Psychological stress monitoring method and device based on heartbeat evoked potential

Abstract

The present invention discloses a method and device for monitoring psychological stress based on heartbeat evoked potentials. The method comprises: obtaining multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states; obtaining multiple heartbeat evoked potential samples based on the multi-channel EEG signal sample data and the single-channel ECG signal sample data; performing principal component analysis on the heartbeat evoked potential samples, removing outlier heartbeat evoked potential samples, and constructing a feature vector; performing classification prediction of psychological stress levels based on the constructed feature vectors, and configuring target parameters for a target model based on the prediction results; and performing predictive analysis on target EEG and ECG signals based on the target model to obtain corresponding psychological stress level monitoring results. The present invention is convenient and highly accurate, and can be widely applied in the field of data processing technology.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and device for monitoring psychological stress based on heartbeat evoked potentials. Background Art

[0002] Psychological stress is a psychological stress response to internal and external stimuli, often accompanied by changes in physiological activity. Long-term, high levels of psychological stress are a direct risk factor for psychological disorders such as anxiety and depression, as well as for physiological conditions such as coronary heart disease, endocrine disorders, and immunodeficiency. Therefore, introducing real-time psychological stress monitoring methods into work and living environments, and alerting individuals experiencing high stress levels to take appropriate intervention measures, is a key foundation for maintaining physical and mental health.

[0003] Methods for assessing psychological stress can be broadly divided into three categories: (1) methods based on psychological questionnaires and interviews; (2) biomarkers based on biochemical analysis; and (3) biomarkers based on electrophysiological signals. Among them, psychological stress monitoring based on electrophysiological signal characteristics has the advantages of real-time performance and ease of implementation of intelligent wearable monitoring systems.

[0004] Existing psychological stress monitoring based on electrophysiological biomarkers is mostly based on a single physiological parameter such as EEG or ECG, and the stress monitoring effect is generally around 80%; some use several physiological parameters simultaneously, and the stress monitoring accuracy is improved to greater than 90%; no personalized parameter settings are seen in the monitoring; and there are no reports on psychological stress level monitoring based on heart rate evoked potentials. Summary of the Invention

[0005] In view of this, an embodiment of the present invention provides a highly accurate and convenient psychological stress monitoring method and device based on heartbeat evoked potentials.

[0006] One aspect of an embodiment of the present invention provides a method for monitoring psychological stress based on heartbeat evoked potentials, comprising:

[0007] Obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states;

[0008] Obtaining a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data;

[0009] Performing principal component analysis on the heartbeat evoked potential samples, removing outlier heartbeat evoked potential samples, and constructing a feature vector;

[0010] Performing classification prediction of psychological stress levels based on the constructed characteristic vector, and configuring target parameters of the target model based on the prediction results;

[0011] According to the target model, the target EEG and ECG signals are predicted and analyzed to obtain corresponding psychological stress level monitoring results.

[0012] Optionally, the method further includes: performing signal preprocessing on the ECG signal sample data, which step includes:

[0013] The ECG data are filtered using a zero-phase high-pass filter and a low-pass filter;

[0014] Obtaining filtered ECG data and detecting the R peak of the ECG data;

[0015] The detected R wave peak is corrected to complete the signal preprocessing of the electrocardiogram signal sample data.

[0016] Optionally, acquiring filtered ECG data and detecting an R peak of the ECG data includes:

[0017] Calculate the standard deviation of the ECG data and find all the peaks in the ECG data;

[0018] When the peak amplitude is greater than the set multiple of the standard deviation, the corresponding peak is determined to be an R peak, and the time point of the R peak is recorded;

[0019] The correcting process for the detected R wave peak to complete the signal preprocessing of the electrocardiogram signal sample data includes:

[0020] Calculate the mean of the intervals between all two adjacent R peaks;

[0021] Comparing the intervals of the newly detected R peaks in sequence, if the new interval is greater than the set multiple of the mean, it is determined that the detection of the R peak at the current detection position is missing;

[0022] At the detection missing point, take the midpoint of the two adjacent detected R peak time points, then take the neighborhood of the midpoint, find the maximum peak within the range of the neighborhood, record it as the R peak, and record the time point of the R peak.

[0023] Optionally, the method further includes: performing signal preprocessing on the EEG signal sample data, which step includes:

[0024] The EEG data were filtered using a high-pass filter and a low-pass filter with zero phase;

[0025] Perform artifact subspace reconstruction filtering on EEG data.

[0026] Optionally, obtaining a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data includes:

[0027] According to the time point of the R wave peak of the ECG signal, the EEG signal is segmented to obtain EEG segments;

[0028] All EEG segments are superimposed and averaged to obtain the average heartbeat evoked potential signal;

[0029] All EEG segments are strided and averaged to obtain the heartbeat evoked potential signal samples.

[0030] Optionally, performing classification prediction of psychological stress level based on the constructed characteristic vector, and configuring target parameters of the target model according to the prediction result, includes:

[0031] According to the sorting of the mean differences of the average heartbeat evoked potentials corresponding to different types of psychological stress obtained from multiple channels, the channel with the largest heartbeat evoked potential difference is selected as the optimal channel; wherein the optimal channel is used for feature extraction;

[0032] Based on the optimal channel, the heartbeat evoked potential samples of the same subject under two psychological stresses were statistically analyzed using a paired t-test; based on the paired t-test results, the heartbeat evoked potential amplitudes corresponding to the two time points were selected to construct a feature vector;

[0033] The support vector machine machine learning algorithm is used to classify the psychological stress level, and the two optimal time points for constructing the feature vector are determined based on the classification effect.

[0034] Optionally, the method further includes: performing an accuracy assessment on the psychological stress level monitoring result, which step includes:

[0035] Calculate the TP index, TN index, FP index and FN index of the psychological stress level monitoring results;

[0036] Calculate the sensitivity index, specificity index, and accuracy index of the psychological stress level monitoring results based on the TP index, TN index, FP index, and FN index;

[0037] The accuracy evaluation result of the psychological stress level monitoring result is determined based on the sensitivity index, specificity index, and accuracy index.

[0038] Another aspect of the present invention provides a device for monitoring psychological stress based on heartbeat evoked potentials, comprising:

[0039] The first module is used to obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states;

[0040] The second module is used to obtain a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data;

[0041] The third module is used to perform principal component analysis on the heartbeat evoked potential samples, remove outlier heartbeat evoked potential samples, and construct a feature vector;

[0042] The fourth module is used to classify and predict the psychological stress level based on the constructed feature vector, and configure target parameters of the target model according to the prediction results;

[0043] The fifth module is used to perform predictive analysis on the target EEG and ECG signals according to the target model to obtain corresponding psychological stress level monitoring results.

[0044] Another aspect of an embodiment of the present invention further provides an electronic device, including a processor and a memory;

[0045] The memory is used to store programs;

[0046] The processor executes the program to implement the method described above.

[0047] Another aspect of the embodiments of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the method described above.

[0048] The present invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the above method.

[0049] The embodiments of the present invention obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states; obtain multiple heartbeat evoked potential samples based on the multi-channel EEG signal sample data and the single-channel ECG signal sample data; perform principal component analysis on the heartbeat evoked potential samples, remove outlier heartbeat evoked potential samples, and construct a feature vector; perform psychological stress level classification prediction based on the constructed feature vector, and configure target parameters of the target model based on the prediction results; perform predictive analysis on target EEG and ECG signals based on the target model to obtain corresponding psychological stress level monitoring results. The present invention is convenient and highly accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0051] Figure 1 This is a flowchart of the overall steps of the psychological stress monitoring method based on heartbeat evoked potential;

[0052] Figure 2 This is the R wave peak detection flow chart of the ECG;

[0053] Figure 3 Flowchart for R peak detection and correction. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0055] To address the problems in the prior art, an embodiment of the present invention provides a method for monitoring psychological stress based on heartbeat evoked potentials, comprising:

[0056] Obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states;

[0057] Obtaining a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data;

[0058] Performing principal component analysis on the heartbeat evoked potential samples, removing outlier heartbeat evoked potential samples, and constructing a feature vector;

[0059] Performing classification prediction of psychological stress levels based on the constructed characteristic vector, and configuring target parameters of the target model based on the prediction results;

[0060] According to the target model, the target EEG and ECG signals are predicted and analyzed to obtain corresponding psychological stress level monitoring results.

[0061] Optionally, the method further includes: performing signal preprocessing on the ECG signal sample data, which step includes:

[0062] The ECG data are filtered using a zero-phase high-pass filter and a low-pass filter;

[0063] Obtaining filtered ECG data and detecting the R peak of the ECG data;

[0064] The detected R wave peak is corrected to complete the signal preprocessing of the electrocardiogram signal sample data.

[0065] Optionally, acquiring filtered ECG data and detecting an R peak of the ECG data includes:

[0066] Calculate the standard deviation of the ECG data and find all the peaks in the ECG data;

[0067] When the peak amplitude is greater than the set multiple of the standard deviation, the corresponding peak is determined to be an R peak, and the time point of the R peak is recorded;

[0068] The correcting process for the detected R wave peak to complete the signal preprocessing of the electrocardiogram signal sample data includes:

[0069] Calculate the mean of the intervals between all two adjacent R peaks;

[0070] Comparing the intervals of the newly detected R peaks in sequence, if the new interval is greater than the set multiple of the mean, it is determined that the detection of the R peak at the current detection position is missing;

[0071] At the detection missing point, take the midpoint of the two adjacent detected R peak time points, then take the neighborhood of the midpoint, find the maximum peak within the range of the neighborhood, record it as the R peak, and record the time point of the R peak.

[0072] Optionally, the method further includes: performing signal preprocessing on the EEG signal sample data, which step includes:

[0073] The EEG data were filtered using a high-pass filter and a low-pass filter with zero phase;

[0074] Perform artifact subspace reconstruction filtering on EEG data.

[0075] Optionally, obtaining a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data includes:

[0076] According to the time point of the R wave peak of the ECG signal, the EEG signal is segmented to obtain EEG segments;

[0077] All EEG segments are superimposed and averaged to obtain the average heartbeat evoked potential signal;

[0078] All EEG segments are strided and averaged to obtain the heartbeat evoked potential signal samples.

[0079] Optionally, performing classification prediction of psychological stress level based on the constructed characteristic vector, and configuring target parameters of the target model according to the prediction result, includes:

[0080] According to the sorting of the mean differences of the average heartbeat evoked potentials corresponding to different types of psychological stress obtained from multiple channels, the channel with the largest heartbeat evoked potential difference is selected as the optimal channel; wherein the optimal channel is used for feature extraction;

[0081] Based on the optimal channel, the heartbeat evoked potential samples of the same subject under two psychological stresses were statistically analyzed using a paired t-test; based on the paired t-test results, the heartbeat evoked potential amplitudes corresponding to the two time points were selected to construct a feature vector;

[0082] The support vector machine machine learning algorithm is used to classify the psychological stress level, and the two optimal time points for constructing the feature vector are determined based on the classification effect.

[0083] Optionally, the method further includes: performing an accuracy assessment on the psychological stress level monitoring result, which step includes:

[0084] Calculate the TP index, TN index, FP index and FN index of the psychological stress level monitoring results;

[0085] Calculate the sensitivity index, specificity index, and accuracy index of the psychological stress level monitoring results based on the TP index, TN index, FP index, and FN index;

[0086] The accuracy evaluation result of the psychological stress level monitoring result is determined based on the sensitivity index, specificity index, and accuracy index.

[0087] Another aspect of the present invention provides a device for monitoring psychological stress based on heartbeat evoked potentials, comprising:

[0088] The first module is used to obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states;

[0089] The second module is used to obtain a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data;

[0090] The third module is used to perform principal component analysis on the heartbeat evoked potential samples, remove outlier heartbeat evoked potential samples, and construct a feature vector;

[0091] The fourth module is used to classify and predict the psychological stress level based on the constructed feature vector, and configure target parameters of the target model according to the prediction results;

[0092] The fifth module is used to perform predictive analysis on the target EEG and ECG signals according to the target model to obtain corresponding psychological stress level monitoring results.

[0093] Another aspect of an embodiment of the present invention further provides an electronic device, including a processor and a memory;

[0094] The memory is used to store programs;

[0095] The processor executes the program to implement the method described above.

[0096] Another aspect of the embodiments of the present invention further provides a computer-readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the method described above.

[0097] The present invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the above method.

[0098] The specific implementation process of the present invention is described in detail below with reference to the accompanying drawings:

[0099] First, it should be noted that the present invention proposes a psychological stress monitoring method based on heartbeat evoked potential (HEP) features. The key technologies of this method include electroencephalogram (EEG) and electrocardiogram (ECG) signal processing algorithms, HEP extraction, personalized setting of optimal channels and feature vectors, and a machine learning algorithm for automatically distinguishing between high and low levels of psychological stress. The basic steps of this method include: (1) EEG signal preprocessing, including bandpass filtering and noise removal; (2) ECG signal bandpass filtering, R-wave peak detection and correction; (3) EEG signal segmentation, HEP screening based on principal component analysis and clustering (removing outlier samples of HEP), and average HEP extraction and HEP sample extraction; (4) HEP feature extraction; and (5) automatic classification of psychological stress levels based on HEP ​​features. Compared with the classic psychological stress monitoring methods based on heart rate variability and EEG characteristics, the present invention proposes a heartbeat evoked potential derived from these two signals, and distinguishes the degree of psychological stress based on the characteristics of the heartbeat evoked potential, and verifies that this method has better accuracy in psychological stress monitoring.

[0100] Among them, heartbeat evoked potential (HEP), also known as heartbeat-induced EEG, reflects the brain's response to the heart's physiological activity. HEP extraction involves synchronously recording ECG and EEG signals, performing appropriate EEG and ECG signal processing, detecting the peak of the ECG R wave, and extracting EEG signals within a time window before and after the R wave peak, for example, from -200 to 600 milliseconds. These EEG signals are then appropriately processed, including stacking and averaging, principal component analysis, and clustering, to obtain valid HEP samples.

[0101] This invention aims to extract key features of psychological stress from heartbeat evoked potentials (HEPs) and, through personalized parameter settings, implement a new method for monitoring psychological stress. Although HEPs are only a single physiological parameter, they incorporate information about brain and heart activity, as well as their interactions, enabling more accurate monitoring of psychological stress.

[0102] refer to Figure 1The method of the present invention includes two major steps. (1) Personalized parameter setting: It is necessary to collect multi-channel EEG and single-channel ECG signals under high / low psychological stress, use bandpass filtering to improve the signal-to-noise ratio of ECG signals, identify R peaks based on signal standard deviation information, and extract the time point of R peaks; use bandpass filtering to perform artifact subspace reconstruction on EEG signals. The signal-to-noise ratio is improved by using ASR (Automatic Reconstruction) filtering, and then the average heartbeat evoked potential waveform is obtained by superposition and averaging based on the time point of the R wave peak; the average heartbeat evoked potential waveform corresponding to high / low psychological stress is subtracted to obtain a difference curve, and the channel with the largest difference curve mean is selected as the optimal channel; based on the optimal channel EEG, EEG segments are obtained according to the time point of the R wave peak, and multiple heartbeat evoked potential samples are obtained by strided superposition and averaging. The heartbeat evoked potential samples are subjected to principal component analysis, and the first and second principal components are selected for cluster analysis to remove outlier heartbeat evoked potential samples; the paired t-test method is used for high / low heartbeat evoked potential samples, and the heartbeat evoked potential amplitudes corresponding to two time points with significant statistical differences are selected to construct a feature vector; the support vector machine machine learning algorithm is used to classify and predict the psychological stress level, and the time point corresponding to the feature vector is adjusted according to the prediction effect, and the final time point is determined to complete the personalized parameter setting. (2) Real-time monitoring of current psychological stress: only the EEG and ECG signals of the optimal channel need to be collected. The ECG signal is band-pass filtered to improve the signal-to-noise ratio, the R wave peak is identified based on the signal standard deviation information, and the time point of the R wave peak is extracted; the EEG signal is band-pass filtered to improve the signal-to-noise ratio, and then the average heartbeat evoked potential waveform is obtained by superposition and averaging based on the time point of the R wave peak; the heartbeat evoked potential amplitude corresponding to the time point determined in (1) is extracted to construct a feature vector; and the psychological stress prediction is completed using the machine learning algorithm to give the current psychological stress level.

[0103] It should be noted that, in the implementation of the entire method, for each user, the above step (1) only needs to be performed once, and the real-time psychological stress level monitoring can be performed using step (2).

[0104] 1. EEG and ECG signal recording

[0105] This method needs to be based on the user's multi-channel EEG and single-channel ECG. In order to verify the feasibility of the method, this embodiment obtained data from 10 college students recorded in the prior art. The electrodes are placed in the 10-20 international standard lead position, the ground electrode is the midpoint of the line connecting FPZ and FZ, the reference electrode is the midpoint of the line connecting CZ and CPZ, and the ECG recording electrodes are symmetrically placed on the outside of the clavicles on both sides. The impedance between each electrode and the skin is less than 5kΩ. The recording channels are shown in Table 1. Table 1 is a description of 22 EEG and 1 ECG recording channels.

[0106] Table 1

[0107]

[0108] The data collection experiment in Table 1 above is a mental arithmetic task experiment. The experiment uses two types of questions to provide two types of psychological pressure environments:

[0109] (1) Simple questions: refers to single-digit or double-digit addition, subtraction, multiplication and division operations, used to provide a low-stress environment;

[0110] (2) Complex questions: These are questions with results of two or more digits, including carry or borrow, addition, subtraction, multiplication and division operations, and are used to provide a high psychological pressure environment.

[0111] Examples of the two question types are shown in Table 2

[0112] Table 2

[0113] Simple mental math example Examples of complex mental math problems 7+62－5＝？ 37×22＝？ 11－2＝？ 917－529＝？ 6×2＝？ 123+789＝？ 15÷3＝？ 728÷26＝？ 2－5＝？ 224÷8＝？

[0114] The experiment involved two mental arithmetic tasks. The first consisted of 48 simple problems, each with a time limit of 5 seconds; the second consisted of 24 complex problems, each with a time limit of 10 seconds. Each mental arithmetic task lasted 4 minutes, with a 5-second rest period between each task, during which no mental arithmetic was performed.

[0115] The specific process of the complete experiment for each subject was: resting period (5 seconds) → simple questions (4 minutes) → resting period (5 seconds) → complex questions (4 minutes) → resting period (5 seconds).

[0116] 2. Signal preprocessing

[0117] 2.1 ECG

[0118] 2.1.1 Bandpass Filtering

[0119] The ECG data were filtered using a zero-phase high-pass filter (cutoff frequency: -3 dB, 2 Hz) and a low-pass filter (cutoff frequency: -3 dB, 30 Hz).

[0120] 2.1.2R peak detection

[0121] The R wave peak of the ECG has the largest amplitude in the time domain diagram. The ECG R wave peak detection flow chart is as follows: Figure 2 As shown, the following algorithm steps are used for detection:

[0122] (1) Use the findpeaks function in Matlab to find all peaks;

[0123] (2) Among all the peaks, the peak whose value is greater than n (e.g., n=2.5) times the standard deviation of the ECG signal is recorded and is considered to be the R wave peak.

[0124] 2.1.3R Peak Detection and Correction

[0125] Some subjects with high ECG heart rate variability may have inaccurate R wave peak detection, which can be supplemented by correction algorithms and manual inspection. Figure 3 As shown, the correction algorithm steps are as follows:

[0126] (1) Calculate the mean of all adjacent R peak intervals;

[0127] (2) Compare the intervals of the detected R peaks in sequence. If the intervals are greater than m times the mean, it is considered that there is a lack of R peak detection.

[0128] (3) At the missing part, take the midpoint of the two adjacent detected R wave peak time points, take the neighborhood of the midpoint [-150, 150] ms, find the maximum peak in this range, and record it as the R wave peak.

[0129] 2.2 EEG

[0130] 2.2.1 Bandpass Filtering

[0131] The EEG data were filtered using a zero-phase high-pass filter (cutoff frequency: -3 dB, 1 Hz) and a low-pass filter (cutoff frequency: -3 dB, 25 Hz).

[0132] 2.2.2 Artifact Subspace Reconstruction Filtering

[0133] Artifact subspace reconstruction filtering is an adaptive spatial filter that can detect and remove high-amplitude data components such as eye blinks, muscle and sensor motion. Artifact subspace reconstruction filtering leverages the statistical properties of the analyzed data, using the statistical properties of a relatively clean signal as a benchmark to automatically identify and remove noise, thus avoiding subjective judgment errors.

[0134] 2.2.3 Superposition average

[0135] Average heartbeat evoked potential waveform extraction: According to the time point of the ECG R wave peak, the EEG is segmented to obtain EEG segments, and all EEG segments are superimposed and averaged to obtain the average heartbeat evoked potential signal.

[0136] Extraction of heartbeat evoked potential samples: Based on the time point of the ECG R wave peak, the EEG is segmented to obtain EEG segments; the EEG segments are strided and averaged to obtain multiple samples of the heartbeat evoked potential. Strided and averaged refers to the superposition and averaging of the EEG segments from n to n+100 (assuming the superposition average is 100) to obtain the nth heartbeat evoked potential sample. Next, a certain step size is moved, such as a step size of 10, and the EEG segments from n+10 to n+110 are superimposed and averaged to obtain the n+1th heartbeat evoked potential sample. This cycle is repeated for the EEG segments from simple questions and complex questions, and two groups of heartbeat evoked potential samples with the same number of EEG segments are obtained.

[0137] 2.2.4 Principal Component Analysis and Cluster Analysis

[0138] Principal component analysis was performed on the heartbeat evoked potential samples, and cluster analysis was performed based on the first and second principal components to remove outlier heartbeat evoked potential samples.

[0139] 3. Classification Prediction

[0140] Constructing feature vectors is a key component of machine learning classification algorithms. The key to feature selection is the ability to significantly distinguish between high and low levels of psychological stress. This can be achieved by using statistically significant differences in the heart rate evoked potentials under the two stress levels as features.

[0141] 3.1 Optimal Channel Selection

[0142] The optimal channel is determined in the first major step of this method. The heartbeat evoked potential differences corresponding to high and low psychological stress vary across different EEG channels. Based on this, the following algorithm logic can be employed: Based on the mean heartbeat evoked potential differences across the 22 channels, the channel with the largest heartbeat evoked potential difference is selected as the optimal channel for feature extraction.

[0143] 3.2 Feature vector construction

[0144] Based on the optimal channel, statistical differences in HEP ​​samples from the same subjects under two psychological stresses were analyzed using a paired t-test. Based on the paired t-test results, the HEP amplitudes corresponding to the two time points were selected to construct a feature vector.

[0145] 3.3 Classification Algorithm Prediction

[0146] In the first step of this method (personalized parameter setting), the support vector machine machine learning algorithm is used to classify the psychological stress level, and the two time points for constructing the feature vector are determined based on the classification effect. The two groups of heart rate evoked potential samples are randomly divided into four equal parts, of which two quarters are used for training, one quarter is used for testing, and the other quarter is used for monitoring in the second step of this method. The prediction effect of the algorithm is evaluated using three classic evaluation indicators based on four basic indicators: TP (True Positive), TN (True Negative), FP (False Positive), and FN (False Negative). The explanation of the four basic indicators is shown in Table 3. The calculation formulas for the three evaluation indicators: sensitivity, specificity, and accuracy are shown in formulas (1) to (3), respectively.

[0147]

[0148]

[0149]

[0150] Table 3 Description of basic indicators TP, TN, FP, FN

[0151]

[0152] Adjust the time point corresponding to the feature vector based on the evaluation index to complete the personalized parameter setting.

[0153] Based on the personalized parameters and the last quarter heartbeat evoked potential sample, the current psychological stress monitoring test was completed, and the results are shown in Table 4. The prediction accuracy of 10 subjects was above 97%, and the accuracy of 50% of the subjects was 100%.

[0154] Table 4 Support vector machine results

[0155]

[0156] In summary, the present invention has the following characteristics:

[0157] 1) This method constructs a feature vector based on heartbeat evoked potential to predict the psychological stress level;

[0158] 2) There is a significant statistical difference in the heart rate evoked potentials corresponding to high and low psychological stress, indicating that the heart rate evoked potential has a good ability to distinguish psychological stress;

[0159] 3) Principal component analysis and cluster analysis were used to ensure the quality of heartbeat evoked potential samples;

[0160] 4) Personalized parameter settings can effectively improve monitoring accuracy;

[0161] 5) This method is expected to develop an economical, lightweight, home-usable and reliable real-time psychological stress monitoring system.

[0162] Compared with the prior art, the present invention has the following advantages:

[0163] (1) It was discovered for the first time that heartbeat evoked potentials have a good ability to distinguish psychological stress.

[0164] (2) A variety of signal processing methods are integrated to ensure the quality of HEP samples. Bandpass filtering and artifact subspace reconstruction filtering are used to improve the signal-to-noise ratio of the recorded signal; R-peak detection and correction algorithms are used to ensure the accuracy of R-peak detection; the superposition averaging method is used to extract HEP; and principal component analysis and cluster analysis are used to remove outliers. High-quality HEP samples ensure the effectiveness of machine learning classification features.

[0165] (3) Personalized parameter settings are used for recording channels and feature vectors to eliminate the influence of individual differences.

[0166] Therefore, this method can achieve consistent performance for each user, that is, the monitoring model has good robustness.

[0167] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operation and logic flow presented herein. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.

[0168] Furthermore, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise indicated, one or more of the functions and / or features described may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It will also be understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present invention. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the ordinary skill of an engineer. Therefore, a person skilled in the art using ordinary skill will be able to implement the present invention set forth in the claims without undue experimentation. It will also be understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0169] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0170] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0171] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0172] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0173] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0174] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

[0175] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A psychological stress monitoring method based on heartbeat evoked potential, characterized in that: include: Obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states; Obtaining a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data; Performing principal component analysis on the heartbeat evoked potential samples, removing outlier heartbeat evoked potential samples, and constructing a feature vector; Performing classification prediction of psychological stress levels based on the constructed characteristic vector, and configuring target parameters of the target model based on the prediction results; According to the target model, predictive analysis is performed on the target EEG and ECG signals to obtain corresponding psychological stress level monitoring results; The obtaining of a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data comprises: According to the time point of the R wave peak of the ECG signal, the EEG signal is segmented to obtain EEG segments; All EEG segments are superimposed and averaged to obtain the average heartbeat evoked potential signal; Perform stride superposition and averaging on all EEG segments to obtain the heartbeat evoked potential signal samples; The method of performing classification prediction of psychological stress level based on the constructed characteristic vector and configuring target parameters of the target model based on the prediction result includes: According to the sorting of the mean differences of the average heartbeat evoked potentials corresponding to different types of psychological stress obtained from multiple channels, the channel with the largest heartbeat evoked potential difference is selected as the optimal channel; wherein the optimal channel is used for feature extraction; Based on the optimal channel, the heartbeat evoked potential samples of the same subject under two psychological stresses were statistically analyzed using a paired t-test; based on the paired t-test results, the heartbeat evoked potential amplitudes corresponding to the two time points were selected to construct a feature vector; The support vector machine machine learning algorithm is used to classify the psychological stress level, and the two optimal time points for constructing the feature vector are determined based on the classification effect.

2. The method for monitoring psychological stress based on heartbeat evoked potential according to claim 1, characterized in that: The method further includes: performing signal preprocessing on the ECG signal sample data, including: The ECG data are filtered using a zero-phase high-pass filter and a low-pass filter; Obtaining filtered ECG data and detecting the R peak of the ECG data; The detected R wave peak is corrected to complete the signal preprocessing of the electrocardiogram signal sample data.

3. The method for monitoring psychological stress based on heartbeat evoked potential according to claim 2, characterized in that: The step of acquiring filtered ECG data and detecting an R peak of the ECG data includes: Calculate the standard deviation of the ECG data and find all the peaks in the ECG data; When the peak amplitude is greater than the set multiple of the standard deviation, the corresponding peak is determined to be an R peak, and the time point of the R peak is recorded; The correcting process for the detected R wave peak to complete the signal preprocessing of the electrocardiogram signal sample data includes: Calculate the mean of the intervals between all two adjacent R peaks; Comparing the intervals of the newly detected R peaks in sequence, if the new interval is greater than the set multiple of the mean, it is determined that the detection of the R peak at the current detection position is missing; At the detection missing point, take the midpoint of the two adjacent detected R peak time points, then take the neighborhood of the midpoint, find the maximum peak within the range of the neighborhood, record it as the R peak, and record the time point of the R peak.

4. The method for monitoring psychological stress based on heartbeat evoked potential according to claim 1, characterized in that: The method further includes: performing signal preprocessing on the EEG signal sample data, including: The EEG data were filtered using a high-pass filter and a low-pass filter with zero phase; Perform artifact subspace reconstruction filtering on EEG data.

5. The method for monitoring psychological stress based on heartbeat evoked potential according to any one of claims 1 to 4, characterized in that: The method further includes: performing an accuracy assessment on the psychological stress level monitoring result, including: Calculate the TP index, TN index, FP index and FN index of the psychological stress level monitoring results; Calculate the sensitivity index, specificity index, and accuracy index of the psychological stress level monitoring results based on the TP index, TN index, FP index, and FN index; The accuracy evaluation result of the psychological stress level monitoring result is determined based on the sensitivity index, specificity index, and accuracy index.

6. A device for implementing the psychological stress monitoring method based on heartbeat evoked potential according to any one of claims 1 to 5, characterized in that: include: The first module is used to obtain multi-channel EEG signal sample data and single-channel ECG signal sample data under different psychological stress states; The second module is used to obtain a plurality of heartbeat evoked potential samples according to the multi-channel EEG signal sample data and the single-channel ECG signal sample data; The third module is used to perform principal component analysis on the heartbeat evoked potential samples, remove outlier heartbeat evoked potential samples, and construct a feature vector; The fourth module is used to classify and predict the psychological stress level based on the constructed feature vector, and configure target parameters of the target model according to the prediction results; The fifth module is used to perform predictive analysis on the target EEG and ECG signals according to the target model to obtain corresponding psychological stress level monitoring results.

7. An electronic device, characterized in that: including a processor and a memory; The memory is used to store programs; The processor executes the program to implement the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The storage medium stores a program, and the program is executed by a processor to implement the method according to any one of claims 1 to 5.

Images