Electroencephalogram signal detection method, electroencephalogram signal detection device, electronic equipment and medium
The pre-trained children's EEG signal detection model detects children's EEG signal, which solves the problem of poor accuracy of EEG signal detection in children and achieves efficient and accurate detection of EEG signal in children.
Patent Information
- Application Number
- CN202411117717.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-12-03
AI Technical Summary
The existing technology is difficult to accurately and effectively detect children's EEG signals, especially because children's EEG signals are greatly affected by age factors. The general adult EEG signal detection model has low accuracy in children and high false positive rates. .
The pre-trained EEG signal detection model for children is used to detect the EEG signal of children. By obtaining the sample EEG signal of multiple samples of children and the corresponding label training model, an EEG signal detection model dedicated to children is obtained. Determine whether the EEG signal is abnormal and sleep state.
Improve the accuracy and effectiveness of EEG signal detection in children, can accurately output whether EEG signal is abnormal and the status during detection, provide auxiliary effects on EEG signal detection in children, overcome the accuracy of detection in children in the prior art Problems with low and high false positive rates.
Smart Images

Figure CN119055249A_ABST
Abstract
Description
Background Art
[0002] In the field of medical technology, the human brain electrical signals can accurately reflect the state of the human body. For example, through the electroencephalogram, neurological diseases such as epilepsy can be effectively judged, and then determine what kind of treatment help or medication strategy to provide for the patient. In the prior art, the detection of electroencephalograms usually relies on expert detection. However, this method, on the one hand, with the gradual increase in the demand for electroencephalogram signal analysis, requires more time and manpower costs, increasing the workload of experts; on the other hand, the detection standards of different experts are not unified, and the accuracy and efficiency of manual detection of electroencephalogram signals are difficult to guarantee. In addition, electroencephalogram signals are greatly affected by age factors, and the general adult electroencephalogram signal detection method cannot effectively detect children's electroencephalogram signals.
[0003] Therefore, how to effectively, accurately and efficiently detect children's EEG signals is a problem that needs to be urgently solved by existing technologies.
[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] The present disclosure provides an EEG signal detection method, an EEG signal detection device, an electronic device and a computer-readable storage medium, thereby overcoming, at least to a certain extent, the problem that the prior art is unable to accurately and effectively detect children's EEG signals.
[0006] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.
[0007] According to one aspect of the present disclosure, there is provided an EEG signal detection method, comprising: obtaining an EEG signal to be detected of a target child; detecting the EEG signal to be detected by using a pre-trained child EEG signal detection model to obtain an EEG signal detection result of the target child; wherein the EEG signal detection result comprises a first detection result and a second detection result, the first detection result being used to reflect whether the EEG signal to be detected is abnormal, and the second detection result being used to reflect the state of the EEG signal to be detected.
[0008] According to one aspect of the present disclosure, there is provided an EEG signal detection device, comprising: a signal acquisition module, for acquiring an EEG signal to be detected of a target child; a signal detection module, for detecting the EEG signal to be detected by using a pre-trained child EEG signal detection model to obtain an EEG signal detection result of the target child; wherein the EEG signal detection result comprises a first detection result and a second detection result, the first detection result being used to reflect whether the EEG signal to be detected is abnormal, and the second detection result being used to reflect the state of the EEG signal to be detected.
[0009] According to one aspect of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the above method by executing the executable instructions.
[0010] According to one aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented.
[0011] The exemplary embodiments of the present disclosure have the following beneficial effects:
[0012] Obtain the EEG signal to be detected of the target child; detect the EEG signal to be detected by a pre-trained child EEG signal detection model to obtain the EEG signal detection result of the target child; wherein the EEG signal detection result includes a first detection result and a second detection result, the first detection result is used to reflect whether the EEG signal to be detected is abnormal, and the second detection result is used to reflect the state of the EEG signal to be detected. On the one hand, this exemplary embodiment proposes a new brain signal detection method, which can detect the EEG signal of a child through a dedicated child EEG signal detection model to obtain the EEG signal detection result of the target child. Compared with the prior art that uses a general alarm signal detection model to detect the EEG signal of a child, this exemplary embodiment uses a child EEG signal detection model trained with a sample EEG signal of a sample child to perform EEG signal detection, which is more targeted and accurate. On the other hand, this exemplary embodiment can obtain an EEG signal detection result including a first detection result and a second detection result by detecting the EEG signal to be detected of the target child through the child EEG signal detection model. The detection result can not only include the result of whether the EEG signal to be detected is abnormal, but also accurately output the state of the EEG signal to be detected during the detection, which provides auxiliary effect for the detection of the EEG signal to be detected and ensures the accuracy and effectiveness of the EEG signal detection result of the target child.
[0013] 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
[0014] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.
[0015] Figure 1 A schematic diagram schematically shows the architecture of an electroencephalogram signal detection method in this exemplary embodiment;
[0016] Figure 2 A schematic diagram schematically shows the architecture of another method for detecting EEG signals in this exemplary embodiment;
[0017] Figure 3 A flowchart of a method for detecting an electroencephalogram signal in this exemplary embodiment is schematically shown;
[0018] Figure 4 A schematic diagram schematically shows an EEG signal of a sample child in a sleeping state;
[0019] Figure 5 A schematic diagram of an EEG signal of a sample child in a waking state is shown schematically;
[0020] Figure 6 A sub-flow chart schematically illustrates a method for detecting an electroencephalogram signal in this exemplary embodiment;
[0021] Figure 7 A sub-flow chart schematically illustrates another method for detecting an electroencephalogram signal in this exemplary embodiment;
[0022] Figure 8 A training flow chart of a child's EEG signal detection model in this exemplary embodiment is schematically shown;
[0023] Fig. 9 A block diagram schematically shows a structure of an electroencephalogram signal detection device in this exemplary embodiment;
[0024] Fig.10 An electronic device for implementing the above method in this exemplary embodiment is schematically shown. DETAILED DESCRIPTION
[0025] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the disclosure will be more comprehensive and complete and to fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
[0026] An exemplary embodiment of the present disclosure first provides a method for detecting an electroencephalogram signal.
[0027] Figure 1 A schematic diagram of the system architecture of the operating environment of this exemplary embodiment is shown. Figure 1 As shown, the system 100 may include a target child 110 and an EEG signal acquisition device 120, wherein the EEG signal acquisition device may be a device in various forms including a terminal, a helmet, a hat, etc. configured with detection electrodes, such as a computer, a server, a wearable device, or other smart devices, etc., and a child EEG signal detection model may be configured in the EEG signal acquisition device accordingly. By connecting the detection electrodes of the EEG signal acquisition device 120 with the target child 110, the EEG signal to be detected of the target child 110 may be acquired. After the EEG signal to be detected is acquired, it may be detected by the EEG signal detection model configured in the EEG signal acquisition device to obtain the EEG signal detection result of the target child 110.
[0028] Specifically, during the acquisition process, the multiple detection electrodes of the EEG signal acquisition device 120 can be fixed at various positions on the head of the target child 110. After the interface of the EEG signal acquisition device 120 is connected to the multiple detection electrodes respectively, the multi-channel EEG signal of the target child 110 can be acquired, and the multi-channel EEG signal can be used as the EEG signal to be detected. At the same time, a camera can also be configured in the system to synchronously capture the patient's face and torso video of the target child through the camera, record the user's movement state, and combine the movement state judgment to assist in judging the validity of the EEG signal detection result.
[0029] Figure 2 Another system architecture diagram of the operating environment of this exemplary embodiment is shown, referring to Figure 2As shown, the system 200 may include a target child 210, an EEG signal acquisition device 220 and a server 230. The EEG signal acquisition device 220 is used to acquire the EEG signal to be detected of the target child 210. After acquiring the EEG signal to be detected, it can be sent to the server 230, and detected by the EEG signal detection model configured in the server 230 to obtain the EEG signal detection result of the target child 110, and return it to the EEG signal acquisition device 220, such as returning the EEG signal detection result and displaying the EEG signal detection result in the display terminal corresponding to the EEG signal acquisition device. In addition, the EEG signal detection result can also be returned to a designated terminal, such as a terminal device such as a user's mobile phone or computer.
[0030] Based on the above description, the method in this exemplary embodiment can be applied to Figure 1 The EEG signal acquisition device 120 shown in the figure or Figure 2 On the server 230 shown.
[0031] The following is combined with Figure 3 This exemplary embodiment is further described as follows. Figure 3 As shown, the EEG signal detection method may include the following steps S310-S320:
[0032] Step S310, obtaining the EEG signal to be detected of the target child.
[0033] The target child refers to a child whose EEG signal needs to be detected, and the EEG signal to be detected may refer to the brain discharge signal of the target child, such as brain waves, electroencephalograms formed by brain waves, etc. This exemplary embodiment can obtain the target child's EEG signal to be detected through a variety of EEG signal acquisition devices, such as using a specific EEG signal acquisition device to contact the target child's head skin through electrodes and interfaces to collect the target child's brain discharge signal, and use the collected brain discharge signal as the target child's EEG signal to be detected.
[0034] EEG signals, such as brain discharge signals or electroencephalograms, as an important detection method in the current medical detection field, can accurately reflect whether there are problems with the human body state. For example, electroencephalograms can be used as an important inspection tool for the diagnosis and treatment of epilepsy. Epileptic discharges are an important marker of epilepsy, and their clinical significance has been widely confirmed. Especially in pediatric patients, electroencephalograms are of great significance for diagnosis, treatment and drug withdrawal. In related technologies, in addition to detecting interictal epileptic discharges, staff will also use the discharge signals during the attack period for detection. However, considering that the incidence of epileptic seizures is much lower than that of interictal epileptic discharges, and is usually combined with clinical symptoms that can be recognized by patients and caregivers, the EEG signal to be detected in this exemplary embodiment can be an interictal epileptic discharge signal, that is, an EEG signal between epileptic seizures. This signal can better explore whether the detection object is abnormal or potentially abnormal, and broaden the detection range of epilepsy.
[0035] After acquiring the EEG signal to be detected, the present exemplary embodiment can also perform preprocessing on the EEG signal, such as intelligent detection of abnormally connected leads / ECG / eye movement / myography artifacts, and establish an intelligent preprocessing model to ensure its effectiveness. The EEG signal to be detected is tested for effectiveness through various pre-trained artifact models to eliminate artifacts with strong interference. When it is determined to be an invalid EEG signal (i.e., artifact), its cause can be determined, such as abnormal lead connection, ECG, eye movement, EMG, baseline drift, etc. According to the detection cause of the invalid EEG signal, it can be further processed, such as removing abnormal leads, removing eye movement / ECG signals, or intelligently filling in the signal trends before and after the missing signal, etc., to re-determine the valid EEG signal to be detected for detection.
[0036] Step S320, detecting the EEG signal to be detected by using the pre-trained children's EEG signal detection model to obtain the EEG signal detection result of the target child;
[0037] Among them, the children's EEG signal detection model is trained by using sample EEG signals of multiple sample children and the first category sample labels and second category sample labels corresponding to the sample EEG signals.
[0038] The children's EEG signal detection model refers to a pre-trained neural network model for detecting children's EEG signals, and its model structure can be customized as needed. The model can be trained by sample EEG signals of sample children, as well as the corresponding first-category sample labels and second-category sample labels.
[0039] Considering that EEG is greatly affected by age, there are obvious differences in EEG activities at different ages. Compared with adults, children's EEG has characteristic age-related physiological EEG patterns. For example, the following aspects can be considered: first, there is no occipital dominant rhythm in the EEG background of newborns. As age increases, the occipital dominant rhythm gradually forms in infants and preschoolers, and gradually develops from slow waves to slow alpha rhythms. It is not until adolescence that the occipital alpha rhythm consistent with that of adults is formed. Compared with adults, children have higher amplitude of rhythm in the occipital region, with many inserted slow waves, which are easily confused with sharp slow waves. Second, children may have significant rhythm changes in the anterior leads during the sleep-wake period and the transition from sleep to wakefulness, which also need to be differentiated from rhythmic epileptic discharges. Sometimes small spikes of no more than 100uV may be inserted between rhythmic slow waves, which are immature spike-slow complex waves. This is a phenomenon without clinical diagnostic significance in children, but it is meaningful in adults. Third, the physiological waves of sleep in early childhood, such as the sharp wave components of the top wave and K-complex wave, are prominent and sharp, with a high amplitude, which can be easily confused with generalized spike-slow complex waves. Fourth, the types of epilepsy in children and adults are different. Idiopathic generalized discharges are more common in children, and characteristic EEG abnormal patterns of epilepsy syndromes such as high-amplitude arrhythmias also appear in child patients. Fifth, child patients have poor compliance during EEG monitoring, and there are more family members accompanying and interference, resulting in more artifacts than adults. It can be determined that the pathological and physiological differences between children and adults' EEGs include but are not limited to the above examples, which results in low accuracy and many false positives if the adult EEG interictal epileptiform discharge detection model is directly used for children. If the detection mode of adult EEG signals is directly transferred to children's EEG signal detection, it will be difficult to achieve the expected effect due to the difference between the physiological EEG pattern and the interictal epileptiform discharge pattern and the interference of activity artifacts, and the accuracy and sensitivity are poor. Therefore, this exemplary embodiment obtains a children's EEG signal detection model specifically used to detect children's EEG signals through training of sample EEG signals of multiple sample children and the first type of sample labels and the second type of sample labels corresponding to the sample EEG signals.
[0040] Among them, the first type of sample labels and the second type of sample labels refer to sample labels of two dimensions of different types. The first type of sample labels can be classification data used to determine whether the EEG signal is an abnormal signal, and the second type of sample labels can be classification data used to determine whether the EEG signal is in a sleep state and what level the sleep state is, etc.
[0041] After detecting the EEG signal to be detected by using a dedicated children's EEG signal detection model, this exemplary embodiment may include two detection results: a first detection result and a second detection result. The first detection result may reflect whether the EEG signal to be detected is abnormal, for example, determining whether the EEG signal to be detected is an epileptic-like discharge pattern; the second detection result may reflect the state of the EEG signal to be detected, for example, whether the EEG signal to be detected is detected in a sleeping state or a non-sleeping state, and when in a sleeping state, what is the specific level of the sleeping state, and so on.
[0042] Based on the above description, in this exemplary embodiment, the EEG signal to be detected of the target child is obtained; the EEG signal to be detected is detected by a pre-trained child EEG signal detection model to obtain the EEG signal detection result of the target child; wherein the EEG signal detection result includes a first detection result and a second detection result, the first detection result is used to reflect whether the EEG signal to be detected is abnormal, and the second detection result is used to reflect the state of the EEG signal to be detected. On the one hand, this exemplary embodiment proposes a new brain signal detection method, which can detect the EEG signal of a child through a dedicated child EEG signal detection model to obtain the EEG signal detection result of the target child. Compared with the prior art that uses a general alarm signal detection model to detect the EEG signal of a child, this exemplary embodiment uses a child EEG signal detection model trained with a sample EEG signal of a sample child to detect EEG signals, which is more targeted and accurate. On the other hand, this exemplary embodiment can obtain an EEG signal detection result including a first detection result and a second detection result by detecting the EEG signal to be detected of the target child through the child EEG signal detection model. The detection result can not only include the result of whether the EEG signal to be detected is abnormal, but also accurately output the state of the EEG signal to be detected during the detection, which provides auxiliary effect for the detection of the EEG signal to be detected and ensures the accuracy and effectiveness of the EEG signal detection result of the target child.
[0043] In an exemplary embodiment, the above-mentioned children's EEG signal detection model is trained by the following method:
[0044] Obtaining sample EEG signals of multiple sample children;
[0045] Obtain the first type sample label and the second type sample label corresponding to the sample EEG signal;
[0046] The children's EEG signal detection model is trained based on the sample EEG signals, the first category sample labels and the second category sample labels.
[0047] The training of the children's EEG signal detection model can be achieved by obtaining a large number of sample EEG signals of sample children, as well as the corresponding first-category sample labels and second-category sample labels, and using these data as training data to train the machine learning model.
[0048] In an exemplary embodiment, the first type of sample label is a classification label of whether the sample EEG signal is an abnormal EEG signal, such as a label of whether the pre-labeled signal is an epileptic discharge signal, which can be represented by a "1" or "0" mark, such as "1" for abnormality and "0" for normal. The second type of sample label is a level label of the sleep state corresponding to the sample EEG signal, such as pre-segmenting the sample EEG signal to determine whether the sample EEG signal is a sleep state or a non-sleep state. Specifically, the second type of sample label can be a detailed level label of the sleep state in different time periods, which can reflect the state of sleep or not, or the state of sleep depth, such as "0" for non-sleep state, "1" for sleep state level 1 or light sleep, "2" for sleep state level 2 or moderate sleep, "3" for sleep state level 3 or deep sleep, etc. The level label of the specific sleep yellow platform can be customized in advance as needed.
[0049] Figure 4 A schematic diagram of a sample EEG signal of a sample child in a sleeping state in this exemplary embodiment is shown. The sample EEG signal may be a multi-channel EEG signal, including multiple EEG signal waveforms, representing data detected from different parts of the head. For example, the waveforms of EEG signals detected from the front, left, right, back, and top of the head may have certain differences. Figure 5 The schematic diagram of the EEG signal of the sample child in the awake state is shown, including multiple EEG signal waveforms. It can be seen that the EEG signals of the child in the sleeping state and the awake state have certain differences, and the fluctuation in the sleeping state is more obvious. Therefore, through the above process, this exemplary embodiment can determine the abnormal judgment result of the EEG signal to be detected on the one hand, and on the other hand, it can also obtain the level of the EEG signal to be detected in the awake or sleeping state, or even the sleeping state, and can accurately and effectively detect and judge the EEG signal to be detected from a richer dimension.
[0050] In an exemplary embodiment, if Figure 6 As shown, the above-mentioned training of the children's EEG signal detection model based on the sample EEG signal, the first type of sample label and the second type of sample label may include the following steps:
[0051] Step S610, inputting the sample EEG signal into the child EEG signal detection model to be trained to obtain the sample EEG signal detection result;
[0052] Step S620, calculating a loss function based on the sample EEG signal detection result and the first type sample label and the second type sample label;
[0053] Step S630, optimizing the child EEG signal detection model to be trained by minimizing the loss function to obtain a trained child EEG signal detection model.
[0054] The children's EEG signal detection model to be trained can be an initial untrained machine learning model, and the parameters therein can be customized and initialized as needed, or can be randomly set. This exemplary embodiment can calculate the loss function based on the sample EEG signal detection results and the first type of sample labels and the second type of sample labels, and optimize the children's EEG signal detection model to be trained by minimizing the loss function to obtain a trained children's EEG signal detection model. The process of optimizing the children's EEG signal detection model to be trained is essentially the process of optimizing the parameters in the model.
[0055] Figure 7 A flowchart of a child EEG signal detection model training is shown, which may specifically include: obtaining a sample EEG signal 710, inputting it into a child EEG signal detection model 720 to be trained, and obtaining a predicted sample EEG signal detection result 730, which may include a first detection result of whether the sample EEG signal is abnormal, and also includes a state result of the sample EEG signal; then, combining the first category sample label and the second category sample label 740 of the pre-acquired sample EEG signal, calculating a loss function 750, and by minimizing the loss function 750, the parameters of the child EEG signal detection model 720 to be trained can be optimized to obtain a trained child EEG signal detection model.
[0056] In an exemplary embodiment, the sample EEG signal detection result includes a first sample detection result and a second sample detection result; the above step S620 may include:
[0057] Determine a first loss function according to the difference between the first sample detection result and the first class sample label;
[0058] Determine a second loss function according to the difference between the second sample detection result and the second class sample label;
[0059] Determine a target loss function according to the first loss function and the second loss function;
[0060] The above step S630 may include:
[0061] By minimizing the target loss function, the children's EEG signal detection model to be trained is optimized to obtain a trained children's EEG signal detection model.
[0062] Taking into account that the present exemplary embodiment includes sample labels of two different dimensions, when calculating the loss function, different loss functions can be calculated separately, and then fused into a target loss function according to the different loss functions, and the children's EEG signal detection model to be trained can be optimized based on the target loss function.
[0063] Specifically, after processing the sample EEG signals, the children's EEG signal detection model to be trained can obtain two detection results, namely, the first sample detection result and the second sample detection result, wherein the first sample detection result can be a classification result of whether the EEG signal is abnormal, and the first loss function can be determined by calculating the difference between the first sample detection result and the first category sample label; the second sample detection result can be the state result of the EEG signal, such as the sleep state sequence of the EEG signal, such as the awake state from t1 to t2, the first sleep stage from t2 to t3, the second sleep stage from t3 to t4, etc. Based on the time series and the corresponding state, a sleep state sequence of the EEG signal can be generated, that is, the second sample detection result, and the second category sample label can also be a sleep state sequence label. Therefore, the second loss function is determined by calculating the difference between the second sample detection result and the second category sample label. Furthermore, the target loss function can be determined based on the first loss function and the second loss function. For example, the target loss function can be determined by performing weighted calculation on the first loss function and the second loss function. Finally, the trained child EEG signal detection model can be optimized by minimizing the target loss function.
[0064] In an exemplary embodiment, if Figure 8 As shown, the above-mentioned acquisition of the first type of sample labels and the second type of sample labels corresponding to the sample EEG signals may include the following steps:
[0065] Step S810, segmenting the sample EEG signal according to a preset timing rule to obtain sub-signal segments of the sample EEG signal within different time ranges;
[0066] Step S820, obtaining a first label of each sub-signal segment, and determining a first type of sample label of the sample EEG signal according to the first label;
[0067] Step S830: Obtain a second label for each sub-signal segment, and determine a second type of sample label for the sample EEG signal according to the second label.
[0068] Taking into account that EEG signals are often brain discharge data over a period of time, detection and analysis based on data at a certain moment, or summarizing long-term data results with short-term detection results are often not accurate enough. Therefore, this exemplary embodiment can disassemble and analyze the sample EEG signals in detail and comprehensively from the time dimension. Specifically, the sample EEG signals can be segmented according to preset timing rules. For example, multiple sub-signal segments can be divided into regular or irregular time windows in the sample EEG signal, and the time windows corresponding to the multiple sub-signal segments may or may not have intersections. Then, the first label and the second label of each sub-signal segment are obtained respectively, and the first type of sample label is determined according to the first label, and the second type of sample label is determined according to the second label. The difference between the first label and the first type of sample label here is that the first label can be the label of the sub-signal segment, and the first labels of different sub-signal segments can be different. For example, the sample EEG signal in the time period t can be divided into sub-signal segments corresponding to the three time ranges of t1-t2, t2-t3, and t3-t4. The signal classification information corresponding to t1-t2, t2-t3, and t3-t4 is "abnormal, abnormal, abnormal" or "abnormal", "abnormal", and "normal". The first type of sample label can be the label of the entire sample EEG signal composed of the sub-signal segments. The first type of sample label can be determined according to the first label. For example, the signal classification information corresponding to t1-t2, t2-t3, and t3-t4 is " Abnormal, abnormal, abnormal" or "abnormal", "abnormal", "normal", it can be seen that most of the sub-signal segments show abnormal results, then the first category sample label can be determined as "abnormal"; for another example, the second labels corresponding to t1-t2, t2-t3, and t3-t4 are "non-sleep state, sleep state level 1, sleep state level 1", and the second category sample label can be determined as "non-sleep state, sleep state level 1, sleep state level 1" or "0-1-1", and the second labels corresponding to t1-t2, t2-t3, and t3-t4 are "non-sleep state, sleep state level 1, sleep state level 2", and the second category sample label can be determined as "non-sleep state, sleep state level 1, sleep state level 1" or "0-1-2". This is just a simple example, and the expression structure of the actual sample label can be set with a variety of different characters or logic settings as needed.
[0069] In this exemplary embodiment, the first category sample labels and the second category sample labels can be determined by expert annotation. For example, after the sample EEG signal is segmented, the expert annotates each sub-signal segment with the first label and the second label, and then the expert determines the first category sample labels based on the statistical results of the first label, and determines the second category sample labels based on the statistical results of the second label. In addition, the first category sample labels and the second category sample labels can also be determined by comparative analysis and detection with historical data.
[0070] It should be noted that, considering that the first-class sample labels and the second-class sample labels reflect different meanings, for example, whether the first-class sample labels are abnormal or not can often be judged from a global perspective, and the status of the second-class sample labels can be determined locally. Therefore, when obtaining the first-class sample labels and the second-class sample labels, different timing rules can be used to divide the sub-signal segments. For example, when determining the first-class sample labels, the sample EEG signal can be divided into one sub-signal segment, that is, the original sample EEG signal is used as a sub-signal segment, and the signal classification information of the original sample EEG signal is directly used as the first-class sample label; when determining the second-class sample label, the sample EEG signal is divided into multiple sub-signal segments. That is, according to different division standards, the sample EEG signal can be subjected to a first segmentation process according to the first standard, and the first-class sample label can be determined according to the signal classification information of the sub-signal segment after the segmentation process, and the sample EEG signal can be subjected to a second segmentation process according to the second standard, and the second-class sample label can be determined according to the signal status information of the sub-signal segment after the segmentation process, and so on.
[0071] In an exemplary embodiment, after obtaining the EEG signal to be detected of the target child, the EEG signal detection method may include:
[0072] Extract the basic attribute information of the target children and determine the age range of the target children;
[0073] According to the age range of the target children, a pre-trained children's EEG signal detection model corresponding to the age group is determined.
[0074] Among them, the basic attribute information can be basic attribute information such as the age, school age, and gender of the target child. According to the basic attribute information, the age range of the target child can be accurately determined, for example, the age group to which the target child belongs can be determined. This exemplary embodiment takes into account the specificity of the EEG signals of the child group. The characteristics of the EEG signals of children in different age ranges are also different. For example, the EEG signals of newborns and children over 10 years old may also be different. Therefore, this exemplary embodiment can train multiple children's EEG signal detection models, each of which corresponds to a child's age range. After determining the age range of the target child, the child EEG signal detection model corresponding to the age range can be determined first, and the EEG signal to be detected is processed by the child EEG signal detection model corresponding to the age range to improve the accuracy, effectiveness and pertinence of EEG signal detection.
[0075] The exemplary embodiments of the present disclosure also provide an electroencephalogram signal detection device. Fig. 9The device 900 may include a signal acquisition module 910, which is used to acquire the EEG signal to be detected of the target child; a signal detection module 920, which is used to detect the EEG signal to be detected by using a pre-trained child EEG signal detection model to obtain the EEG signal detection result of the target child; wherein the EEG signal detection result includes a first detection result and a second detection result, the first detection result is used to reflect whether the EEG signal to be detected is abnormal, and the second detection result is used to reflect the state of the EEG signal to be detected.
[0076] In an exemplary embodiment, the child EEG signal detection model is configured to be trained by the following units: a sample signal acquisition unit, used to acquire sample EEG signals of multiple sample children; a sample label acquisition unit, used to acquire first-category sample labels and second-category sample labels corresponding to the sample EEG signals; and a model training unit, used to train the child EEG signal detection model based on the sample EEG signals, the first-category sample labels and the second-category sample labels.
[0077] In an exemplary embodiment, the model training unit includes: a sample detection result acquisition subunit, which is used to input the sample EEG signal into the child EEG signal detection model to be trained to obtain the sample EEG signal detection result; a loss function calculation subunit, which is used to calculate the loss function based on the sample EEG signal detection result and the first category sample label and the second category sample label; a model optimization subunit, which is used to optimize the child EEG signal detection model to be trained by minimizing the loss function to obtain a trained child EEG signal detection model.
[0078] In an exemplary embodiment, the loss function calculation subunit includes: a first loss function determination subunit, used to determine the first loss function based on the difference between the first sample detection result and the first category sample label; a second loss function determination subunit, used to determine the second loss function based on the difference between the second sample detection result and the second category sample label; a target loss function determination subunit, used to determine the target loss function based on the first loss function and the second loss function; a model optimization subunit, used to optimize the child EEG signal detection model to be trained by minimizing the target loss function to obtain a trained child EEG signal detection model.
[0079] In an exemplary embodiment, the sample signal acquisition unit includes: a signal segmentation subunit, which is used to segment the sample EEG signal according to a preset timing rule to obtain sub-signal segments of the sample EEG signal within different time ranges; a first label acquisition subunit, which is used to obtain a first label for each sub-signal segment, and determine a first type of sample label for the sample EEG signal according to the first label; and a second label acquisition subunit, which is used to obtain a second label for each sub-signal segment, and determine a second type of sample label for the sample EEG signal according to the second label.
[0080] In an exemplary embodiment, the first type of sample label is a classification label of whether the sample EEG signal is an abnormal EEG signal, and the second type of sample label is a level label of the sample EEG signal corresponding to the sleep state.
[0081] In an exemplary embodiment, after obtaining the EEG signal to be detected of the target child, the EEG signal detection device includes: an age range determination unit, used to extract basic attribute information of the target child and determine the age range of the target child; and a model determination unit, used to determine a pre-trained child EEG signal detection model corresponding to the age range based on the age range of the target child.
[0082] The specific details of each module / unit in the above-mentioned device have been described in detail in the embodiments of the method part. The undisclosed details can be found in the embodiments of the method part, so they will not be repeated here.
[0083] Exemplary embodiments of the present disclosure also provide an electronic device capable of implementing the above method.
[0084] Those skilled in the art will appreciate that various aspects of the present disclosure may be implemented as systems, methods or program products. Therefore, various aspects of the present disclosure may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to herein as "circuits", "modules" or "systems".
[0085] Refer to the following Fig.10 1000 according to such an exemplary embodiment of the present disclosure is described. Fig.10 The electronic device 1000 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.
[0086] like Fig.10 As shown, the electronic device 1000 is in the form of a general computing device. The components of the electronic device 1000 may include but are not limited to: the at least one processing unit 1010, the at least one storage unit 1020, a bus 1030 connecting different system components (including the storage unit 1020 and the processing unit 1010), and a display unit 1040.
[0087] The storage unit stores program codes, which can be executed by the processing unit 1010, so that the processing unit 1010 executes the steps according to various exemplary embodiments of the present disclosure described in the above “Exemplary Method” section of this specification. For example, the processing unit 1010 can execute Figure 3 , Figure 7 or Fig. 9 wait.
[0088] The storage unit 1020 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 1021 and / or a cache memory unit 1022 , and may further include a read-only memory unit (ROM) 1023 .
[0089] The storage unit 1020 may also include a program / utility 1024 having a set (at least one) of program modules 1025, such program modules 1025 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 an implementation of a network environment.
[0090] Bus 1030 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
[0091] The electronic device 1000 may also communicate with one or more external devices 1100 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 1000, and / or may communicate with any device that enables the electronic device 1000 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 1050. Furthermore, the electronic device 1000 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 1060. As shown, the network adapter 1060 communicates with other modules of the electronic device 1000 via a bus 1030. 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 1000, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0092] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation 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, including 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 exemplary embodiment of the present disclosure.
[0093] The exemplary embodiments of the present disclosure also provide a computer-readable storage medium on which a program product capable of implementing the above-mentioned method of the present specification is stored. In some possible implementations, various aspects of the present disclosure may 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 execute the steps according to various exemplary implementations of the present disclosure described in the above-mentioned "Exemplary Method" section of the present specification.
[0094] The exemplary embodiments of the present disclosure also provide a program product for implementing the above method, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto, and in this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.
[0095] The program product may use any combination of one or more readable media. The readable 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 of the above. 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 of the above.
[0096] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, in which 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. Readable signal media 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.
[0097] The program code embodied 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.
[0098] Program code for performing the operations of the present disclosure 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 device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through 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., through the Internet using an Internet service provider).
[0099] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.
[0100] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to exemplary embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.
[0101] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.
[0102] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims
1. A method for detecting an electroencephalogram signal, characterized in that: include: Obtaining the EEG signal to be detected of the target child; The EEG signal to be detected is detected by using a pre-trained child EEG signal detection model to obtain an EEG signal detection result of the target child; The EEG signal detection result includes a first detection result and a second detection result. The first detection result is used to reflect whether the EEG signal to be detected is abnormal, and the second detection result is used to reflect the state of the EEG signal to be detected.
2. The method according to claim 1, characterized in that The children's EEG signal detection model is trained in the following way: Obtaining sample EEG signals of multiple sample children; Obtaining a first type of sample label and a second type of sample label corresponding to the sample EEG signal; A child EEG signal detection model is trained based on the sample EEG signals, the first category sample labels and the second category sample labels.
3. The method according to claim 2, characterized in that: The step of training the child EEG signal detection model according to the sample EEG signal, the first type of sample label and the second type of sample label comprises: Inputting the sample EEG signal into the child EEG signal detection model to be trained to obtain the sample EEG signal detection result; Calculate a loss function based on the sample EEG signal detection result, the first category sample label and the second category sample label; The child's EEG signal detection model to be trained is optimized by minimizing the loss function to obtain a trained child's EEG signal detection model.
4. The method according to claim 3, characterized in that The sample EEG signal detection result includes a first sample detection result and a second sample detection result; and the loss function is calculated based on the sample EEG signal detection result and the first category sample label and the second category sample label, including: Determining a first loss function according to the difference between the first sample detection result and the first type of sample label; Determining a second loss function according to the difference between the second sample detection result and the second type of sample label; Determine a target loss function according to the first loss function and the second loss function; The step of optimizing the child's EEG signal detection model to be trained by minimizing the loss function to obtain a trained child's EEG signal detection model includes: The child's EEG signal detection model to be trained is optimized by minimizing the target loss function to obtain a trained child's EEG signal detection model.
5. The method according to claim 2, characterized in that: The step of obtaining the first type of sample labels and the second type of sample labels corresponding to the sample EEG signals includes: Segmenting the sample EEG signal according to a preset timing rule to obtain sub-signal segments of the sample EEG signal within different time ranges; Obtaining a first label of each sub-signal segment, and determining a first type of sample label of the sample EEG signal according to the first label; A second label of each sub-signal segment is obtained, and a second type of sample label of the sample EEG signal is determined according to the second label.
6. The method according to claim 2, characterized in that The first type of sample label is a classification label of whether the sample EEG signal is an abnormal EEG signal, and the second type of sample label is a level label of the sleep state corresponding to the sample EEG signal.
7. The method according to claim 1, characterized in that After obtaining the EEG signal to be detected of the target child, the method includes: Extracting basic attribute information of the target child and determining the age range of the target child; According to the age range of the target children, a pre-trained children's EEG signal detection model corresponding to the age range is determined.
8. An electroencephalogram signal detection device, characterized in that: include: A signal acquisition module, used to acquire the EEG signal to be detected of the target child; A signal detection module, used to detect the EEG signal to be detected by using a pre-trained child EEG signal detection model to obtain an EEG signal detection result of the target child; The EEG signal detection result includes a first detection result and a second detection result. The first detection result is used to reflect whether the EEG signal to be detected is abnormal, and the second detection result is used to reflect the state of the EEG signal to be detected.
9. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.
Citation Information
Cited By
Brain abnormal discharge detection method and device, readable storage medium and equipment
CN121910389A
Brain abnormal discharge detection method and device, readable storage medium and equipment
CN121910389B
