Method, device, electronic device and computer-readable storage medium for localizing epileptic focus

Abstract

The embodiments of the present application provide an epileptic focus localization method, an epileptic focus localization device, an electronic device, and a computer-readable storage medium, and relate to the field of computer technology. The method comprises: obtaining intracranial EEG characteristic data of the subject; obtaining a monitoring video of the subject; inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptic focus localization model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model. It can be seen that the implementation of the embodiments of the present application can improve the accuracy and efficiency of the positioning of the target epileptic focus by inputting the intracranial EEG characteristic data and the monitoring video into the trained target epileptic focus localization model, so as to locate the target epileptic focus of the subject through the target epileptic focus localization model.

Description

Background Art

[0002] This section is intended to provide a background or context to embodiments of the invention that are recited in the claims. No description herein is admitted to be prior art by inclusion in this section.

[0003] Epilepsy is a common disease of the nervous system. Surgical operations can terminate or improve drug-resistant epilepsy. However, effective surgical operations rely on accurate localization of the epileptogenic focus. Head MRI and other imaging methods are currently used clinically to assist in localizing the epileptogenic focus. However, this localization method relies on the multidisciplinary assistance of experienced specialists, neurosurgeons, and radiologists, resulting in inconsistencies in the localization results and a decrease in the accuracy of localization of the epileptogenic focus.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] Based on the above problems, the inventors have conducted corresponding thinking and made targeted improvements, providing an epileptic focus localization method, an epileptic focus localization device, an electronic device and a computer-readable storage medium, which can input intracranial EEG feature data and monitoring video into a trained target epileptic focus localization model, so as to locate the target epileptic focus of the subject through the target epileptic focus localization model, thereby improving the accuracy of the located target epileptic focus.

[0006] According to a first aspect of an embodiment of the present application, a method for locating an epileptic focus is disclosed, comprising:

[0007] Acquiring intracranial EEG characteristic data of the tested subject;

[0008] Acquiring a monitoring video of the subject;

[0009] The intracranial EEG characteristic data and the monitoring video are input into a trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model.

[0010] In one embodiment, based on the above scheme, the step of obtaining intracranial EEG characteristic data of the subject includes:

[0011] Acquiring original intracranial electroencephalogram (EEG) monitoring data of the subject collected by a biomedical monitoring device;

[0012] Preprocessing the raw intracranial EEG monitoring data;

[0013] The intracranial EEG characteristic data is obtained based on the preprocessed raw intracranial EEG monitoring data.

[0014] In one embodiment, based on the aforementioned solution, the raw intracranial EEG monitoring data includes: multi-channel EEG monitoring data collected from multiple locations in the brain of the subject;

[0015] The intracranial EEG characteristic data is obtained according to the preprocessed original intracranial EEG monitoring data, including:

[0016] The potential difference between each channel and the reference electrode is calculated based on the preprocessed multi-channel EEG monitoring data to obtain the initial characteristic data of the EEG signal;

[0017] Intracranial EEG characteristic data is extracted according to the initial EEG signal characteristic data.

[0018] In one embodiment, based on the above scheme, the intracranial EEG characteristic data includes EEG signal waveform characteristics; the step of extracting the intracranial EEG characteristic data according to the EEG signal initial characteristic data includes:

[0019] Generate a corresponding EEG signal waveform diagram according to the initial EEG signal characteristic data;

[0020] Extracting EEG signal waveform features according to the EEG signal waveform diagram.

[0021] In one embodiment, based on the above scheme, the intracranial EEG feature data includes EEG signal time-frequency features; the step of extracting the intracranial EEG feature data based on the EEG signal initial feature data includes:

[0022] Performing Fourier transform on the initial characteristic data of the EEG signal to obtain time-frequency data of the EEG signal corresponding to the characteristic data of the EEG signal;

[0023] The time-frequency features of the EEG signal are extracted according to the time-frequency data of the EEG signal.

[0024] In one embodiment, based on the aforementioned scheme, the preprocessing of the original intracranial EEG monitoring data includes at least one of the following processing: resampling, filtering, noise data removal, and numerical normalization.

[0025] In one embodiment, based on the above solution, after acquiring the monitoring video of the object under test, the method further includes:

[0026] Obtaining human body feature information and motion feature information of the subject from the monitoring video;

[0027] The potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information and the motion characteristic information are input into the trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model.

[0028] In one embodiment, based on the above solution, obtaining the human body feature information and motion feature information of the subject from the monitoring video includes:

[0029] Extracting a plurality of frame images corresponding to the monitoring video respectively;

[0030] Detecting object key points corresponding to the detected object and key point spatial coordinates corresponding to the object key points from the multiple frame images respectively; the object key points include object facial key points and body key points;

[0031] Constructing human body feature information according to the spatial coordinates of all key points of the tested object for the key points of the target object;

[0032] The human body feature information is subjected to human body motion analysis to obtain the motion feature information of the subject.

[0033] In one embodiment, based on the above scheme, the step of inputting the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information and the motion characteristic information into the trained target epileptogenic focus localization model includes:

[0034] Taking a preset time point as a starting time, based on a preset time interval, the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information, and the motion characteristic information are divided respectively to obtain a plurality of divided segments;

[0035] The multiple segmented segments are input into the trained target epileptogenic focus localization model.

[0036] In one embodiment, based on the above scheme, the target epileptogenic focus localization model includes a feature processing layer, an attention layer, and an output layer;

[0037] The step of inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model, comprises:

[0038] Extracting video feature data from the monitoring video using the feature processing layer, and fusing the intracranial EEG feature data and the video feature data to obtain a fusion feature;

[0039] Using the attention layer to characterize the fused features to obtain embedded features;

[0040] The output layer is used to map the embedded features to an output space to obtain a positioning result of a target epileptogenic focus corresponding to the subject.

[0041] In one embodiment, based on the above solution, the method further includes:

[0042] Acquire training samples, wherein the training samples include multiple patient data corresponding to multiple epilepsy patients; the multiple patient data include intracranial electroencephalogram feature data, patient monitoring videos, and epilepsy attack labels; the epilepsy attack labels are used to characterize the epilepsy attack state and epileptogenic focus of the epilepsy patient;

[0043] Inputting the training samples into the initial epileptogenic focus localization model to be trained to obtain a model output result; the model output result includes an epileptic seizure state output result and an epileptogenic focus output result;

[0044] Determining a loss value of the initial epileptic focus localization model to be trained according to the model output result and the epileptic seizure label;

[0045] According to the loss value, the initial epileptic focus localization model to be trained is iteratively trained until a trained target epileptic focus localization model is obtained.

[0046] In one embodiment, based on the above scheme, the training samples include positive samples and negative samples, the positive samples include multiple patient data corresponding to multiple epilepsy patients with abnormal discharges, and the negative samples include multiple patient data corresponding to multiple epilepsy patients without abnormal discharges.

[0047] In one embodiment, based on the above scheme, the inputting of the training sample into the initial epileptogenic focus localization model to be trained to obtain the model output result includes:

[0048] Inputting the training samples into the initial epileptogenic focus localization model to be trained to obtain a prediction probability; the prediction probability includes an epileptic seizure prediction probability and an epileptogenic focus prediction probability;

[0049] Determine the model output result corresponding to the training sample according to the predicted probability.

[0050] In one embodiment, based on the above scheme, the iterative training of the initial epileptic focus localization model to be trained according to the loss value until a trained target epileptic focus localization model is obtained includes:

[0051] According to the loss value, updating the model parameters of the initial epileptic focus localization model to be trained;

[0052] When the loss value satisfies a preset convergence condition, a trained target epileptogenic focus localization pattern is obtained.

[0053] According to a second aspect of an embodiment of the present application, a device for locating an epileptic focus is disclosed, the device comprising:

[0054] An intracranial characteristic data acquisition module is used to acquire intracranial EEG characteristic data of the subject;

[0055] A monitoring video acquisition module is used to acquire the monitoring video of the object under test;

[0056] The target epileptogenic focus positioning module is used to input the intracranial EEG characteristic data and the monitoring video into a trained target epileptogenic focus positioning model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus positioning model.

[0057] According to a third aspect of an embodiment of the present application, an electronic device is disclosed, including: a processor; and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the method for locating an epileptic focus as disclosed in the first aspect is implemented.

[0058] According to a fourth aspect of an embodiment of the present application, a computer program medium is disclosed, on which computer-readable instructions are stored. When the computer-readable instructions are executed by a processor of a computer, the computer executes the method for locating an epileptic focus disclosed in the first aspect of the present application.

[0059] The embodiment of the present application can input the intracranial EEG characteristic data and the monitoring video into the trained target epileptogenic focus positioning model, so as to locate the target epileptogenic focus of the subject through the target epileptogenic focus positioning model. Compared with the prior art, the implementation of the embodiment of the present application, on the one hand, avoids the determination of the epileptogenic focus through the form of multi-doctor consultation, and combines the intracranial EEG characteristic data and the monitoring video to improve the accuracy and efficiency of locating the epileptogenic focus; on the other hand, the method of locating the epileptogenic focus can be applied not only during the patient's illness, but also before surgery, which improves the accuracy of preoperative positioning and increases the chance of successful surgery.

[0060] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0061] It should be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become readily understood. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-limiting manner, wherein:

[0063] Figure 1 Shown is a flow chart of a method for locating an epileptic focus according to an exemplary embodiment of the present application;

[0064] Figure 2 What is shown is an EEG signal waveform diagram generated according to initial EEG signal feature data according to an exemplary embodiment of the present application;

[0065] Figure 3 Shown is a schematic diagram of key points of a target object according to an exemplary embodiment of the present application;

[0066] Figure 4 What is shown is a flow chart of inputting intracranial EEG characteristic data and monitoring video into a trained target epileptogenic focus localization model to obtain a localization result according to an exemplary embodiment of the present application;

[0067] Figure 5 What is shown is a flow chart of training an initial epileptic focus localization model to be trained to obtain a trained target epileptic focus localization model according to an optional exemplary implementation of the present application;

[0068] Figure 6 Shown is a schematic diagram of a process for locating a target epileptogenic focus according to an optional implementation manner of the present application;

[0069] Figure 7 Shown is a schematic diagram of a scenario for locating a target epileptic focus of a subject before surgery or during hospitalization according to an exemplary embodiment of the present application;

[0070] Figure 8 Shown is a structural block diagram of an epileptic focus positioning device according to another optional exemplary embodiment of the present application;

[0071] Fig. 9 Shown is an electronic device according to another alternative exemplary embodiment of the present application.

[0072] In the drawings, the same or corresponding reference numerals represent the same or corresponding parts. DETAILED DESCRIPTION

[0073] The principles and spirit of the present application will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present application, and are not intended to limit the scope of the present application in any way. On the contrary, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0074] Those skilled in the art know that the embodiments of the present application can be implemented as a device, apparatus, method or computer program product. Therefore, the present application can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0075] According to the embodiments of the present application, a method for locating an epileptic focus, an apparatus for locating an epileptic focus, an electronic device, and a computer-readable storage medium are proposed.

[0076] Any number of elements in the drawings is for illustration and not limitation, and any naming is for distinction only and does not have any limiting meaning.

[0077] The principles and spirit of the present application are explained in detail below with reference to several representative implementations of the present application. SUMMARY OF THE INVENTION

[0079] At present, the existing technology for localizing epileptogenic focus usually relies on experienced specialists, neurosurgeons, and radiologists to assist in localization. However, in the process of localization, there will be inconsistencies in the conclusions given by doctors. Therefore, localizing the epileptogenic focus using the above method will reduce the accuracy and efficiency of the located epileptogenic focus.

[0080] Based on the above problems, the applicant has thought of avoiding the need for multiple doctors to assist in locating the epileptic focus during the process of locating the epileptic focus. By inputting the intracranial EEG feature data and the monitoring video data into the trained target epileptic focus localization model, the target epileptic focus corresponding to the subject can be located through the target epileptic focus localization model. It can be seen that the target epileptic focus corresponding to the subject can be located through the target epileptic focus localization model, which avoids the phenomenon of multiple doctors assisting in positioning in the prior art and improves the accuracy and efficiency of epileptic focus localization.

[0081] Application Scenario Overview

[0082] It should be noted that the following application scenarios are only shown to facilitate understanding of the spirit and principle of the present application, and the implementation of the present application is not limited in this respect. On the contrary, the implementation of the present application can be applied to any applicable scenario.

[0083] When used in the scenario of localizing the epileptic focus for preoperative diagnosis, it can avoid the phenomenon of multiple doctors assisting in the localization, thereby improving the accuracy and efficiency of localizing the epileptic focus.

[0084] Exemplary Methods

[0085] In combination with the above application scenarios, refer to Figure 1 and Fig. 9 The following describes a method for localizing an epileptic focus according to an exemplary embodiment of the present application.

[0086] See also Figure 1 , Figure 1FIG. 2 is a flow chart of a method for locating an epileptic focus according to an exemplary embodiment of the present application. Figure 1 As shown, the method for locating the epileptogenic focus may include:

[0087] Step S110: Acquire intracranial EEG characteristic data of the subject.

[0088] Step S120: Acquire monitoring video of the object under test.

[0089] Step S130: inputting the intracranial EEG characteristic data and the monitoring video into the trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model.

[0090] Implementation Figure 1 The epileptogenic focus localization method shown can input intracranial EEG characteristic data and monitoring video into the trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus of the subject through the target epileptogenic focus localization model. Compared with the prior art, the implementation of the embodiment of the present application, on the one hand, avoids determining the epileptogenic focus through consultation of multiple doctors, and combines intracranial EEG characteristic data and monitoring video to improve the accuracy and efficiency of localizing the epileptogenic focus; on the other hand, the method of localizing the epileptogenic focus can be applied not only during the patient's illness, but also before surgery, which improves the accuracy of preoperative localization and increases the chance of successful surgery.

[0091] These steps are described in detail below.

[0092] In step S110, intracranial EEG characteristic data of the subject is obtained.

[0093] Specifically, the test subject refers to a patient who needs to locate the target epileptogenic focus at this time, for example, a patient who is about to undergo epilepsy surgery. The intracranial EEG characteristic data can reflect the EEG fluctuation of the test subject at this time. It is worth noting that the intracranial EEG characteristic data is obtained from the intracranial position of the test subject.

[0094] As an optional embodiment, obtaining intracranial EEG characteristic data of the subject includes: obtaining original intracranial EEG monitoring data of the subject collected by a biomedical monitoring device; preprocessing the original intracranial EEG monitoring data; and obtaining intracranial EEG characteristic data based on the preprocessed original intracranial EEG monitoring data.

[0095] Specifically, biomedical testing equipment refers to equipment used to collect raw intracranial EEG monitoring data. The equipment can specifically be multiple monitoring devices set in the skull of the subject. Raw intracranial EEG monitoring data refers to data directly collected by biomedical monitoring equipment without any preprocessing. Specifically, raw intracranial EEG monitoring data specifically includes EEG data, ECG data, and EMG data.

[0096] It is worth noting that the original intracranial EEG monitoring data directly collected using biomedical monitoring equipment may contain some data that may affect the subsequent prediction of the target epileptogenic focus. In order to eliminate these influences, the original intracranial EEG monitoring data needs to be preprocessed to obtain intracranial EEG characteristic data that can reflect the actual EEG fluctuations of the subject.

[0097] It can be seen that the implementation of this optional embodiment can avoid directly using the original intracranial EEG monitoring data to locate the epileptogenic focus, thereby improving the accuracy of locating the epileptogenic focus.

[0098] As an optional embodiment, the original intracranial EEG monitoring data includes: multi-channel EEG monitoring data collected from multiple locations in the skull of the subject; obtaining intracranial EEG characteristic data based on the preprocessed original intracranial EEG monitoring data, including: calculating the potential difference between each channel and the reference electrode based on the preprocessed multi-channel EEG monitoring data to obtain initial characteristic data of the EEG signal; extracting intracranial EEG characteristic data based on the initial characteristic data of the EEG signal.

[0099] Specifically, the original intracranial EEG monitoring data is a multi-channel EEG monitoring data, which refers to the EEG monitoring data collected from multiple locations in the skull of the subject. For example, detection electrodes are placed at multiple locations in the skull of the subject, and each detection electrode at the skull can collect intracranial EEG monitoring data of one channel. Furthermore, multi-channel EEG monitoring data can be collected by placing detection electrodes at multiple locations in the skull.

[0100] The reference electrode is used to measure whether the collected multi-channel intracranial EEG monitoring data is in an abnormal discharge state at this time. Specifically, the potential difference between the intracranial EEG monitoring data under each channel and the reference electrode is calculated. According to the potential difference, the initial characteristic data of the EEG signal reflecting whether the channel EEG monitoring data is in an abnormal discharge state at this time can be obtained.

[0101] Since the multi-channel EEG monitoring data specifically includes EEG data, ECG data and EMG data, after obtaining the initial feature data of the EEG signal, it is necessary to extract the intracranial EEG feature data that reflects the brain discharge status of the subject at this time from the initial feature data of the EEG signal.

[0102] It can be seen that by implementing this optional embodiment, the potential difference between each channel and the reference electrode can be calculated to finally obtain intracranial EEG characteristic data that can reflect whether the brain of the subject is in abnormal discharge at this time.

[0103] As an optional embodiment, the intracranial EEG feature data includes EEG signal waveform features; extracting the intracranial EEG feature data based on the initial EEG signal feature data includes: generating a corresponding EEG signal waveform graph based on the initial EEG signal feature data; and extracting the EEG signal waveform features based on the EEG signal waveform graph.

[0104] Specifically, the intracranial EEG characteristic data is specifically displayed as an EEG signal waveform. By performing Fourier transformation on the initial characteristics of the EEG signal, an EEG signal waveform diagram can be obtained. Figure 2 FIG. 1 shows an EEG signal waveform diagram generated according to the initial feature data of the EEG signal according to an embodiment of the present application, such as Figure 2 As shown, each row of waveform corresponds to a channel. The fluctuation of the EEG signal waveform can reflect the initial characteristics of the EEG signal of the subject in real time.

[0105] It can be seen that by implementing this optional embodiment, an EEG signal waveform diagram corresponding to the initial characteristic data of the EEG signal can be obtained, and the EEG fluctuation of the subject can be visually reflected in real time.

[0106] As an optional embodiment, the intracranial EEG feature data includes EEG signal time-frequency features; extracting the intracranial EEG feature data based on the EEG signal initial feature data includes: performing Fourier transform on the EEG signal initial feature data to obtain EEG signal time-frequency data corresponding to the EEG signal feature data; extracting the EEG signal time-frequency features based on the EEG signal time-frequency data.

[0107] Specifically, the intracranial EEG feature data includes the EEG signal time-frequency feature, which is used to reflect the correspondence between the EEG of the subject and each specific moment. Specifically, the time-frequency spectrum of the EEG signal feature data (i.e., the EEG signal time-frequency data) can be obtained by performing Fourier transformation on the initial EEG signal feature data.

[0108] Since the EEG signal time-frequency data includes EEG time-frequency data, ECG time-frequency data and EMG time-frequency data, it is necessary to extract EEG signal time-frequency features that can be used to reflect the EEG fluctuations of the subject from the EEG signal time-frequency data.

[0109] It can be seen that by implementing this optional embodiment, the time-frequency characteristics of the EEG signal reflecting the real-time EEG changes of the subject can be obtained by performing Fourier transform on the initial characteristic data of the EEG signal.

[0110] As an optional embodiment, the original intracranial EEG monitoring data is preprocessed, including at least one of the following processes: resampling, filtering, noise data removal, and numerical standardization.

[0111] Specifically, after the original intracranial EEG monitoring data is collected, the original intracranial EEG monitoring data can be first resampled according to a unified standard to obtain data A; then the data A is filtered to obtain data B to reduce the impact of alternating current signals and non-EEG signal frequency bands on data A; then, the noise data in data B is eliminated to obtain data C; finally, the EEG data, ECG data, and EMG data included in data C are multiplied by corresponding constants, respectively, so that data C has a unified numerical range, so as to increase the numerical stability of the subsequent target epileptogenic focus localization model.

[0112] For example, the original intracranial EEG monitoring data is resampled at 500 Hz to obtain data A; the data in each channel in data A is notch filtered at 50 Hz to reduce the impact of the alternating current signal on data A, and the data in some channels are band-pass filtered from 0.1 Hz to 70 Hz to obtain data B to reduce the impact of non-EEG signal frequency bands on data A; the noise caused by poor contact in data B is eliminated to obtain data C; finally, the EEG data, ECG data and EMG data included in data C are multiplied by corresponding constants, so that data C has a uniform numerical range.

[0113] It can be seen that by implementing this optional embodiment, by preprocessing the original intracranial EEG monitoring data, the influence of some unnecessary data on the intracranial EEG characteristic data can be reduced, thereby increasing the accuracy of the intracranial EEG characteristic data.

[0114] In step S120, a monitoring video of the object under test is obtained.

[0115] Specifically, the monitoring video refers to the video collected during the process of monitoring the object under test. For example, a monitoring device can be carried on the head of the object under test, and there is a camera in the device to collect dynamic video of the object under test in real time.

[0116] As an optional embodiment, after obtaining the monitoring video of the subject, the method also includes: obtaining the human body feature information and motion feature information of the subject from the monitoring video; inputting the potential difference, EEG signal time-frequency features, human body feature information and motion feature information into a trained target epileptic focus localization model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model.

[0117] Specifically, human body features refer to features that describe a certain part of the human body of the subject. Action features refer to features that describe the actions performed by the subject. Human body feature information can be feature information of the eyes of the subject, feature information of the nose of the subject, feature information of the hands of the subject, feature information of the neck of the subject, or feature information of the feet of the subject, and this exemplary embodiment does not specifically limit this.

[0118] The motion characteristic information may be the motion characteristic information corresponding to the action of raising the arm when the subject is testing, or the motion characteristic information corresponding to the action of twitching the leg when the subject is testing, and this exemplary embodiment does not specifically limit this.

[0119] On this basis, the previously obtained potential difference, EEG signal time-frequency characteristics, human feature information obtained through monitoring video, and motion feature information are input into the trained target epileptogenic focus model, and the target epileptogenic focus of the subject at this time can be obtained. For example, the target epileptogenic focus is the N-1 area of ​​the subject's brain.

[0120] It can be seen that by implementing this optional embodiment, the data input into the trained target epileptic focus localization model includes, in addition to the potential difference and the time-frequency characteristics of the EEG signal, also human body feature information and motion feature information, which increases the accuracy of the target epileptic focus finally determined.

[0121] As an optional embodiment, obtaining the human body feature information and the motion feature information of the subject from the monitoring video includes: extracting a plurality of frame images corresponding to the monitoring video respectively;

[0122] The object key points corresponding to the tested object and the key point spatial coordinates corresponding to the object key points are detected from multiple frame images respectively; the object key points include the object's facial key points and body key points; human body feature information is constructed based on the spatial coordinates of all key points of the tested object for the target object key points; human body feature information is subjected to human motion analysis to obtain the motion feature information of the tested object.

[0123] Specifically, the monitoring video is actually composed of multiple frames of images. According to the human body signs, it is possible to detect which part of the multiple frames of images is the key point of the object under test, and also detect the position of the key point in the multiple frames of images (i.e., the spatial coordinates of the key point).

[0124] It is worth noting that the object key points include the object facial key points of the face of the subject and the body key points of the body of the subject. Figure 3 FIG. 1 is a schematic diagram of key points of a target object according to an exemplary embodiment of the present application. Figure 3As shown, each black dot represents a target object key point. For example, target object key point No. 6 represents the target object key point corresponding to the outside of the right eye, and target object key point No. 23 represents the target object key point corresponding to the left hip.

[0125] The human feature information used to characterize the human features of the subject can be obtained through the key point spatial coordinates corresponding to the key points of the same target object in multiple frame images. For example, the human feature information of the left hip can be obtained through the key point spatial coordinates of the key points of the target object corresponding to the left hip, and then the human feature information can be analyzed by human motion to determine that the left hip of the subject is shaking at this time (the shaking of the left hip is a motion feature information).

[0126] It can be seen that by implementing this optional embodiment, human feature information reflecting key points of the human object and motion feature information reflecting human motion can be obtained by monitoring the video, and then the human feature information and motion feature information can be input into the trained target epileptic focus localization model to improve the accuracy of epileptic focus localization.

[0127] As an optional embodiment, the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information are input into a trained target epileptic focus localization model, including: taking a preset time point as the starting time, and based on a preset time interval, dividing the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information respectively to obtain multiple divided segments; and inputting the multiple divided segments into the trained target epileptic focus localization model.

[0128] Specifically, after obtaining the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information, the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information are not directly input into the trained target epileptic focus localization model, but the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information are divided and processed.

[0129] Specifically, a preset time point is used as the starting time (for example, the starting time of collecting the intracranial EEG characteristic data and monitoring video of the subject), and the potential difference, EEG signal time-frequency characteristics, human body characteristic information, and motion characteristic information are divided and processed according to a preset time interval (for example, 2 seconds), and corresponding divided segments can be obtained. After obtaining multiple divided segments, the multiple divided segments are input into the trained target epileptogenic focus localization model.

[0130] It can be seen that by implementing this optional embodiment, multiple divided segments are input into the target epileptic focus positioning model, and the obtained target epileptic focus corresponds to the potential difference of a specific segment, the time-frequency characteristics of the EEG signal of a segment, the human body characteristic information of a segment, and the action characteristic information of a segment, so as to better understand the characteristic changes caused by the epileptic focus in the subject.

[0131] In step S130, the intracranial EEG feature data and the monitoring video are input into the trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model.

[0132] Specifically, the trained target epileptogenic focus localization model refers to a model used to determine whether the subject is in an epileptic seizure period at this time, and if the subject is indeed in an epileptic seizure period at this time, the epileptogenic focus that causes the subject to have an epileptic seizure can also be directly located.

[0133] Based on this, after obtaining the intracranial EEG feature data and monitoring video, the intracranial EEG feature data and monitoring video are input into the trained target epileptogenic focus model, so that the target epileptogenic focus of the subject at this time can be located.

[0134] As an optional embodiment, the target epileptic focus localization model includes a feature processing layer, an attention layer, and an output layer; the intracranial EEG feature data and the monitoring video are input into the trained target epileptic focus localization model to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model, including: using the feature processing layer to extract video feature data from the monitoring video, and fusing the intracranial EEG feature data and the video feature data to obtain fused features; using the attention layer to characterize the fused features to obtain embedded features; using the output layer to map the embedded features to the output space to obtain the positioning result of the target epileptic focus corresponding to the subject.

[0135] Specifically, the target epileptic focus localization model can be a neural network, and the neural network includes a feature processing layer, an attention layer and an output layer.

[0136] First, the feature processing layer can extract video feature data (i.e., human body feature heart and chest and motion feature information) from the monitoring video input into the trained target epileptogenic focus localization model. Then, the video feature data is aligned with the collected intracranial EEG feature data in feature dimension to perform feature fusion on the video feature data and intracranial EEG feature data, thereby obtaining fusion features.

[0137] Then, the trained target epileptogenic focus localization model characterizes the embedded features according to the fusion features. Specifically, the trained target epileptogenic focus localization model can map the embedded features to an output space (i.e., obtain the probability value corresponding to the intracranial EEG feature data and video feature data input into the trained target epileptogenic focus localization model). The probability value is used to characterize the probability that the subject is in the epileptic seizure period at this time and the probability that the epileptogenic focus that causes the subject to be in the epileptic seizure period is the target epileptogenic focus, and then through the probability value, it can be obtained whether the subject is in the epileptic seizure period at this time and the corresponding target epileptogenic focus (i.e., the positioning result).

[0138] Figure 4 FIG. 1 is a flow chart of inputting intracranial EEG feature data and monitoring video into a trained target epileptogenic focus localization model to obtain a localization result according to an exemplary embodiment of the present application, as shown in FIG. Figure 4 As shown, first obtain the original intracranial EEG monitoring data and monitoring video of the subject. Then, for the original intracranial EEG monitoring data, the original intracranial EEG monitoring data is preprocessed to obtain the initial feature data of the EEG signal, and then the initial feature data of the EEG signal is calculated with the reference electrode to obtain the potential difference. In addition, the initial feature data of the EEG signal is Fourier transformed to obtain the EEG signal time-frequency data, so as to extract the EEG signal time-frequency characteristics from the EEG signal time-frequency data. It is worth noting that the intracranial EEG feature data includes the potential difference and the EEG signal time-frequency characteristics.

[0139] As for the monitoring video, since the monitoring video includes the facial video of the subject and the body video of the subject, they can be processed separately. Specifically, multiple frame images of the facial video of the subject are determined, and the multiple frame images are input into the feature processing layer to extract the facial feature information of the human body. Similarly, the feature processing layer can also extract the body feature information of the human body. In addition, the key point spatial coordinates can also be obtained from the facial video of the subject. The key point spatial coordinates are analyzed for human motion to obtain facial motion feature information. Similarly, body motion feature information can also be obtained. It is worth noting that the human feature information includes the above-mentioned facial feature information and the body feature information, and the motion feature information includes the above-mentioned facial motion feature information and the body motion feature information.

[0140] After obtaining the potential difference, the EEG signal time-frequency features, the human facial feature information, the human body feature information, the facial action feature information and the body action feature information, the potential difference, the EEG signal time-frequency features, the human facial feature information, the human body feature information, the facial action feature information and the body action feature information are subjected to feature fusion to obtain fusion features. The fusion features are input into the attention layer to obtain embedded features, and finally the embedded features are mapped to the output space, so that the localization result of the target epileptogenic focus of the subject can be obtained.

[0141] It can be seen that by implementing this optional embodiment, the intracranial EEG feature data and monitoring video are input into the trained target epileptic focus localization model, which avoids the situation of determining the epileptic focus through consultation with multiple doctors, and improves the accuracy and efficiency of epileptic focus localization.

[0142] As an optional embodiment, the method also includes: obtaining training samples, the training samples include multiple patient data corresponding to multiple epilepsy patients; the multiple patient data include intracranial EEG feature data, patient monitoring videos and epilepsy seizure labels; the epilepsy seizure labels are used to characterize the epileptic seizure state and epileptic focus of epilepsy patients; inputting the training samples into the initial epileptic focus localization type to be trained to obtain the model output result; the model output result includes the epileptic seizure state output result and the epileptic focus output result; according to the model output result and the epileptic seizure label, determining the loss value of the initial epileptic focus localization model to be trained; according to the loss value, iteratively training the initial epileptic focus localization type to be trained until a trained target epileptic focus localization model is obtained.

[0143] Specifically, before the intracranial EEG feature data and the monitoring video are input into the trained target epileptogenic focus localization model, the target epileptogenic focus localization model needs to be trained.

[0144] The training sample refers to the patient data used to train the target epileptic focus localization model. Specifically, the patient data includes the segmented segments of the patient's intracranial EEG feature data, the segmented segments of the patient's monitoring video, and the epileptic seizure labels corresponding to the segmented segments of the intracranial EEG feature data and the segmented segments of the patient's monitoring video. The epileptic seizure label is used to indicate whether the patient corresponding to the patient data is in an epileptic seizure state. If the patient corresponding to the patient data is in an epileptic seizure state, the epileptic seizure label is also used to indicate the epileptic focus that causes the patient to be in an epileptic seizure state.

[0145] The training samples are input into the initial epileptogenic focus localization model that has not been trained, and the model output results can be obtained. The model output results include the epilepsy state output results (used to indicate whether the patient is in the epilepsy attack period) and the epileptogenic focus output results (used to indicate the epileptogenic focus that causes the patient to be in the epilepsy attack period).

[0146] The model output result is the prediction result obtained by the initial epileptogenic focus localization model based on the training sample. Since the prediction result is not necessarily accurate, it is necessary to determine the loss value of the initial epileptogenic focus localization model based on the model output result and the epileptic seizure label. Then, the initial epileptogenic focus localization model is iteratively trained according to the loss value (i.e., the difference between the predicted result and the true result) so that the loss value meets the preset convergence condition, thereby obtaining a trained target epileptogenic focus localization model.

[0147] Figure 5 FIG. 1 is a flow chart of training an initial epileptic focus localization model to be trained to obtain a trained target epileptic focus localization model according to an optional exemplary embodiment of the present application, such as Figure 5 As shown, data 510 is the divided segments of intracranial EEG feature data and the divided segments of patient monitoring videos, model 520 is the initial epileptic focus localization model to be trained, result 530 is the model output result, data 540 is the epileptic seizure label corresponding to the divided segments of intracranial EEG feature data and the divided segments of patient monitoring videos, and data 550 is the loss value determined according to the model output result and the epileptic seizure label.

[0148] It can be seen that, by implementing this optional embodiment, the trained target epileptic focus localization model is obtained by training the initial epileptic focus localization model with multiple training samples, which increases the prediction accuracy of the trained target epileptic focus localization model.

[0149] As an optional embodiment, the training samples include positive samples and negative samples, the positive samples include multiple patient data corresponding to multiple epilepsy patients with abnormal discharges, and the negative samples include multiple patient data corresponding to multiple epilepsy patients without abnormal discharges.

[0150] Specifically, the training samples include positive samples and negative samples. The so-called positive samples refer to patient data corresponding to multiple epilepsy patients with abnormal discharges, that is, positive samples are patient data corresponding to patients in the epileptic seizure period; negative samples are patient data corresponding to multiple epilepsy patients who are not in the epileptic seizure period.

[0151] It can be seen that by implementing this optional embodiment, the training samples include positive samples and negative samples. By inputting the positive samples and negative samples into the initial epileptic focus localization model, the initial epileptic focus localization model can more accurately predict whether the subject is in an epileptic seizure period at this time, and more accurately predict the target epileptic focus that causes the subject to be in an epileptic seizure period.

[0152] As an optional embodiment, the training samples are input into the initial epileptic focus localization type to be trained to obtain the model output result, including: inputting the training samples into the initial epileptic focus localization type to be trained to obtain the predicted probability; the predicted probability includes the predicted probability of epileptic seizure and the predicted probability of epileptic focus; and determining the model output result corresponding to the training sample according to the predicted probability.

[0153] Specifically, the prediction probability refers to the prediction probability of epileptic seizure and the prediction probability of epileptogenic focus. For example, the obtained prediction probability is 98% and 99% corresponding to area D, which proves that the model output result corresponding to the training sample shows that the probability that the patient corresponding to the training sample is in the epileptic seizure period is 98%, and the probability that the target epileptogenic focus that causes the patient to be in the epileptic seizure period is area D is 99%.

[0154] It can be seen that by implementing this optional embodiment, the prediction result of the initial epileptogenic focus localization type to be trained for the training sample can be more intuitively known through the prediction probability.

[0155] As an optional embodiment, the initial epileptic focus localization model to be trained is iteratively trained according to the loss value until a trained target epileptic focus localization model is obtained, including: updating the model parameters of the initial epileptic focus localization model to be trained according to the loss value; until the loss value meets a preset convergence condition, the trained target epileptic focus localization model is obtained.

[0156] Specifically, the loss value refers to the difference between the model output result and the epileptic seizure label, that is, the loss value is the difference between the prediction result of the initial epileptic focus localization model and the actual epileptic seizure situation of the patient (corresponding to the training sample). If the loss value does not meet the preset convergence condition (for example, the condition that the loss value is less than the preset loss threshold), it is necessary to continuously adjust the model parameters in the initial epileptic focus localization model until the loss value meets the preset convergence condition, then stop adjusting the model parameters in the initial epileptic focus localization model to obtain a trained target epileptic focus localization model.

[0157] It can be seen that, when this optional embodiment is implemented, the adjustment of the model parameters of the initial epileptic focus localization model is stopped only when the loss value satisfies the preset convergence condition, so as to improve the prediction accuracy of the trained target epileptic focus localization model.

[0158] By implementing the embodiments of the present application, on the one hand, it is possible to avoid determining the epileptic focus through consultation among multiple doctors, and the accuracy and efficiency of locating the epileptic focus can be improved by combining intracranial EEG characteristic data and monitoring video; on the other hand, the method of locating the epileptic focus can be used not only during the patient's illness, but also before surgery, thereby improving the accuracy of preoperative positioning and increasing the chance of successful surgery.

[0159] Figure 6 FIG. 1 is a flow chart of locating a target epileptogenic focus according to an optional implementation manner of the present application. Figure 6 As shown, data 610 is the intracranial EEG characteristic data of the subject (specifically, the intracranial EEG characteristic data under multiple channels). Data 620 is the monitoring video of the subject. Data 640 (i.e., the human body characteristic information and motion characteristic information of the subject) can be obtained through the key point detection model of model 630. Data 640 and data 610 are input into the trained target epileptogenic focus positioning model 650 to obtain the prediction result 660 (i.e., predicting whether the subject is in the epileptic seizure period, and if in the epileptic seizure period, predicting the target epileptogenic focus that causes the epileptic seizure).

[0160] Figure 7 FIG. 1 is a schematic diagram of a scenario for locating a target epileptogenic focus of a subject before surgery or during hospitalization according to an exemplary embodiment of the present application. Figure 7 As shown, the object 710 is the object under test. The acquisition device 720 is used to collect the original intracranial EEG monitoring data of the object under test 710. During the acquisition process, the electrodes 721, 722 and 723 can be inserted into various locations in the skull of the object under test 710 by surgery, and connected to the electrodes 721, 722 and 723 through the interface of the acquisition device 720 to collect multi-channel original intracranial EEG monitoring data. It is worth noting that although the object under test needs to undergo surgery to insert the electrodes 721, 722 and 723 into various locations in the skull of the object under test in order to collect the original intracranial EEG monitoring data, this acquisition method can often obtain original intracranial EEG monitoring data with higher accuracy, so as to improve the accuracy of the target epileptogenic focus located subsequently. The electrodes 721, 722 and 723 can be installed in a helmet, a hat or any head-mounted device, and this exemplary embodiment does not specifically limit this.

[0161] At the same time, the camera 730 collects monitoring videos of the face and body of the subject. It is worth noting that the acquisition device 720 can be coupled with the abnormal discharge detection device 740. Figure 7 The coupling method is only an exemplary one. The acquisition device 720 and the abnormal discharge detection device 740 can be integrated into the same device, which can be a camera 730 or a head-mounted device.

[0162] After receiving the intracranial EEG characteristic data and the monitoring video, the abnormal discharge detection device 740 can directly use the trained target epileptogenic focus localization model to locate the target epileptogenic focus corresponding to the subject, or can import the intracranial EEG characteristic data and the monitoring video into the server 750. After receiving the intracranial EEG characteristic data and the monitoring video, the server 750 can use the trained target epileptogenic focus localization model to locate the target epileptogenic focus corresponding to the subject. It is worth noting that the server 750 can be any form of data processing server such as a cloud server or a distributed server, and this exemplary embodiment does not specifically limit this.

[0163] After the target epileptogenic focus is obtained, the abnormal discharge detection result corresponding to the target epileptogenic focus may be displayed on the display 760, so that the doctor or the subject can understand the abnormal discharge situation at the specific target epileptogenic focus.

[0164] Exemplary Media

[0165] After introducing the method according to the exemplary embodiment of the present application, next, the medium according to the exemplary embodiment of the present application will be described.

[0166] In some possible embodiments, various aspects of the present application may also be implemented as a medium on which program code is stored, and when the program code is executed by a processor of a device, it is used to implement the steps of the method for localizing the epileptic focus according to various exemplary embodiments of the present application described in the above “Exemplary Method” section of this specification.

[0167] Specifically, when the processor of the device executes the program code, it is used to implement the following steps: obtaining intracranial EEG characteristic data of the subject; obtaining a monitoring video of the subject; inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptic focus localization model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model.

[0168] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: obtaining original intracranial EEG monitoring data of the subject collected by the biomedical monitoring device; preprocessing the original intracranial EEG monitoring data; and obtaining intracranial EEG characteristic data based on the preprocessed original intracranial EEG monitoring data.

[0169] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: multi-channel EEG monitoring data collected from multiple locations in the skull of the subject; obtaining intracranial EEG characteristic data based on the preprocessed original intracranial EEG monitoring data, including: calculating the potential difference between each channel and the reference electrode based on the preprocessed multi-channel EEG monitoring data to obtain initial EEG signal characteristic data; extracting intracranial EEG characteristic data based on the initial EEG signal characteristic data.

[0170] In some embodiments of the present application, the processor of the device is further used to implement the following steps when executing the program code: generating a corresponding EEG signal waveform graph based on initial EEG signal feature data; and extracting EEG signal waveform features based on the EEG signal waveform graph.

[0171] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: performing Fourier transform on the initial feature data of the EEG signal to obtain EEG signal time-frequency data corresponding to the EEG signal feature data; and extracting the EEG signal time-frequency features based on the EEG signal time-frequency data.

[0172] In some embodiments of the present application, the processor of the device is further used to implement the following steps when executing the program code: resampling, filtering, removing noise data, and numerical standardization.

[0173] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: obtaining the human feature information and motion feature information of the subject from the monitoring video; inputting the potential difference, EEG signal time-frequency characteristics, human feature information and motion feature information into the trained target epileptic focus localization model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model.

[0174] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: respectively extracting multiple frame images corresponding to the monitoring video; respectively detecting object key points corresponding to the test object and key point spatial coordinates corresponding to the object key points from the multiple frame images; the object key points include object facial key points and body key points; constructing human body feature information based on all key point spatial coordinates of the test object for the target object key points; performing human motion analysis on the human body feature information to obtain motion feature information of the test object.

[0175] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: taking a preset time point as the starting time, based on a preset time interval, the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information, and the motion characteristic information are respectively divided to obtain multiple divided segments; and the multiple divided segments are input into a trained target epileptogenic focus localization model.

[0176] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: using the feature processing layer to extract video feature data from the monitoring video, and fusing the intracranial EEG feature data and the video feature data to obtain fused features; using the attention layer to characterize the fused features to obtain embedded features; using the output layer to map the embedded features to the output space to obtain the positioning result of the target epileptogenic focus corresponding to the subject.

[0177] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: obtaining training samples, the training samples include multiple patient data corresponding to multiple epilepsy patients; the multiple patient data include intracranial EEG feature data, patient monitoring videos and epilepsy seizure labels respectively; the epilepsy seizure labels are used to characterize the epileptic seizure state and epileptic focus of epilepsy patients; the training samples are input into the initial epileptic focus localization type to be trained to obtain the model output result; the model output result includes the epileptic seizure state output result and the epileptic focus output result; according to the model output result and the epileptic seizure label, determining the loss value of the initial epileptic focus localization model to be trained; according to the loss value, iteratively training the initial epileptic focus localization type to be trained until a trained target epileptic focus localization model is obtained.

[0178] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: the training samples include positive samples and negative samples, the positive samples include multiple patient data corresponding to multiple epilepsy patients with abnormal discharges, and the negative samples include multiple patient data corresponding to multiple epilepsy patients without abnormal discharges.

[0179] In some embodiments of the present application, the processor of the device is also used to implement the following steps when executing the program code: inputting the training sample into the initial epileptic focus localization type to be trained to obtain the predicted probability; the predicted probability includes the predicted probability of epileptic seizure and the predicted probability of epileptic focus; determining the model output result corresponding to the training sample according to the predicted probability.

[0180] In some embodiments of the present application, the processor of the device is further used to implement the following steps when executing the program code: updating the model parameters of the initial epileptic focus localization model to be trained according to the loss value; until the loss value meets the preset convergence condition, the trained target epileptic focus localization model is obtained.

[0181] It should be noted that the above-mentioned medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0182] A readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, wherein readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to: electromagnetic signals, optical signals, or any suitable combination of the above. A readable signal medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0183] The program code contained on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0184] Program code for performing the operations of the present application may be written in any combination of one or more programming languages, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user computing device, partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0185] Exemplary Devices

[0186] After introducing the medium of the exemplary embodiment of the present application, next, reference is made to Figure 8 The device for localizing an epileptic focus according to an exemplary embodiment of the present application is described.

[0187] See also Figure 8 , Figure 8 FIG. 2 shows a structural block diagram of an epileptic focus localization device according to an exemplary embodiment of the present application. Figure 8 As shown, an epileptic focus locating device 800 according to an exemplary embodiment of the present application includes: an intracranial feature data acquisition unit 810, a monitoring video acquisition unit 820, and a target epileptic focus locating unit 830, wherein:

[0188] The intracranial characteristic data acquisition unit 810 is used to acquire intracranial EEG characteristic data of the subject;

[0189] The monitoring video acquisition unit 820 is used to acquire the monitoring video of the object under test;

[0190] The target epileptogenic focus positioning unit 830 is used to input the intracranial EEG feature data and the monitoring video into the trained target epileptogenic focus positioning model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus positioning model.

[0191] It can be seen that implementation Figure 8 The device shown can input intracranial EEG characteristic data and monitoring video into the trained target epileptogenic focus positioning model to locate the target epileptogenic focus of the subject through the target epileptogenic focus positioning model. Compared with the prior art, the implementation of the embodiment of the present application, on the one hand, avoids determining the epileptogenic focus through consultation with multiple doctors, and combines intracranial EEG characteristic data and monitoring video to improve the accuracy and efficiency of locating the epileptogenic focus; on the other hand, the method of locating the epileptogenic focus can be applied not only during the patient's illness, but also before surgery, which improves the accuracy of preoperative positioning and increases the chance of successful surgery.

[0192] In one embodiment, based on the above scheme, the intracranial characteristic data acquisition unit 810 acquires intracranial EEG characteristic data of the subject, including:

[0193] The original intracranial electroencephalogram (EEG) monitoring data of the test subject collected by the biomedical monitoring equipment is obtained; the original intracranial electroencephalogram (EEG) monitoring data is preprocessed; and the intracranial electroencephalogram (EEG) characteristic data is obtained according to the preprocessed original intracranial electroencephalogram (EEG) monitoring data.

[0194] It can be seen that the implementation of this optional embodiment can avoid directly using the original intracranial EEG monitoring data to locate the epileptogenic focus, thereby improving the accuracy of locating the epileptogenic focus.

[0195] In one embodiment, based on the aforementioned scheme, obtaining the original intracranial EEG monitoring data of the intracranial feature data unit 810 includes: multi-channel EEG monitoring data collected from multiple locations in the skull of the subject; obtaining intracranial EEG feature data based on the preprocessed original intracranial EEG monitoring data, including: calculating the potential difference between each channel and the reference electrode based on the preprocessed multi-channel EEG monitoring data to obtain initial feature data of the EEG signal; extracting intracranial EEG feature data based on the initial feature data of the EEG signal.

[0196] It can be seen that by implementing this optional embodiment, the potential difference between each channel and the reference electrode can be calculated to finally obtain intracranial EEG characteristic data that can reflect whether the brain of the subject is in abnormal discharge at this time.

[0197] In one embodiment, based on the above scheme, the intracranial feature data acquisition unit 810 extracts intracranial EEG feature data according to the initial feature data of the EEG signal, including: generating a corresponding EEG signal waveform graph according to the initial feature data of the EEG signal; and extracting EEG signal waveform features according to the EEG signal waveform graph.

[0198] It can be seen that by implementing this optional embodiment, an EEG signal waveform corresponding to the initial characteristic data of the EEG signal can be obtained, and thus the EEG fluctuation of the subject can be visually reflected in real time.

[0199] In one embodiment, based on the aforementioned scheme, the intracranial feature data acquisition unit 810 extracts intracranial EEG feature data based on the initial feature data of the EEG signal, including: performing Fourier transform on the initial feature data of the EEG signal to obtain EEG signal time-frequency data corresponding to the EEG signal feature data; and extracting EEG signal time-frequency features based on the EEG signal time-frequency data.

[0200] It can be seen that by implementing this optional embodiment, the time-frequency characteristics of the EEG signal reflecting the real-time EEG changes of the subject can be obtained by performing Fourier transform on the initial characteristic data of the EEG signal.

[0201] In one embodiment, based on the above scheme, the intracranial feature data acquisition unit 810 pre-processes the original intracranial EEG monitoring data, including at least one of the following processing: resampling, filtering, noise data removal, and numerical normalization.

[0202] It can be seen that by implementing this optional embodiment, by preprocessing the original intracranial EEG monitoring data, the influence of some unnecessary data on the intracranial EEG characteristic data can be reduced, thereby increasing the accuracy of the intracranial EEG characteristic data.

[0203] In one embodiment, based on the aforementioned scheme, after the monitoring video acquisition unit 820 acquires the monitoring video of the subject, the method further includes: obtaining the human body feature information and motion feature information of the subject from the monitoring video; inputting the potential difference, the time-frequency features of the EEG signal, the human body feature information and the motion feature information into the trained target epileptic focus localization model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model.

[0204] It can be seen that by implementing this optional embodiment, the data input into the trained target epileptic focus localization model includes, in addition to the potential difference and the time-frequency characteristics of the EEG signal, also human body feature information and motion feature information, which increases the accuracy of the target epileptic focus finally determined.

[0205] In one embodiment, based on the aforementioned scheme, the monitoring video acquisition unit 820 obtains the human feature information and motion feature information of the test object from the monitoring video, including: extracting multiple frame images corresponding to the monitoring video respectively; detecting the object key points corresponding to the test object and the key point spatial coordinates corresponding to the object key points from the multiple frame images respectively; the object key points include the object facial key points and the body key points; constructing the human feature information according to the spatial coordinates of all key points of the test object for the target object key points; performing human motion analysis on the human feature information to obtain the motion feature information of the test object.

[0206] It can be seen that by implementing this optional embodiment, human feature information reflecting key points of the human object and motion feature information reflecting human motion can be obtained by monitoring the video, and then the human feature information and motion feature information can be input into the trained target epileptic focus localization model to improve the accuracy of epileptic focus localization.

[0207] In one embodiment, based on the aforementioned scheme, the monitoring video unit 820 is used to input the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information into a trained target epileptic focus localization model, including: taking a preset time point as the starting time, and based on a preset time interval, dividing the potential difference, EEG signal time-frequency characteristics, human body characteristic information and motion characteristic information respectively to obtain multiple divided segments; and inputting the multiple divided segments into the trained target epileptic focus localization model.

[0208] It can be seen that by implementing this optional embodiment, multiple divided segments are input into the target epileptic focus positioning model, and the obtained target epileptic focus corresponds to the potential difference of a specific segment, the time-frequency characteristics of the EEG signal of a segment, the human body characteristic information of a segment, and the action characteristic information of a segment, so as to better understand the characteristic changes caused by the epileptic focus in the subject.

[0209] In one embodiment, based on the aforementioned scheme, the target epileptic focus positioning unit 830 target epileptic focus positioning model includes a feature processing layer, an attention layer, and an output layer; the intracranial EEG feature data and the monitoring video are input into the trained target epileptic focus positioning model to locate the target epileptic focus corresponding to the subject through the target epileptic focus positioning model, including: using the feature processing layer to extract video feature data from the monitoring video, and fusing the intracranial EEG feature data and the video feature data to obtain fused features; using the attention layer to characterize the fused features to obtain embedded features; using the output layer to map the embedded features to the output space to obtain the positioning result of the target epileptic focus corresponding to the subject.

[0210] It can be seen that by implementing this optional embodiment, the intracranial EEG feature data and monitoring video are input into the trained target epileptic focus localization model, which avoids the situation of determining the epileptic focus through consultation with multiple doctors, and improves the accuracy and efficiency of epileptic focus localization.

[0211] In one embodiment, based on the aforementioned scheme, the target epileptic focus positioning unit 830 also includes: obtaining training samples, the training samples include multiple patient data corresponding to multiple epilepsy patients respectively; the multiple patient data include intracranial EEG feature data, patient monitoring videos and epileptic seizure labels respectively; the epileptic seizure labels are used to characterize the epileptic seizure state and epileptic focus of epileptic patients; the training samples are input into the initial epileptic focus positioning type to be trained to obtain the model output result; the model output result includes the epileptic seizure state output result and the epileptic focus output result; according to the model output result and the epileptic seizure label, determining the loss value of the initial epileptic focus positioning model to be trained; according to the loss value, iteratively training the initial epileptic focus positioning type to be trained until a trained target epileptic focus positioning model is obtained.

[0212] It can be seen that, by implementing this optional embodiment, the trained target epileptic focus localization model is obtained by training the initial epileptic focus localization model with multiple training samples, which increases the prediction accuracy of the trained target epileptic focus localization model.

[0213] In one embodiment, based on the above scheme, the training samples include positive samples and negative samples. The positive samples include multiple patient data corresponding to multiple epilepsy patients with abnormal discharges, and the negative samples include multiple patient data corresponding to multiple epilepsy patients without abnormal discharges.

[0214] It can be seen that by implementing this optional embodiment, the training samples include positive samples and negative samples. By inputting the positive samples and negative samples into the initial epileptic focus localization model, the initial epileptic focus localization model can more accurately predict whether the subject is in an epileptic seizure period at this time, and more accurately predict the target epileptic focus that causes the subject to be in an epileptic seizure period.

[0215] In one embodiment, based on the aforementioned scheme, the target epileptic focus positioning unit 830 inputs the training sample into the initial epileptic focus positioning type to be trained to obtain a model output result, including: inputting the training sample into the initial epileptic focus positioning type to be trained to obtain a predicted probability; the predicted probability includes the predicted probability of epileptic seizure and the predicted probability of epileptic focus; and determining the model output result corresponding to the training sample according to the predicted probability.

[0216] It can be seen that by implementing this optional embodiment, the prediction result of the initial epileptogenic focus localization type to be trained for the training sample can be more intuitively known through the prediction probability.

[0217] In one embodiment, based on the above scheme, the target epileptic focus localization unit 830 iteratively trains the initial epileptic focus localization model to be trained according to the loss value until a trained target epileptic focus localization model is obtained, including:

[0218] According to the loss value, the model parameters of the initial epileptic focus localization model to be trained are updated; until the loss value meets the preset convergence condition, the trained target epileptic focus localization model is obtained.

[0219] It can be seen that, when this optional embodiment is implemented, the adjustment of the model parameters of the initial epileptic focus localization model is stopped only when the loss value satisfies the preset convergence condition, so as to improve the prediction accuracy of the trained target epileptic focus localization model.

[0220] It should be noted that although several modules or units of the epileptogenic focus localization device are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiment of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0221] Exemplary Electronic Devices

[0222] After introducing the method, medium, and apparatus according to the exemplary embodiment of the present application, next, an electronic device according to another exemplary embodiment of the present application is introduced.

[0223] Those skilled in the art will appreciate that various aspects of the present application may be implemented as a system, method or program product. Therefore, various aspects of the present application may be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to as "circuit", "module" or "system" herein.

[0224] Refer to the following Fig. 9 An electronic device 900 according to such an embodiment of the present invention will be described. Fig. 9 The electronic device 900 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0225] like Fig. 9 As shown, the electronic device 900 is in the form of a general computing device. The components of the electronic device 900 may include, but are not limited to: the at least one processing unit 910, the at least one storage unit 920, a bus 930 connecting different system components (including the storage unit 920 and the processing unit 910), and a display unit 940.

[0226] The storage unit stores program codes, which can be executed by the processing unit 910, so that the processing unit 910 executes the steps according to various exemplary embodiments of the present invention described in the above “Exemplary Method” section of this specification.

[0227] The storage unit 920 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 921 and / or a cache memory unit 922 , and may further include a read-only memory unit (ROM) 923 .

[0228] The storage unit 920 may also include a program / utility 924 having a set (at least one) of program modules 925, such program modules 925 including but not limited to: an operating system, one or more application programs, other program modules and program data, each of which or some combination may include the reality of a network environment.

[0229] The bus 930 may include a data bus, an address bus, and a control bus.

[0230] The electronic device 900 may also communicate with one or more external devices 970 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), and such communication may be performed via an input / output (I / O) interface 950. Furthermore, the electronic device 900 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 960. As shown, the network adapter 960 communicates with other modules of the electronic device 900 via a bus 930. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 900, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0231] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the embodiment of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiment of the present disclosure.

[0232] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the above method of the present specification is stored. In some possible embodiments, various aspects of the present invention can also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to perform the steps according to various exemplary embodiments of the present invention described in the above "Exemplary Method" section of the present specification.

[0233] Although the spirit and principle of the present application have been described with reference to several specific embodiments, it should be understood that the present application is not limited to the specific embodiments of the invention, and the division of various aspects does not mean that the features in these aspects cannot be combined to benefit, and such division is only for the convenience of expression. The present application is intended to cover various modifications and equivalent arrangements included in the spirit and scope of the attached claims.

Claims

1. A method for localizing epileptogenic foci, It is characterized in that The method comprises: Acquiring intracranial EEG characteristic data of the tested subject; Acquiring a monitoring video of the subject; Inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model; After obtaining the monitoring video of the object under test, the method further includes: Obtaining human feature information and motion feature information of the subject from the monitoring video; inputting the potential difference, the time-frequency characteristics of the EEG signal, the human feature information and the motion feature information into the trained target epileptic focus localization model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus localization model; Among them, the potential difference is obtained in the following way: the potential difference between each channel and the reference electrode is calculated based on the multi-channel EEG monitoring data of the subject; the time-frequency characteristics of the EEG signal are obtained in the following way: Fourier transform is performed on the initial characteristic data of the EEG signal of the subject to obtain the time-frequency data of the EEG signal, and the time-frequency characteristics of the EEG signal are extracted based on the time-frequency data of the EEG signal.

2. The method according to claim 1, It is characterized in that The step of obtaining intracranial EEG characteristic data of the subject includes: Acquiring original intracranial electroencephalogram (EEG) monitoring data of the subject collected by a biomedical monitoring device; Preprocessing the raw intracranial EEG monitoring data; The intracranial EEG characteristic data is obtained based on the preprocessed raw intracranial EEG monitoring data.

3. The method according to claim 2, It is characterized in that The original intracranial EEG monitoring data includes: multi-channel EEG monitoring data collected from multiple locations in the brain of the subject; The intracranial EEG characteristic data is obtained according to the preprocessed original intracranial EEG monitoring data, including: The potential difference between each channel and the reference electrode is calculated based on the preprocessed multi-channel EEG monitoring data to obtain the initial characteristic data of the EEG signal; Intracranial EEG characteristic data is extracted according to the initial EEG signal characteristic data.

4. The method according to claim 3, It is characterized in that The intracranial EEG characteristic data includes EEG signal waveform characteristics; The step of extracting intracranial EEG characteristic data according to the initial characteristic data of the EEG signal comprises: Generate a corresponding EEG signal waveform diagram according to the initial EEG signal characteristic data; Extracting EEG signal waveform features according to the EEG signal waveform diagram.

5. The method according to claim 3, It is characterized in that The intracranial EEG characteristic data includes EEG signal time-frequency characteristics; The step of extracting intracranial EEG characteristic data according to the initial characteristic data of the EEG signal comprises: Performing Fourier transform on the initial characteristic data of the EEG signal to obtain time-frequency data of the EEG signal corresponding to the characteristic data of the EEG signal; The time-frequency features of the EEG signal are extracted according to the time-frequency data of the EEG signal.

6. The method according to claim 2, It is characterized in that The preprocessing of the original intracranial EEG monitoring data includes at least one of the following processing: resampling, filtering, noise data removal, and numerical standardization.

7. The method according to claim 1, It is characterized in that The subject carries a monitoring device on his head, and the monitoring device includes a camera for collecting the monitoring video in real time.

8. The method according to claim 1, It is characterized in that The obtaining of the human body feature information and the motion feature information of the subject from the monitoring video includes: Extracting a plurality of frame images corresponding to the monitoring video respectively; Detecting object key points corresponding to the detected object and key point spatial coordinates corresponding to the object key points from the multiple frame images respectively; the object key points include object facial key points and body key points; Constructing human body feature information according to the spatial coordinates of all key points of the tested object for the key points of the target object; The human body feature information is subjected to human body motion analysis to obtain the motion feature information of the subject.

9. The method according to claim 1, It is characterized in that The step of inputting the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information and the motion characteristic information into the trained target epileptogenic focus localization model comprises: Taking a preset time point as a starting time, based on a preset time interval, the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information, and the motion characteristic information are divided respectively to obtain a plurality of divided segments; The multiple segmented segments are input into the trained target epileptogenic focus localization model.

10. The method according to claim 1, It is characterized in that The target epileptogenic focus localization model includes a feature processing layer, an attention layer, and an output layer; The step of inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model, comprises: Extracting video feature data from the monitoring video using the feature processing layer, and fusing the intracranial EEG feature data and the video feature data to obtain a fusion feature; Using the attention layer to characterize the fused features to obtain embedded features; The output layer is used to map the embedded features to an output space to obtain a positioning result of a target epileptogenic focus corresponding to the subject.

11. The method according to claim 1, It is characterized in that The method further comprises: Acquire training samples, wherein the training samples include multiple patient data corresponding to multiple epilepsy patients; the multiple patient data include intracranial electroencephalogram feature data, patient monitoring videos, and epilepsy attack labels; the epilepsy attack labels are used to characterize the epilepsy attack state and epileptogenic focus of the epilepsy patient; Inputting the training samples into the initial epileptogenic focus localization model to be trained to obtain a model output result; the model output result includes an epileptic seizure state output result and an epileptogenic focus output result; Determining a loss value of the initial epileptic focus localization model to be trained according to the model output result and the epileptic seizure label; According to the loss value, the initial epileptic focus localization model to be trained is iteratively trained until a trained target epileptic focus localization model is obtained.

12. The method according to claim 11, It is characterized in that The training samples include positive samples and negative samples. The positive samples include multiple patient data corresponding to multiple epilepsy patients with abnormal discharges, and the negative samples include multiple patient data corresponding to multiple epilepsy patients without abnormal discharges.

13. The method according to claim 11, It is characterized in that The step of inputting the training sample into the initial epileptogenic focus localization model to be trained to obtain a model output result includes: Inputting the training samples into the initial epileptogenic focus localization model to be trained to obtain a prediction probability; the prediction probability includes an epileptic seizure prediction probability and an epileptogenic focus prediction probability; Determine the model output result corresponding to the training sample according to the predicted probability.

14. The method according to claim 11, It is characterized in that The iterative training of the initial epileptic focus localization model to be trained according to the loss value until a trained target epileptic focus localization model is obtained includes: According to the loss value, updating the model parameters of the initial epileptic focus localization model to be trained; When the loss value satisfies a preset convergence condition, a trained target epileptogenic focus localization model is obtained.

15. An epileptogenic focus positioning device, It is characterized in that The device comprises: An intracranial characteristic data acquisition unit is used to acquire intracranial EEG characteristic data of the subject; A monitoring video acquisition unit is used to acquire the monitoring video of the object under test; a target epileptic focus positioning unit, used for inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptic focus positioning model, so as to locate the target epileptic focus corresponding to the subject through the target epileptic focus positioning model; The device is further used to obtain the monitoring video of the subject, and then obtain the human body feature information and motion feature information of the subject from the monitoring video; input the potential difference, the time-frequency characteristics of the EEG signal, the human body feature information and the motion feature information into the trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model; Among them, the potential difference is obtained in the following way: the potential difference between each channel and the reference electrode is calculated based on the multi-channel EEG monitoring data of the subject; the time-frequency characteristics of the EEG signal are obtained in the following way: Fourier transform is performed on the initial characteristic data of the EEG signal of the subject to obtain the time-frequency data of the EEG signal, and the time-frequency characteristics of the EEG signal are extracted based on the time-frequency data of the EEG signal.

16. The device according to claim 15, It is characterized in that The step of obtaining intracranial EEG characteristic data of the subject includes: Acquiring original intracranial electroencephalogram (EEG) monitoring data of the subject collected by a biomedical monitoring device; Preprocessing the raw intracranial EEG monitoring data; The intracranial EEG characteristic data is obtained based on the preprocessed raw intracranial EEG monitoring data.

17. The device according to claim 16, It is characterized in that The original intracranial EEG monitoring data includes: multi-channel EEG monitoring data collected from multiple locations in the brain of the subject; The intracranial EEG characteristic data is obtained according to the preprocessed original intracranial EEG monitoring data, including: The potential difference between each channel and the reference electrode is calculated based on the preprocessed multi-channel EEG monitoring data to obtain the initial characteristic data of the EEG signal; Intracranial EEG characteristic data is extracted according to the initial EEG signal characteristic data.

18. The device according to claim 17, It is characterized in that The intracranial EEG characteristic data includes EEG signal waveform characteristics; The step of extracting intracranial EEG characteristic data according to the initial characteristic data of the EEG signal comprises: Generate a corresponding EEG signal waveform diagram according to the initial EEG signal characteristic data; Extracting EEG signal waveform features according to the EEG signal waveform diagram.

19. The device according to claim 17, It is characterized in that The intracranial EEG characteristic data includes EEG signal time-frequency characteristics; The step of extracting intracranial EEG characteristic data according to the initial characteristic data of the EEG signal comprises: Performing Fourier transform on the initial characteristic data of the EEG signal to obtain time-frequency data of the EEG signal corresponding to the characteristic data of the EEG signal; The time-frequency features of the EEG signal are extracted according to the time-frequency data of the EEG signal.

20. The device according to claim 16, It is characterized in that The preprocessing of the original intracranial EEG monitoring data includes at least one of the following processing: resampling, filtering, noise data removal, and numerical standardization.

21. The device according to claim 15, It is characterized in that The subject carries a monitoring device on his head, and the monitoring device includes a camera for collecting the monitoring video in real time.

22. The device according to claim 15, It is characterized in that The obtaining of the human body feature information and the motion feature information of the subject from the monitoring video includes: Extracting a plurality of frame images corresponding to the monitoring video respectively; Detecting object key points corresponding to the detected object and key point spatial coordinates corresponding to the object key points from the multiple frame images respectively; the object key points include object facial key points and body key points; Constructing human body feature information according to the spatial coordinates of all key points of the tested object for the key points of the target object; The human body feature information is subjected to human body motion analysis to obtain the motion feature information of the subject.

23. The device according to claim 15, It is characterized in that The step of inputting the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information and the motion characteristic information into the trained target epileptogenic focus localization model comprises: Taking a preset time point as a starting time, based on a preset time interval, the potential difference, the time-frequency characteristics of the EEG signal, the human body characteristic information, and the motion characteristic information are divided respectively to obtain a plurality of divided segments; The multiple segmented segments are input into the trained target epileptogenic focus localization model.

24. The device according to claim 15, It is characterized in that The target epileptogenic focus localization model includes a feature processing layer, an attention layer, and an output layer; The step of inputting the intracranial EEG characteristic data and the monitoring video into a trained target epileptogenic focus localization model, so as to locate the target epileptogenic focus corresponding to the subject through the target epileptogenic focus localization model, comprises: Extracting video feature data from the monitoring video using the feature processing layer, and fusing the intracranial EEG feature data and the video feature data to obtain a fusion feature; Using the attention layer to characterize the fused features to obtain embedded features; The output layer is used to map the embedded features to an output space to obtain a positioning result of a target epileptogenic focus corresponding to the subject.

25. The device according to claim 15, It is characterized in that The device also includes: Acquire training samples, wherein the training samples include multiple patient data corresponding to multiple epilepsy patients; the multiple patient data include intracranial electroencephalogram feature data, patient monitoring videos, and epilepsy attack labels; the epilepsy attack labels are used to characterize the epilepsy attack state and epileptogenic focus of the epilepsy patient; Inputting the training samples into the initial epileptogenic focus localization model to be trained to obtain a model output result; the model output result includes an epileptic seizure state output result and an epileptogenic focus output result; Determining a loss value of the initial epileptic focus localization model to be trained according to the model output result and the epileptic seizure label; According to the loss value, the initial epileptic focus localization model to be trained is iteratively trained until a trained target epileptic focus localization model is obtained.

26. The device according to claim 25, It is characterized in that The training samples include positive samples and negative samples. The positive samples include multiple patient data corresponding to multiple epilepsy patients with abnormal discharges, and the negative samples include multiple patient data corresponding to multiple epilepsy patients without abnormal discharges.

27. The device according to claim 25, It is characterized in that The step of inputting the training sample into the initial epileptogenic focus localization model to be trained to obtain a model output result includes: Inputting the training samples into the initial epileptogenic focus localization model to be trained to obtain a prediction probability; the prediction probability includes an epileptic seizure prediction probability and an epileptogenic focus prediction probability; Determine the model output result corresponding to the training sample according to the predicted probability.

28. The device according to claim 25, It is characterized in that The iterative training of the initial epileptic focus localization model to be trained according to the loss value until a trained target epileptic focus localization model is obtained includes: According to the loss value, updating the model parameters of the initial epileptic focus localization model to be trained; When the loss value satisfies a preset convergence condition, a trained target epileptogenic focus localization model is obtained.

29. An electronic device, It is characterized in that include: processor; A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1-14 by executing the executable instructions.

30. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 14 is implemented.

Images