Cognitive disorder monitoring system and method based on brain wave analysis
By creating a brain wave-based cognitive impairment monitoring model, the problem that the existing technology cannot monitor and predict cognitive impairment in real time is solved, real-time prediction and monitoring of patients' cognitive status is achieved, and safety is ensured.
Patent Information
- Application Number
- CN202410977836.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-01
AI Technical Summary
The existing technology fails to effectively utilize the relationship between brain waves and cognitive impairment, and cannot monitor and predict the cognitive status of patients such as Alzheimer's disease in real time, resulting in difficulty in monitoring and medical intervention.
Create a brain wave-based cognitive impairment monitoring model, and divide it into training sets, test sets and verification sets by collecting and preprocessing brain wave data, training and verification sets, and collecting brain wave data in real time for feature extraction and output Monitoring data is transmitted to the guardian terminal through the wireless transmission module.
Real-time prediction of the cognitive status of some patients is achieved, and guardians can intervene in time or conduct medical intervention to prevent safety problems.
Smart Images

Figure CN118873097A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of computer fault diagnosis, and in particular relates to a cognitive impairment monitoring system and method based on brain wave analysis. Background Art
[0002] Brain waves come from the electrical activity of brain nerve tissue. When nerve cells transmit signals, brain waves are generated. Brain wave activity has certain regular characteristics and has a certain degree of correspondence with the brain's consciousness or cognition.
[0003] The study found that the phenomenon of moving waves is common in the human cerebral cortex and increases coordination when the brain processes tasks. This means that brain waves are closely related to behavior, memory, perception, attention and even consciousness, and are an important but often overlooked brain mechanism.
[0004] At present, many scholars and neurologists have conducted research on human brain waves. According to the research results, it is indeed found that there are different brain waves under different cognitive conditions, different behaviors, and different memory states. Brain waves are closely related to cognition, behavior, memory and even consciousness.
[0005] However, no one has yet established the relationship between brain waves and cognition. As a result, although there have been many studies on brain waves, they have never been used to solve practical problems in real life. For example, after someone suffers from Alzheimer's disease, if the brain waves of the patient can be analyzed, their cognitive state can be monitored in real time, and it can be predicted whether they have cognitive impairment. The guardian will be able to understand the patient's condition in real time to achieve safe and effective monitoring of the patient.
[0006] Therefore, the present invention hopes to achieve real-time prediction of the cognitive status of certain patients by creating a cognitive impairment monitoring model, using brain waves as the input of the cognitive impairment monitoring model, and using cognitive prediction results as the output of the cognitive impairment prediction model. Guardians can intervene or perform medical intervention in a timely manner to prevent safety or other problems. Summary of the invention
[0007] The purpose of the present invention is to provide a real-time prediction of the cognitive state of certain patients by creating a cognitive impairment monitoring model, using brain waves as the input of the cognitive impairment monitoring model, and using cognitive prediction results as the output of the cognitive impairment prediction model. Guardians can intervene or perform medical intervention in a timely manner to prevent safety problems or other problems.
[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:
[0009] In a first aspect, a method for monitoring cognitive impairment based on brain wave analysis is provided, comprising the following steps:
[0010] S1: Collect brain wave data of a specified period of cognitive impairment population and obtain cognitive impairment data within the corresponding period;
[0011] S2: extracting brain wave data in a time period before cognitive impairment occurs within a specified period, and preprocessing the extracted brain wave data;
[0012] S3: Divide the preprocessed EEG data into training set, test set and validation set;
[0013] S4: creating a cognitive impairment prediction model, and inputting the training set, the test set, and the validation set into the cognitive impairment prediction model in sequence, respectively, to train, test, and validate the cognitive impairment prediction model;
[0014] S5: collecting brain wave data of the monitored person in real time, and inputting the real-time collected brain wave data into the cognitive impairment prediction model;
[0015] S6: The cognitive impairment prediction model extracts features from the real-time collected brain wave data and outputs cognitive impairment monitoring data;
[0016] S7: The cognitive impairment monitoring data is transmitted to the guardian's terminal device via a wireless transmission module.
[0017] Preferably, the brain wave data in step S1 includes brain wave frequency and brain wave waveform; and the cognitive impairment data acquired within a corresponding period of time includes language data and behavior data.
[0018] Preferably, the specific process of preprocessing the extracted brain wave data in step S2 is as follows:
[0019] S21: deleting invalid brain wave data and duplicate brain wave data in the extracted brain wave data;
[0020] S22: Processing abnormal values in the brain wave data by deleting invalid brain wave data and duplicate brain wave data, including interpolation or replacement of missing values.
[0021] Preferably, the specific process of training, testing and validating the cognitive impairment prediction model in step S4 is as follows:
[0022] S41: inputting the training set into the cognitive impairment prediction model to train it;
[0023] S42: inputting the test set into the trained cognitive impairment prediction model to test it, and determining whether the test result meets the preset test requirements, if so, executing step S43, if not, re-obtaining a new training set to train the cognitive impairment prediction model, and then testing the model until the test result meets the preset test requirements;
[0024] S43: Input the verification set into the cognitive impairment prediction model after the test to verify its test results, and determine whether the verification results meet the verification requirements. If so, execute step S5; if not, re-execute steps S41 to S43 until the verification results meet the verification requirements.
[0025] Preferably, the cognitive impairment prediction model has at least three layers, including an input layer, a hidden layer and an output layer. A nonlinear spatial transformation of the EEG data is performed between the input layer and the hidden layer, specifically, the nonlinear spatial transformation of the EEG is achieved through the Gaussian function of the radial basis function, and a linear spatial transformation is performed between the hidden layer and the output layer. The specific method is to train the EEG data for a specified number of times through a clustering algorithm or the least orthogonal squares method to obtain a cluster center matrix and a weight matrix.
[0026] Preferably, the specific training process in step S41 is:
[0027] S411: Initializing the cognitive impairment model and using the training set as a clustering center;
[0028] S412: grouping the input training set according to the nearest neighbor rule, and assigning the data in the training set to different cluster sets according to the Euclidean distance between the data in the training set and the cluster center;
[0029] S413: Update cluster centers: calculate the average value of the data of the training set in each cluster set, and determine whether the new cluster center is the same as the cluster center before the update. If so, use the new cluster center as the final basis function center of the cognitive impairment prediction model; if not, re-execute step and S412 to perform the next round of cluster center solution;
[0030] S414: solving the variance based on the Gaussian function of the radial basis function;
[0031] S415: Calculating the weight between the hidden layer and the output layer of the cognitive impairment prediction model by a clustering algorithm or a least orthogonal squares method;
[0032] S416: Outputting prediction results through the cognitive impairment prediction model.
[0033] In a second aspect, a cognitive impairment monitoring system based on brain wave analysis is provided to implement any one of the cognitive impairment monitoring methods based on brain wave analysis, including a data acquisition module, a brain wave data extraction module, a data preprocessing module, a model creation module, a cognitive impairment prediction model and a wireless transmission module; the signal output end of the brain wave acquisition module is connected to the signal input end of the brain wave data extraction module, the signal output end of the brain wave data extraction module is connected to the signal input end of the data preprocessing module, the signal output end of the data preprocessing module is connected to the output layer of the cognitive impairment prediction model, the output layer of the cognitive impairment prediction model is connected to the signal input end of the wireless transmission module, and the signal output end of the wireless transmission module is connected to the signal input end of a terminal device;
[0034] The data collection module is used to collect brain wave data of a specified group of people within a specified period and obtain cognitive impairment data within the corresponding period;
[0035] The brain wave data extraction module is used to extract brain wave data in a time period before the onset of cognitive impairment within a specified period;
[0036] The data preprocessing module is used to preprocess the extracted brain wave data;
[0037] The model creation module is used to create a cognitive impairment prediction model;
[0038] The cognitive impairment prediction model is used to predict cognitive impairment based on input brain wave data;
[0039] The wireless transmission module is used to transmit the cognitive impairment prediction result of the cognitive impairment prediction model to the guardian's terminal device.
[0040] The beneficial effects of the present invention include:
[0041] The cognitive impairment monitoring system and method based on brain wave analysis provided by the present invention collect brain wave data of people with cognitive impairment; extract brain wave data in a time period before the occurrence of cognitive impairment within a specified period, and pre-process the extracted brain wave data; divide the pre-processed brain wave data into a training set, a test set and a validation set, and input them into a created cognitive impairment prediction model in sequence; collect brain wave data of the monitored person in real time, input the real-time collected brain wave data into the cognitive impairment prediction model, extract features from the real-time collected brain wave data, and output cognitive impairment monitoring data; the cognitive impairment monitoring data is transmitted to the guardian's terminal device through a wireless transmission module.
[0042] By creating a cognitive impairment monitoring model, using brain waves as the input of the cognitive impairment monitoring model and cognitive prediction results as the output of the cognitive impairment prediction model, real-time prediction of the cognitive status of certain patients can be achieved. Guardians can intervene or conduct medical intervention in a timely manner to prevent safety or other problems. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 It is a flow chart of the cognitive impairment monitoring method based on brain wave analysis of the present invention. DETAILED DESCRIPTION
[0044] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.
[0045] The following is combined with Figure 1 The present invention is further described in detail:
[0046] The method for monitoring cognitive impairment based on brain wave analysis includes the following steps:
[0047] S1: Collect brain wave data of a specified period of cognitive impairment population and obtain cognitive impairment data within the corresponding period;
[0048] S2: extracting brain wave data in a time period before cognitive impairment occurs within a specified period, and preprocessing the extracted brain wave data;
[0049] S3: Divide the preprocessed EEG data into training set, test set and validation set;
[0050] S4: creating a cognitive impairment prediction model, and inputting the training set, the test set, and the validation set into the cognitive impairment prediction model in sequence, respectively, to train, test, and validate the cognitive impairment prediction model;
[0051] S5: collecting brain wave data of the monitored person in real time, and inputting the real-time collected brain wave data into the cognitive impairment prediction model;
[0052] S6: The cognitive impairment prediction model extracts features from the real-time collected brain wave data and outputs cognitive impairment monitoring data;
[0053] S7: The cognitive impairment monitoring data is transmitted to the guardian's terminal device via a wireless transmission module.
[0054] The cognitive impairment monitoring method based on brain wave analysis first collects brain wave data within a specified period of the cognitive impairment population, and simultaneously obtains cognitive impairment data within a corresponding time period, wherein the brain wave data includes brain wave frequency and brain wave waveform; the cognitive impairment data obtained within the corresponding time period includes language data and behavioral data. The data collection here is to provide a data basis for the subsequent cognitive impairment prediction model. Through the brain wave data and cognitive impairment data here, the relationship between the waveform or frequency in the brain wave data and cognitive impairment is learned through the cognitive impairment model, and then the brain wave data of the person before the cognitive impairment occurs can be identified. This enables the cognitive impairment prediction model to have the ability to predict cognitive impairment based on real-time brain wave data. The brain wave data within the specified period includes both the brain wave data before the occurrence of cognitive impairment and the brain wave data when cognitive impairment occurs.
[0055] When obtaining the corresponding EEG data and cognitive impairment data, data preprocessing is required before it can be used for training, testing and verification of cognitive impairment prediction models. Because there is a high probability that there are some invalid data, duplicate data or abnormal data in the collected EEG data and cognitive impairment data, in order to avoid invalid data, duplicate data or abnormal data affecting the subsequent training, testing and verification process of cognitive impairment prediction models. Therefore, it is necessary to preprocess these invalid data, duplicate data or abnormal data.
[0056] The specific process of preprocessing the extracted brain wave data in step S2 is as follows:
[0057] S21: deleting invalid brain wave data and duplicate brain wave data in the extracted brain wave data;
[0058] S22: Processing abnormal values in the brain wave data by deleting invalid brain wave data and duplicate brain wave data, including interpolation or replacement of missing values.
[0059] After preprocessing the EEG data and cognitive impairment data, the preprocessed data needs to be divided into training sets, test sets, and validation sets to achieve the training, testing, and validation of the cognitive impairment prediction model. The purpose of training the cognitive impairment prediction model is to enable the model to identify the EEG data before cognitive impairment occurs. The purpose of testing the model is to test whether the model has the ability to identify the EEG data before cognitive impairment occurs, that is, to achieve the ability to predict cognitive impairment before it occurs. The purpose of validating the model is to verify whether the aforementioned test results are accurate, and to further ensure that the cognitive impairment prediction model has the ability to predict cognitive impairment in advance based on EEG data.
[0060] The specific process of training, testing and validating the cognitive impairment prediction model in step S4 is as follows:
[0061] S41: inputting the training set into the cognitive impairment prediction model to train it;
[0062] S42: inputting the test set into the trained cognitive impairment prediction model to test it, and determining whether the test result meets the preset test requirements, if so, executing step S43, if not, re-obtaining a new training set to train the cognitive impairment prediction model, and then testing the model until the test result meets the preset test requirements;
[0063] S43: Input the verification set into the cognitive impairment prediction model after the test to verify its test results, and determine whether the verification results meet the verification requirements. If so, execute step S5; if not, re-execute steps S41 to S43 until the verification results meet the verification requirements.
[0064] The cognitive impairment prediction model has at least three layers, including an input layer, a hidden layer and an output layer. A nonlinear spatial transformation of brain wave data is performed between the input layer and the hidden layer, specifically, the nonlinear spatial transformation of brain waves is realized by the Gaussian function of the radial basis function. A linear spatial transformation is performed between the hidden layer and the output layer. The specific method is to train the brain wave data for a specified number of times through a clustering algorithm or a least orthogonal squares method to obtain a cluster center matrix and a weight matrix.
[0065] The specific training process in step S41 is:
[0066] S411: Initializing the cognitive impairment model and using the training set as a clustering center;
[0067] S412: grouping the input training set according to the nearest neighbor rule, and assigning the data in the training set to different cluster sets according to the Euclidean distance between the data in the training set and the cluster center;
[0068] S413: Update cluster centers: calculate the average value of the data of the training set in each cluster set, and determine whether the new cluster center is the same as the cluster center before the update. If so, use the new cluster center as the final basis function center of the cognitive impairment prediction model; if not, re-execute step and S412 to perform the next round of cluster center solution;
[0069] S414: solving the variance based on the Gaussian function of the radial basis function;
[0070] S415: Calculating the weight between the hidden layer and the output layer of the cognitive impairment prediction model by a clustering algorithm or a least orthogonal squares method;
[0071] S416: Outputting prediction results through the cognitive impairment prediction model.
[0072] For this embodiment, since a large amount of brain wave data and cognitive impairment data of people with cognitive impairment are collected, the cognitive impairment monitoring method and cognitive impairment prediction model can be applied to other people with cognitive impairment or people with potential cognitive impairment. In other words, it can also be used in people with a high incidence of cognitive impairment, such as people with family-inherited cognitive impairment or some elderly people, to predict cognitive impairment in advance, so that corresponding medical intervention can be carried out earlier or the guardian can understand the cognitive status of the monitored person in real time in advance, and make preparations in advance before the monitored person is found to have cognitive impairment, so as to avoid safety problems or getting lost when cognitive impairment occurs.
[0073] In another embodiment, the brain wave data collected within a specified period and the cognitive impairment data within the corresponding time period can be obtained for the monitored person, including the brain wave data of the monitored person during the normal cognitive time and the brain wave data when cognitive impairment occurs. Because in general, among the people who are preparing for cognitive impairment, most people do not have cognitive impairment all the time, but will have cognitive impairment in some time periods, so it is possible to collect enough brain wave data of the monitored person in the normal cognitive time period and the brain wave data of the cognitive impairment time period, as the data basis of the cognitive impairment prediction model of the monitored person, to realize the training, testing and verification of the cognitive impairment prediction model of the monitored person. In this embodiment, it is more suitable for people who have already had cognitive impairment, so it is convenient to collect brain wave data and cognitive impairment data in their corresponding time period.
[0074] A cognitive impairment monitoring system based on brain wave analysis, which implements any one of the cognitive impairment monitoring methods based on brain wave analysis, comprises a data acquisition module, a brain wave data extraction module, a data preprocessing module, a model creation module, a cognitive impairment prediction model and a wireless transmission module; the signal output end of the brain wave acquisition module is connected to the signal input end of the brain wave data extraction module, the signal output end of the brain wave data extraction module is connected to the signal input end of the data preprocessing module, the signal output end of the data preprocessing module is connected to the output layer of the cognitive impairment prediction model, the output layer of the cognitive impairment prediction model is connected to the signal input end of the wireless transmission module, and the signal output end of the wireless transmission module is connected to the signal input end of a terminal device;
[0075] The data collection module is used to collect brain wave data of a specified group of people within a specified period and obtain cognitive impairment data within the corresponding period;
[0076] The brain wave data extraction module is used to extract brain wave data in a time period before the onset of cognitive impairment within a specified period;
[0077] The data preprocessing module is used to preprocess the extracted brain wave data;
[0078] The model creation module is used to create a cognitive impairment prediction model;
[0079] The cognitive impairment prediction model is used to predict cognitive impairment based on input brain wave data;
[0080] The wireless transmission module is used to transmit the cognitive impairment prediction result of the cognitive impairment prediction model to the guardian's terminal device.
[0081] In summary, the cognitive impairment monitoring system and method based on brain wave analysis provided by the present invention first collects brain wave data of people with cognitive impairment; then extracts brain wave data in a time period before the occurrence of cognitive impairment within a specified period, and preprocesses the extracted brain wave data; divides the preprocessed brain wave data into a training set, a test set, and a validation set, and sequentially inputs them into the created cognitive impairment prediction model; then collects brain wave data of the monitored person in real time, and inputs the real-time collected brain wave data into the cognitive impairment prediction model, extracts features from the real-time collected brain wave data, and outputs cognitive impairment monitoring data; finally, the cognitive impairment monitoring data is transmitted to the guardian's terminal device through a wireless transmission module.
[0082] By creating a cognitive impairment monitoring model, using brain waves as the input of the cognitive impairment monitoring model and cognitive prediction results as the output of the cognitive impairment prediction model, real-time prediction of the cognitive status of certain patients can be achieved. Guardians can intervene or conduct medical intervention in a timely manner to prevent safety or other problems.
[0083] The above-mentioned embodiments only express the specific implementation method of the present application, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the protection scope of the present application. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the technical solution of the present application, and these all belong to the protection scope of the present application.
Claims
1. Cognitive impairment monitoring method based on brain wave analysis, It is characterized in that The following steps are involved: S1: Collect brain wave data of a specified period of cognitive impairment population and obtain cognitive impairment data within the corresponding period; S2: extracting brain wave data in a time period before cognitive impairment occurs within a specified period, and preprocessing the extracted brain wave data; S3: Divide the preprocessed EEG data into training set, test set and validation set; S4: creating a cognitive impairment prediction model, and inputting the training set, the test set, and the validation set into the cognitive impairment prediction model in sequence, respectively, to train, test, and validate the cognitive impairment prediction model; S5: collecting brain wave data of the monitored person in real time, and inputting the real-time collected brain wave data into the cognitive impairment prediction model; S6: The cognitive impairment prediction model extracts features from the real-time collected brain wave data and outputs cognitive impairment monitoring data; S7: The cognitive impairment monitoring data is transmitted to the guardian's terminal device via a wireless transmission module.
2. The method for monitoring cognitive impairment based on brain wave analysis according to claim 1, It is characterized in that The brain wave data in step S1 includes brain wave frequency and brain wave waveform; the cognitive impairment data acquired within the corresponding time includes language data and behavior data.
3. The method for monitoring cognitive impairment based on brain wave analysis according to claim 1, It is characterized in that The specific process of preprocessing the extracted brain wave data in step S2 is as follows: S21: deleting invalid brain wave data and duplicate brain wave data in the extracted brain wave data; S22: Processing abnormal values in the brain wave data by deleting invalid brain wave data and duplicate brain wave data, including interpolation or replacement of missing values.
4. The method for monitoring cognitive impairment based on brain wave analysis according to claim 1, It is characterized in that The specific process of training, testing and validating the cognitive impairment prediction model in step S4 is as follows: S41: inputting the training set into the cognitive impairment prediction model to train it; S42: inputting the test set into the trained cognitive impairment prediction model to test it, and determining whether the test result meets the preset test requirements, if so, executing step S43, if not, re-obtaining a new training set to train the cognitive impairment prediction model, and then testing the model until the test result meets the preset test requirements; S43: Input the verification set into the cognitive impairment prediction model after the test to verify its test results, and determine whether the verification results meet the verification requirements. If so, execute step S5; if not, re-execute steps S41 to S43 until the verification results meet the verification requirements.
5. The method for monitoring cognitive impairment based on brain wave analysis according to claim 4, It is characterized in that The cognitive impairment prediction model has at least three layers, including an input layer, a hidden layer and an output layer. A nonlinear spatial transformation of brain wave data is performed between the input layer and the hidden layer, specifically, the nonlinear spatial transformation of brain waves is realized by the Gaussian function of the radial basis function. A linear spatial transformation is performed between the hidden layer and the output layer. The specific method is to train the brain wave data for a specified number of times through a clustering algorithm or a least orthogonal squares method to obtain a cluster center matrix and a weight matrix.
6. The method for monitoring cognitive impairment based on brain wave analysis according to claim 4, It is characterized in that The specific training process in step S41 is: S411: Initializing the cognitive impairment model and using the training set as a clustering center; S412: grouping the input training set according to the nearest neighbor rule, and assigning the data in the training set to different cluster sets according to the Euclidean distance between the data in the training set and the cluster center; S413: Update cluster centers: calculate the average value of the data of the training set in each cluster set, and determine whether the new cluster center is the same as the cluster center before the update. If so, use the new cluster center as the final basis function center of the cognitive impairment prediction model; if not, re-execute step and S412 to perform the next round of cluster center solution; S414: solving the variance based on the Gaussian function of the radial basis function; S415: Calculating the weight between the hidden layer and the output layer of the cognitive impairment prediction model by a clustering algorithm or a least orthogonal squares method; S416: Outputting prediction results through the cognitive impairment prediction model.
7. A cognitive impairment monitoring system based on brain wave analysis, which implements the cognitive impairment monitoring method based on brain wave analysis as described in any one of claims 1 to 5, It is characterized in that It includes a data acquisition module, a brain wave data extraction module, a data preprocessing module, a model creation module, a cognitive impairment prediction model and a wireless transmission module; the signal output end of the brain wave acquisition module is connected to the signal input end of the brain wave data extraction module, the signal output end of the brain wave data extraction module is connected to the signal input end of the data preprocessing module, the signal output end of the data preprocessing module is connected to the output layer of the cognitive impairment prediction model, the output layer of the cognitive impairment prediction model is connected to the signal input end of the wireless transmission module, and the signal output end of the wireless transmission module is connected to the signal input end of the terminal device; The data collection module is used to collect brain wave data of a specified group of people within a specified period and obtain cognitive impairment data within the corresponding period; The brain wave data extraction module is used to extract brain wave data in a time period before the onset of cognitive impairment within a specified period; The data preprocessing module is used to preprocess the extracted brain wave data; The model creation module is used to create a cognitive impairment prediction model; The cognitive impairment prediction model is used to predict cognitive impairment based on input brain wave data; The wireless transmission module is used to transmit the cognitive impairment prediction result of the cognitive impairment prediction model to the guardian's terminal device.
Citation Information
Cited By
Cognitive impairment electroencephalogram signal analysis system and method based on multi-dimensional feature fusion
CN122701360A
