Real-time epilepsy monitoring method, device and equipment and storage medium

Abstract

The invention provides an epilepsy real-time monitoring method, device and equipment and a storage medium, and the method comprises the steps: obtaining electroencephalogram signals of multiple paths of electrodes in real time, and carrying out data preprocessing on the electroencephalogram signals to obtain preprocessed signals; linear features and local phase values in the preprocessed signals are extracted, the linear features exceeding a first threshold value are transmitted to a monitoring model, the local phase values exceeding a second threshold value are transmitted to the monitoring model, and when the proportion of the number of the local phase values transmitted to the monitoring model exceeds a preset proportion, the monitoring model is started; a global phase value is calculated according to all the local phase values, and if the global phase value exceeds a third threshold value, the global phase value is transmitted to a monitoring model; and splicing all the received values as fusion features through a monitoring model, and carrying out real-time monitoring on epilepsy based on the fusion features. According to the invention, the features can be obtained progressively according to the disease development condition, so that the monitoring model can obtain effective information capable of reflecting the nature, and the precision and effect of epilepsy monitoring are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring technology, and in particular to a real-time epilepsy monitoring method, device, equipment and storage medium. Background Art

[0002] Epilepsy is a disease in which sudden abnormal discharges of brain neurons lead to brain dysfunction. The corresponding intelligent monitoring method can assist in neurological diagnosis.

[0003] In the related technology, online epilepsy monitoring can obtain the linear and nonlinear features in epilepsy signals through learning, and combine machine learning algorithms for real-time monitoring. Although it can achieve high monitoring performance, it requires high computing energy consumption. However, if it is a monitoring algorithm that only relies on a single feature, the false detection rate is high. Therefore, in the current real-time epilepsy detection, there is still a contradiction between the monitoring performance and computing energy consumption that is difficult to balance. In addition, epilepsy signals have obvious characteristic differences in the three stages of pre-ictal, interictal and ictal.

[0004] Based on the above analysis of the development status of this technical field, the existing technology lacks a monitoring solution based on linear features and nonlinear features, and configures corresponding thresholds to filter out features that can reflect the essence for monitoring, so as to balance performance and computing energy consumption. Summary of the invention

[0005] The purpose of the present invention is to provide a real-time epilepsy monitoring method, device, equipment and storage medium, aiming to solve the above-mentioned problems in the prior art.

[0006] According to a first aspect of an embodiment of the present invention, a real-time epilepsy monitoring method is provided, comprising:

[0007] Acquire EEG signals of multiple electrodes in real time, perform data preprocessing on the EEG signals, and obtain preprocessed signals;

[0008] Extracting linear features and local phase values ​​from the preprocessed signal, transmitting linear features exceeding a first threshold to a monitoring model, transmitting local phase values ​​exceeding a second threshold to a monitoring model, and when the proportion of local phase values ​​transmitted to the monitoring model exceeds a preset proportion, calculating a global phase value based on all local phase values, and transmitting the global phase value to the monitoring model if the global phase value exceeds a third threshold;

[0009] All received values ​​are spliced ​​together as fusion features through the monitoring model, and real-time monitoring of epilepsy is performed based on the fusion features.

[0010] According to a second aspect of an embodiment of the present invention, there is provided a real-time epilepsy monitoring device, comprising:

[0011] A preprocessing module is used to obtain EEG signals of multiple electrodes in real time, perform data preprocessing on the EEG signals, and obtain preprocessed signals;

[0012] A progressive screening module, used to extract linear features and local phase values ​​from the preprocessed signal, transmit linear features exceeding a first threshold to the monitoring model, transmit local phase values ​​exceeding a second threshold to the monitoring model, and when the proportion of the number of local phase values ​​transmitted to the monitoring model exceeds a preset proportion, calculate a global phase value based on all local phase values, and if the global phase value exceeds a third threshold, transmit the global phase value to the monitoring model;

[0013] The hierarchical monitoring module is used to splice all received values ​​as fusion features through the monitoring model, and perform real-time monitoring of epilepsy based on the fusion features.

[0014] According to a third aspect of an embodiment of the present invention, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the real-time epilepsy monitoring method provided in the first aspect of the present disclosure are implemented.

[0015] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which a program for implementing information transmission is stored, and when the program is executed by a processor, the steps of the real-time epilepsy monitoring method provided in the first aspect of the present disclosure are implemented.

[0016] The technical solution provided by the embodiment of the present invention includes the following beneficial effects: using multiple electrodes to perform zoned monitoring of EEG, while considering linear and nonlinear features, and starting global phase synchronization calculation only when the local phase synchronization calculation in the nonlinear feature exceeds the corresponding threshold by a certain proportion, and being able to progressively acquire features according to the progression of the disease, and regardless of whether it is a linear feature or a nonlinear feature, only the features that exceed the corresponding threshold are sent to the monitoring model, so that the monitoring model can obtain effective information that can reflect the essence of the corresponding stage, thereby improving the accuracy and effect of epilepsy monitoring.

[0017] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate one or more embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0019] Figure 1 is a flow chart of a method for real-time monitoring of epilepsy according to an embodiment of the present invention;

[0020] Figure 2 is a schematic diagram of an overall framework of real-time monitoring according to an embodiment of the present invention;

[0021] Figure 3 is a schematic diagram of a real-time monitoring device for epilepsy according to an embodiment of the present invention;

[0022] Figure 4 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will be combined with the drawings in one or more embodiments of this specification to clearly and completely describe the technical solutions in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0024] Method Embodiment

[0025] According to an embodiment of the present invention, a real-time monitoring method for epilepsy is provided. Figure 1 is a flow chart of a real-time epilepsy monitoring method according to an embodiment of the present invention. Figure 1 As shown, the real-time epilepsy monitoring method according to an embodiment of the present invention specifically includes:

[0026] In step S110, the electroencephalogram signals of multiple electrodes are acquired in real time, and data preprocessing is performed on the electroencephalogram signals to obtain preprocessed signals, which specifically includes:

[0027] In the embodiment of the present invention, the brain circuit signal is an EEG signal in the field;

[0028] The high-frequency components in the EEG signal are removed to obtain low-frequency signals. The removed high-frequency components are signals above 50 Hz, and the removed high-frequency components are not related to epilepsy.

[0029] At this time, the low-frequency signal also contains other task-related signals. The task-irrelevant signals in the low-frequency signal are removed to obtain the preprocessed signal. The preprocessed signal is more conducive to the feature extraction of EEG data, thereby revealing the physiological or pathological state of the brain.

[0030] In step S120, linear features and local phase values ​​in the preprocessed signal are extracted, linear features exceeding a first threshold are transmitted to the monitoring model, and local phase values ​​exceeding a second threshold are transmitted to the monitoring model. When the proportion of the number of local phase values ​​transmitted to the monitoring model exceeds a preset proportion, a global phase value is calculated based on all local phase values. If the global phase value exceeds a third threshold, the global phase value is transmitted to the monitoring model, specifically including:

[0031] During an epileptic seizure, the amplitude of the EEG wave rises sharply, and high-amplitude and high-frequency spike discharges are generated inside the brain, making the collected EEG signal have obvious time domain characteristics. The computational complexity of these linear features is low. In the interval and pre-seizure period, the time domain and frequency domain information of the linear features of the EEG signal is not clear enough. However, EEG phase synchronization, as a macroscopic characteristic of epilepsy, can reflect the state of the epileptic patient. The computational complexity is more complicated than that of linear features, but it can improve the accuracy of monitoring. Due to the differences in each stage, using only a single feature cannot be accurate.

[0032] Therefore, the embodiments of the present invention consider both linear and nonlinear characteristics;

[0033] Analyze the time domain waveform of the preprocessed signal, extract the amplitude, energy and fluctuation relationship of the preprocessed signal as time domain features, perform wavelet transform or empirical mode analysis on the preprocessed signal to obtain frequency domain features, and use the time domain features and frequency domain features as linear features; preferably, the time domain waveform can be transformed into the frequency domain to obtain linear features at different frequencies;

[0034] Each electrode region can obtain a corresponding linear feature;

[0035] Based on the preprocessed signal, the phase synchronization degree between the two electrodes is calculated as the corresponding local phase value between the two electrodes. The local phase value is the phase synchronization degree between two specific electrodes under the two leads. All local phase values ​​are calculated two by two in sequence. For example, the multi-lead electrodes in the embodiment of the present invention are 4 electrodes, that is, 4 EEG areas are monitored, and the local phase values ​​between 12, 13, 14, 23, 24, and 34 need to be calculated.

[0036] When the proportion of the number of local phase values ​​transmitted to the monitoring model exceeds a preset proportion, the global phase evaluation is started. In the embodiment of the present invention, the preset proportion is 50%, that is, when more than 3 of the 6 local phase values ​​exceed the second threshold, the global phase calculation is started;

[0037] As epilepsy progresses, the synchronization between brain regions will continue to increase, and the synchronization between some brain regions will spread from one region to other regions, causing the overall synchronization of brain regions to increase. The average of all local phase values ​​is calculated as the global phase value, which is directly calculated from the values ​​monitored in the previous two leads.

[0038] In the embodiment of the present invention, the first threshold, the second threshold and the third threshold are all obtained by summarizing the experience of historical data, and preferably, can be adjusted according to the actual situation of the patient to be detected.

[0039] In step S130, all received values ​​are spliced ​​as fusion features through the monitoring model, and real-time monitoring of epilepsy is performed based on the fusion features, specifically including:

[0040] All received linear features, local phase values ​​and global phase values ​​are concatenated to obtain fused features;

[0041] Since only features exceeding their respective corresponding thresholds can be transmitted to the monitoring model, the types or quantities of features received at each moment are different, so the fused features corresponding to different stages or situations can reflect the essential characteristics, while the features that cannot express the essential characteristics do not participate in the execution of the monitoring model;

[0042] As the severity of epilepsy progresses, its fusion characteristics have a progressive grading characteristic;

[0043] Based on the fusion features, the monitoring model can be used to directly obtain the prediction results of the epilepsy stage, where the monitoring model is a pre-trained neural network model, and the epilepsy stage includes the pre-ictal period, interictal period and ictal period;

[0044] For example, during the attack period with obvious characteristics, a large number of obvious linear features can be obtained, and the monitoring model can use a smaller computational complexity to draw the conclusion that the patient is in the attack period. In the interval and early stage of the attack, fewer linear features are obtained, and the monitoring model needs to use nonlinear features to obtain accurate conclusions. Although the computational complexity is larger than that of the attack period model, the fusion feature is the key feature after threshold screening, which reduces the computational energy consumption to a certain extent compared to the unscreened one, and ensures the monitoring accuracy.

[0045] The monitoring model can also be used to determine the severity of the disease in each EEG area of ​​multiple electrodes. Preferably, improving the performance of the neural network model can assist in more complex analysis tasks;

[0046] In the embodiment of the present invention, the neural network model includes but is not limited to: a convolutional neural network, a recurrent neural network, a fully connected neural network, etc.

[0047] The method further comprises:

[0048] In step S140, the value of the first threshold, the second threshold or the third threshold is adaptively adjusted according to the result of real-time monitoring.

[0049] The above technical solutions of the embodiments of the present invention are illustrated with reference to the following drawings.

[0050] Figure 2 Schematic diagram of the overall real-time monitoring framework of an embodiment of the present invention. Figure 2 As shown, the complete framework design of progressive graded prediction is demonstrated, with linear features and nonlinear features as the dual branches of the framework, and the first threshold, the second threshold and the third threshold are set to filter the features transmitted to the monitoring model, and the preset proportion is used to start the global phase calculation, so that the monitoring model can obtain key features to reflect the progressive development of the disease.

[0051] To sum up, in response to the existing problems, the real-time monitoring method for epilepsy invented this time adopts multi-channel electrodes to monitor EEG in different areas, while considering linear and nonlinear features; the global phase synchronization calculation is started only when the local phase synchronization calculation in the nonlinear feature exceeds the corresponding threshold by more than a certain proportion, and the features can be acquired progressively according to the development of the disease. The local phase value refers to the phase synchronization degree between the two electrodes, and the global phase value refers to the overall synchronization degree, that is, the overall situation needs to be further evaluated only when the number of local asynchronies increases; and no matter it is a linear feature or a nonlinear feature, only the features exceeding the corresponding threshold are sent to the monitoring model, so that the fusion features obtained by the monitoring model can reflect the effective information of the essence of the corresponding stage; in general, the use of a rich variety of features ensures the inspection performance to a certain extent, and the threshold setting reduces the computational complexity while strengthening the degree of distinction of the fusion features in different situations, thereby improving the accuracy and effect of epilepsy monitoring and balancing the contradiction between the monitoring performance effect and the computational energy consumption.

[0052] Device Embodiment

[0053] According to an embodiment of the present invention, a real-time monitoring device for epilepsy is provided. Figure 3 Schematic diagram of a real-time epilepsy monitoring device according to an embodiment of the present invention. Figure 3 As shown, the real-time epilepsy monitoring device according to an embodiment of the present invention specifically includes:

[0054] The preprocessing module 30 is used to obtain EEG signals of multiple electrodes in real time, perform data preprocessing on the EEG signals, and obtain preprocessed signals, which are specifically used for:

[0055] Remove the high-frequency components in the EEG signal to obtain the low-frequency signal;

[0056] The task-irrelevant signals in the low-frequency signals are removed to obtain the preprocessed signals.

[0057] The progressive screening module 32 is used to extract linear features and local phase values ​​in the preprocessed signal, transmit the linear features exceeding the first threshold to the monitoring model, transmit the local phase values ​​exceeding the second threshold to the monitoring model, and when the proportion of the number of local phase values ​​transmitted to the monitoring model exceeds the preset proportion, calculate the global phase value according to all local phase values, and if the global phase value exceeds the third threshold, transmit the global phase value to the monitoring model, specifically for:

[0058] The amplitude, energy and fluctuation relationship of the preprocessed signal are extracted as time domain features, and the preprocessed signal is subjected to wavelet transform or empirical mode analysis to obtain frequency domain features, and the time domain features and frequency domain features are used as linear features;

[0059] Based on the preprocessed signal, the phase synchronization degree between the two electrodes is calculated as the local phase value corresponding to the two electrodes, and all the local phase values ​​are calculated in pairs in sequence.

[0060] The average of all local phase values ​​is calculated as the global phase value.

[0061] The hierarchical monitoring module 34 is used to combine all received values ​​as fusion features through the monitoring model and perform real-time monitoring of epilepsy based on the fusion features, specifically for:

[0062] All received linear features, local phase values, and global phase values ​​are concatenated to obtain fused features.

[0063] Based on the fusion features, the monitoring model is used to obtain the prediction results of the epilepsy stage, where the monitoring model is a pre-trained neural network model. The epilepsy stages include the pre-ictal period, interictal period and ictal period; and the severity of the disease in each EEG area of ​​multiple electrodes is judged.

[0064] The device further comprises:

[0065] The adjustment module 36 is used to adaptively adjust the value of the first threshold, the second threshold or the third threshold according to the result of real-time monitoring.

[0066] To sum up, in response to the existing problems, the real-time monitoring device for epilepsy invented this time adopts multi-channel electrodes to monitor EEG in different areas, taking into account linear and nonlinear characteristics at the same time; the global phase synchronization calculation is started only when the local phase synchronization calculation in the nonlinear characteristics exceeds the corresponding threshold number by more than a certain proportion, and the characteristics can be acquired progressively according to the development of the disease. The local phase value refers to the phase synchronization degree between the two electrodes, and the global phase value refers to the overall synchronization degree, that is, when the number of local asynchronies increases, it is necessary to further evaluate the whole; and no matter it is a linear feature or a nonlinear feature, only the features exceeding the corresponding threshold are sent to the monitoring model, so that the fusion features obtained by the monitoring model can reflect the effective information of the essence of the corresponding stage; in general, the use of a rich variety of features ensures the inspection performance to a certain extent, and the threshold setting reduces the calculation complexity while strengthening the degree of distinction of the fusion features in different situations, thereby improving the accuracy and effect of epilepsy monitoring and balancing the contradiction between the monitoring performance effect and the computing energy consumption.

[0067] Electronic device embodiment

[0068] Figure 4 4 is a schematic diagram of an electronic device according to an embodiment of the present invention. The electronic device 400 may include at least one processor 410 and a memory 420. The processor 410 may execute instructions stored in the memory 420. The processor 410 is connected to the memory 420 through a data bus. In addition to the memory 420, the processor 410 may also be connected to an input device 430, an output device 440, and a communication device 450 through a data bus.

[0069] The processor 410 may be any conventional processor, such as a commercially available CPU. The processor may also include a graphics processor (Graphic Process Unit, GPU), a field programmable gate array (Field Programmable Gate Array, FPGA), a system on chip (System on Chip, SOC), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC) or a combination thereof.

[0070] Memory 420 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0071] In the embodiment of the present disclosure, executable instructions are stored in the memory 420, and the processor 410 can read the executable instructions from the memory 420 and execute the instructions to implement all or part of the steps of any of the real-time epilepsy monitoring methods in the above exemplary embodiments.

[0072] Computer Readable Storage Medium Embodiments

[0073] In addition to the above-mentioned methods and apparatus, an exemplary embodiment of the present disclosure may also be a computer program product or a computer-readable storage medium storing the computer program product, wherein the computer product includes computer program instructions that can be executed by a processor to implement all or part of the steps described in any of the real-time epilepsy monitoring methods in the above-mentioned exemplary embodiments.

[0074] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages ​​such as Java, C++, etc., and also conventional procedural programming languages ​​such as "C" language or similar programming languages ​​and scripting languages ​​(e.g., Python). The program code may be executed entirely on the user computing device, partially on the user computing device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0075] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples of readable storage media include: a static random access memory (SRAM) with one or more wires electrically connected, an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk, or any suitable combination of the above.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A real-time epilepsy monitoring method, characterized in that: include: Acquire electroencephalogram signals of multiple electrodes in real time, perform data preprocessing on the electroencephalogram signals, and obtain preprocessed signals; Extracting linear features and local phase values ​​from the preprocessed signal, transmitting linear features exceeding a first threshold to a monitoring model, transmitting local phase values ​​exceeding a second threshold to a monitoring model, and when the proportion of the number of local phase values ​​transmitted to the monitoring model exceeds a preset proportion, calculating a global phase value based on all local phase values, and transmitting the global phase value to the monitoring model if the global phase value exceeds a third threshold; All received values ​​are spliced ​​together as fusion features through the monitoring model, and real-time monitoring of epilepsy is performed based on the fusion features.

2. The method according to claim 1, characterized in that: The performing data preprocessing on the electroencephalogram signal to obtain the preprocessed signal specifically comprises: Removing high-frequency components from the electroencephalogram signal to obtain a low-frequency signal; The task-irrelevant signal in the low-frequency signal is removed to obtain a preprocessed signal.

3. The method according to claim 1, characterized in that The extracting of the linear features and the local phase value in the preprocessed signal specifically includes: Extracting the amplitude, energy and fluctuation relationship of the preprocessed signal as time domain features, performing wavelet transform or empirical mode analysis on the preprocessed signal to obtain frequency domain features, and using the time domain features and the frequency domain features as linear features; Based on the preprocessed signal, the phase synchronization degree between the two electrodes is calculated as the local phase value corresponding to the two electrodes, and all the local phase values ​​are calculated in pairs in sequence.

4. The method according to claim 1, characterized in that Calculating the global phase value according to all local phase values ​​specifically includes: calculating an average value of all local phase values ​​as the global phase value.

5. The method according to claim 1, characterized in that: The splicing of all received values ​​as fused features through the monitoring model specifically includes: splicing all received linear features, local phase values ​​and global phase values ​​to obtain the fused features.

6. The method according to claim 1, characterized in that The real-time monitoring of epilepsy based on the fusion feature specifically includes: Based on the fusion features, the monitoring model is used to obtain a prediction result of the epilepsy stage, wherein the monitoring model is a pre-trained neural network model, and the epilepsy stage includes the pre-ictal period, the interictal period and the ictal period; and the severity of the disease in each EEG area of ​​multiple electrodes is determined.

7. The method according to claim 1, characterized in that The method further comprises: According to the result of real-time monitoring, the value of the first threshold, the second threshold or the third threshold is adaptively adjusted.

8. A real-time epilepsy monitoring device, characterized in that: include: A preprocessing module, used for acquiring EEG signals of multiple electrodes in real time, performing data preprocessing on the EEG signals, and obtaining preprocessed signals; A progressive screening module, used to extract linear features and local phase values ​​in the preprocessed signal, transmit linear features exceeding a first threshold to a monitoring model, transmit local phase values ​​exceeding a second threshold to the monitoring model, and when the proportion of the number of local phase values ​​transmitted to the monitoring model exceeds a preset proportion, calculate a global phase value based on all local phase values, and if the global phase value exceeds a third threshold, transmit the global phase value to the monitoring model; The hierarchical monitoring module is used to splice all received values ​​as fusion features through the monitoring model, and perform real-time monitoring of epilepsy based on the fusion features.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of the real-time epilepsy monitoring method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by a processor, the steps of the real-time epilepsy monitoring method according to any one of claims 1 to 7 are implemented.

Images