ablation length to obtain the secondary ablation parameters. Compared with the determination of ablation parameters based entirely on ablation length and ablation depth, the accuracy of the determination of the secondary ablation parameters is effectively improved. In the specific implementation, impedance is introduced as additional information, the Bayesian formula is re-established, and a Markov network is established based on the formula. The network is deployed to the training server for AI training. Each set of training data contains secondary ablation parameters, impedance data, ablation length and ablation depth after ablation discharge, and the first ablation parameter determination model Similarly, during training, the training data set consisting of secondary ablation parameters, impedance data, ablation depth and ablation length is used to determine the ablation length and ablation depth through the secondary ablation parameters and impedance data. However, in actual prediction, the secondary ablation parameters need to be determined through impedance data, the contour and depth information of the lesion, i.e., the ablation length and ablation depth. The Markov network needs to construct a Markov chain. Therefore, the ablation length and ablation depth, secondary ablation parameters and impedance data after the above-mentioned ablation discharge are used as states for probability deduction, thereby forming a Markov model. The Bayesian formula is also used for probability calculation.

[0042] Specifically, the mathematical expression of the Bayesian formula is: ； in, P(A|B, I) Indicates in the event B Events under the conditions A, I The probability of occurrence, P(B|A, I) Indicates in the event A Events under the conditions B, I The probability of occurrence, P(A|I) Indicates in the event A The probability of event I occurring under the condition that P(B|I) Indicates in the event B Events under the conditions IThe probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}, I The impedance data.

[0043] Exemplarily, after the discharge is completed, the impedance data 5 seconds after the discharge is completed, and the contour and depth of the lesion after this ablation discharge are input into the Markov network again. At this time, the Markov network can be adjusted according to the results of this discharge to obtain the ablation parameters required for the second discharge, namely the secondary ablation parameters.

[0044] In summary, the method for determining ablation parameters of high-frequency pulse electric field ablation in the above-mentioned embodiment of the present invention can determine the lesion box containing three-dimensional information of the lesion by acquiring medical image data, so that the contour and depth of the lesion and the electrode contour can be determined according to the potential information about the length, width and depth of the lesion inside the lesion box, so as to use the contour and depth of the lesion to determine the determination of the ablation parameters each time, and can accurately provide corresponding ablation parameters when facing different lesions. The problem of inaccuracy in determining ablation parameters in the prior art is solved. And the use of a collision algorithm to detect whether the electrode and the lesion are in contact can further improve the accuracy of lesion elimination and avoid thermal damage.

[0045] Embodiment 3 See also Figure 2 , shown is an ablation parameter determination device for high-frequency pulse electric field ablation proposed in the third embodiment of the present invention, the device comprising: An acquisition module 100 is used to acquire the collected medical image data containing depth information, and determine a corresponding lesion box containing three-dimensional information of the lesion according to the medical image data; The segmentation module 200 is used to analyze and segment the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion; The collision module 300 is used to take a plurality of equally spaced points on the boundary of the electrode outline as starting points, respectively emit a preset number of rays to the surroundings, obtain the number of intersections between the rays and the lesion box, so as to determine whether the electrode fits the lesion; A determination module 400, configured to determine corresponding ablation parameters according to the contour and depth of the lesion when the electrode is in contact with the lesion; or used to determine corresponding ablation parameters according to the contour and depth of the lesion when the electrode is attached to the lesion, and used to obtain impedance data, contour and depth of the lesion after the corresponding ablation parameters complete one ablation discharge, and determine corresponding secondary ablation parameters based on the impedance data and the contour and depth of the lesion; The ablation parameters and the secondary ablation parameters both include pulse voltage, pulse width, number of pulse trains and number of discharges.

[0046] Furthermore, in the above-mentioned device for determining ablation parameters of high-frequency pulse electric field ablation, the step of acquiring the collected medical image data containing depth information and determining the corresponding lesion box containing three-dimensional information of the lesion according to the medical image data comprises: Inputting the medical image data into a pre-trained lesion localization model to obtain a corresponding lesion box containing three-dimensional information of the lesion; Wherein, the training process of the lesion localization model is: Collecting a preset number of historical medical image data, each historical medical image data corresponds to a lesion range framed by a preset frame and annotated with a lesion type to obtain a training data set; The training data set is divided into a training set, a validation set and a test set according to a preset ratio, and then the training data set is input into a preset neural network for training until the model parameters tend to be stable to obtain the lesion localization model.

[0047] Furthermore, in the above-mentioned device for determining ablation parameters of high-frequency pulse electric field ablation, the step of analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion comprises: Inputting the lesion box into a pre-trained lesion segmentation model to obtain the corresponding lesion contour and depth and the contour of the electrode used to ablate the lesion; Wherein, the training process of the lesion segmentation model is: Collecting a preset number of historical lesion boxes and the contours and depths of the corresponding lesions and the contours of electrodes used to ablate the lesions to obtain a training data set; The training data set is divided into a training set, a validation set and a test set according to a preset ratio, and then the training data set is input into a preset neural network for training until the model parameters tend to be stable to obtain the lesion segmentation model.

[0048] Furthermore, in the above-mentioned device for determining ablation parameters of high-frequency pulse electric field ablation, the step of obtaining the number of intersections between the ray and the lesion box to determine whether the electrode fits the lesion comprises: When the number of intersections between the ray and the lesion box is an even number, it is determined that the electrode is not attached to the lesion; When the number of intersections between the ray and the lesion box is an odd number, it is determined that the electrode is in contact with the lesion.

[0049] Furthermore, in the above-mentioned device for determining ablation parameters of high-frequency pulse electric field ablation, the step of determining the corresponding ablation parameters according to the contour and depth of the lesion comprises: Inputting the contour and depth of the lesion into a preset first ablation parameter determination model to obtain corresponding ablation parameters; The training process of the first ablation parameter determination model is as follows: Collecting historical ablation parameters and corresponding ablation depths and ablation lengths as a training data set, and inputting the training data set into a naive Bayesian network for training to obtain the first ablation parameter determination model; The naive Bayesian network is based on the Bayesian probability formula and adjusts the optimal relationship between the ablation depth, ablation length and ablation parameters through multiple trainings; Among them, the mathematical expression of the Bayesian formula is: ； in, P(A|B) Indicates in the event B Events under the conditions A The probability of occurrence, P(B|A) Indicates in the event A Events under the conditions B The probability of occurrence, P(A) It is an event A The prior probability of independent occurrence, P(B) It is an event B The edge probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}.

[0050] Furthermore, the ablation parameter determination device for high-frequency pulse electric field ablation, wherein after the step of determining the corresponding ablation parameter according to the contour and depth of the lesion when the electrode is attached to the lesion, further comprises: After completing an ablation discharge using the ablation parameters, acquiring impedance data, contour, and depth of the lesion within a preset time period after the ablation discharge ends; The impedance data and the contour and depth of the lesion after the ablation discharge are input into a pre-trained second ablation parameter determination model to obtain corresponding secondary ablation parameters.

[0051] Furthermore, in the above-mentioned ablation parameter determination device for high-frequency pulse electric field ablation, the training process of the second ablation parameter determination model is: Collecting historical impedance data, secondary ablation parameters, and corresponding ablation depth and ablation length as a training data set, and inputting the training data set into a Markov network for training to obtain the second ablation parameter determination model; The Markov network is based on the Bayesian probability formula and adjusts the optimal relationship between the ablation depth, ablation length and secondary ablation parameters through multiple trainings; Among them, the mathematical expression of the Bayesian formula is: ； in, P(A|B, I) Indicates in the event B Events under the conditions A, I The probability of occurrence, P(B|A, I) Indicates in the event A Events under the conditions B, I The probability of occurrence, P(A|I) Indicates in the event A The probability of event I occurring under the condition that P(B|I) Indicates in the event B Events under the conditions I The probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}, I The impedance data.

[0052] The functions or operation steps implemented when the above modules are executed are substantially the same as those in the above method embodiments, and will not be repeated here.

[0053] Embodiment 4 Another aspect of the present invention further provides a readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the steps of the method described in any one of the above embodiments 1 to 2 are implemented.

[0054] Embodiment 5 Another aspect of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in any one of the above-mentioned embodiments one to two when executing the program.

[0055] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0056] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable storage medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, a "computer-readable storage medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0057] More specific examples (a non-exhaustive list) of computer-readable storage media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable storage medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0058] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0059] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0060] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for determining ablation parameters of high-frequency pulse electric field ablation, characterized in that: The method comprises: Acquire the collected medical image data containing depth information, and determine a corresponding lesion box containing three-dimensional information of the lesion according to the medical image data; Analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion; Taking multiple equally spaced points on the boundary of the electrode outline as starting points, respectively emitting a preset number of rays to the surroundings, and obtaining the number of intersections between the rays and the lesion box to determine whether the electrode fits the lesion; When the electrode is in contact with the lesion, corresponding ablation parameters are determined according to the contour and depth of the lesion, wherein the ablation parameters include pulse voltage, pulse width, number of pulse trains, and number of discharges.

2. The method for determining ablation parameters of high-frequency pulse electric field ablation according to claim 1, characterized in that: The step of acquiring the collected medical image data containing depth information and determining the corresponding lesion box containing three-dimensional information of the lesion according to the medical image data comprises: Inputting the medical image data into a pre-trained lesion localization model to obtain a corresponding lesion box containing three-dimensional information of the lesion; Wherein, the training process of the lesion localization model is: Collecting a preset number of historical medical image data, each historical medical image data corresponds to a lesion range framed by a preset frame and annotated with a lesion type to obtain a training data set; The training data set is divided into a training set, a validation set and a test set according to a preset ratio, and then the training data set is input into a preset neural network for training until the model parameters tend to be stable to obtain the lesion localization model.

3. The method for determining ablation parameters of high-frequency pulse electric field ablation according to claim 1, characterized in that: The step of analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used to ablate the lesion comprises: Inputting the lesion box into a pre-trained lesion segmentation model to obtain the corresponding lesion contour and depth and the contour of the electrode used to ablate the lesion; Wherein, the training process of the lesion segmentation model is: Collecting a preset number of historical lesion boxes and the contours and depths of the corresponding lesions and the contours of electrodes used to ablate the lesions to obtain a training data set; The training data set is divided into a training set, a validation set and a test set according to a preset ratio, and then the training data set is input into a preset neural network for training until the model parameters tend to be stable to obtain the lesion segmentation model.

4. The method for determining ablation parameters of high-frequency pulse electric field ablation according to claim 1, characterized in that: The step of obtaining the number of intersections between the ray and the lesion box to determine whether the electrode fits the lesion comprises: When the number of intersections between the ray and the lesion box is an even number, it is determined that the electrode is not attached to the lesion; When the number of intersections between the ray and the lesion box is an odd number, it is determined that the electrode is in contact with the lesion.

5. The method for determining ablation parameters of high-frequency pulse electric field ablation according to claim 1, characterized in that: The step of determining corresponding ablation parameters according to the contour and depth of the lesion comprises: Inputting the contour and depth of the lesion into a preset first ablation parameter determination model to obtain corresponding ablation parameters; The training process of the first ablation parameter determination model is as follows: Collecting historical ablation parameters and corresponding ablation depths and ablation lengths as a training data set, and inputting the training data set into a naive Bayesian network for training to obtain the first ablation parameter determination model; The naive Bayesian network is based on the Bayesian probability formula and adjusts the optimal relationship between the ablation depth, ablation length and ablation parameters through multiple trainings; Among them, the mathematical expression of the Bayesian formula is: ； in, P(A|B) Indicates in the event B Events under the conditions A The probability of occurrence, P(B|A) Indicates in the event A Events under the conditions B The probability of occurrence, P(A) It is an event A The prior probability of independent occurrence, P(B) It is an event B The edge probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}.

6. The method for determining ablation parameters of high-frequency pulse electric field ablation according to claim 1, characterized in that: After the step of determining the corresponding ablation parameters according to the contour and depth of the lesion when the electrode is attached to the lesion, the method further comprises: After completing an ablation discharge using the ablation parameters, acquiring impedance data, contour, and depth of the lesion within a preset time period after the ablation discharge ends; The impedance data and the contour and depth of the lesion after the ablation discharge are input into a pre-trained second ablation parameter determination model to obtain corresponding secondary ablation parameters.

7. The method for determining ablation parameters of high-frequency pulse electric field ablation according to claim 6, characterized in that: The training process of the second ablation parameter determination model is: Collecting historical impedance data, secondary ablation parameters, and corresponding ablation depth and ablation length as a training data set, and inputting the training data set into a Markov network for training to obtain the second ablation parameter determination model; The Markov network is based on the Bayesian probability formula and adjusts the optimal relationship between the ablation depth, ablation length and secondary ablation parameters through multiple trainings; Among them, the mathematical expression of the Bayesian formula is: ； in, P(A|B,I) Indicates in the event B Events under the conditions A, I The probability of occurrence, P(B|A,I) Indicates in the event A Events under the conditions B, I The probability of occurrence, P(A|I) Indicates in the event A The probability of event I occurring under the condition that P (B∣I) Indicates in the event B Events under the conditions I The probability of occurrence, event A = {pulse voltage, pulse width, number of pulse trains, number of discharges}, event B ={ablation length, ablation depth}, I The impedance data.

8. A device for determining ablation parameters of high-frequency pulse electric field ablation, characterized in that: The device comprises: An acquisition module, used to acquire the collected medical image data containing depth information, and determine a corresponding lesion box containing three-dimensional information of the lesion according to the medical image data; A segmentation module, used for analyzing and segmenting the lesion box to obtain the contour and depth of the lesion and the contour of the electrode used for ablating the lesion; A collision module is used to take a plurality of equally spaced points on the boundary of the electrode outline as starting points, respectively emit a preset number of rays to the surroundings, obtain the number of intersections between the rays and the lesion box, so as to determine whether the electrode fits the lesion; A determination module, configured to determine corresponding ablation parameters according to the contour and depth of the lesion when the electrode is in contact with the lesion; or used to determine corresponding ablation parameters according to the contour and depth of the lesion when the electrode is attached to the lesion, and used to obtain impedance data, contour and depth of the lesion after the corresponding ablation parameters complete one ablation discharge, and determine corresponding secondary ablation parameters based on the impedance data and the contour and depth of the lesion; The ablation parameters and the secondary ablation parameters both include pulse voltage, pulse width, number of pulse trains and number of discharges.

9. A readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when executing the program.

Images