CN118177835A - Fatigue degree prediction method and system based on electroencephalogram signals - Google Patents
Fatigue degree prediction method and system based on electroencephalogram signals Download PDFInfo
- Publication number
- CN118177835A CN118177835A CN202410351984.1A CN202410351984A CN118177835A CN 118177835 A CN118177835 A CN 118177835A CN 202410351984 A CN202410351984 A CN 202410351984A CN 118177835 A CN118177835 A CN 118177835A
- Authority
- CN
- China
- Prior art keywords
- eeg
- electroencephalogram
- historical
- fatigue
- testees
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 238000010183 spectrum analysis Methods 0.000 claims description 15
- 238000007781 pre-processing Methods 0.000 claims description 14
- 238000005070 sampling Methods 0.000 claims description 10
- 238000012360 testing method Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 210000004556 brain Anatomy 0.000 description 18
- 238000012549 training Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a fatigue degree prediction method and a system based on an electroencephalogram signal, which relate to the technical field of fatigue detection. The method has higher time resolution, reduces the dimension of EEG data and the interference of information irrelevant to the fatigue degree on a model by extracting the electrode channel most relevant to the fatigue degree, and ensures that the EEG data and the fatigue degree label are more easy to correspond. The continuous EEG signal in the time domain is converted into a sample time window of 1s by the data slicing operation, and the model is built according to the sample time window of 1s during modeling, so that fatigue degree prediction is performed with the time resolution of 1s, and the high efficiency and the high prediction accuracy are ensured.
Description
Technical Field
The invention relates to the technical field of fatigue detection, in particular to a fatigue degree prediction method and system based on an electroencephalogram signal.
Background
The effectiveness of EEG-based fatigue monitoring techniques is currently high enough, but there are many problems in practical use. The traditional EEG fatigue monitoring steps are as follows: EEG (brain wave) data acquisition of individuals under different fatigue degrees, EEG (brain wave) data preprocessing, noise reduction, slicing of EEG (brain wave) signals which are continuous in time into EEG (brain wave) samples with fixed time length, input of a subsequent model, selection of a specific algorithm to build a model, and conversion from EEG (brain wave) signals with fixed time length to output of different fatigue degrees of the individuals are achieved, so that fatigue degree prediction based on EEG (brain wave) is achieved.
But has the following technical problems:
1. since the individual differences of EEG (brain waves) are relatively large, a new individual needs to perform a long, complicated and tedious calibration task before performing an EEG fatigue level prediction task for a specific new individual: that is, after the EEG (brain wave) data of the new individual is acquired under different fatigue degrees of the known new individual, a higher model prediction accuracy can be ensured, which leads to low acceptability of the new individual.
2. The real-time performance of predicting the fatigue degree based on EEG (brain wave) is poor, and due to the characteristics of low signal-to-noise ratio, multiple dimensions and the like of EEG (brain wave) signals collected from the multi-channel brain wave cap, higher accuracy is ensured, and the time length of EEG (brain wave) samples adopted in model establishment is long enough, so that enough information can be collected, which affects the real-time performance of model prediction, and the situation that the fatigue degree needs to be predicted in real time is not satisfied.
Therefore, how to improve the universality and the real-time performance of the fatigue model based on brain waves is a problem which needs to be solved by the person skilled in the art.
Disclosure of Invention
In view of the above, the present invention provides a method and a system for predicting the fatigue degree in real time based on the electroencephalogram signals, so as to solve the problems existing in the background technology.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
in one aspect, a fatigue degree prediction method based on an electroencephalogram signal is provided, including:
collecting EEG original data and labels of a plurality of historical testees;
Performing spectrum analysis on the EEG original data of the plurality of historical testees, and selecting an electroencephalogram acquisition channel to obtain respective electroencephalogram channel sets of the historical testees;
acquiring EEG original data of a to-be-detected tested person, performing frequency spectrum analysis and selecting an electroencephalogram acquisition channel to obtain a target electroencephalogram channel set;
Selecting a historical testee which is the same as the target electroencephalogram channel set of the testee to be detected from the historical testees;
modeling is carried out based on EEG original data and labels of the selected historical testees, so that a fatigue prediction model is obtained;
And predicting the to-be-detected testee by using the fatigue prediction model to obtain the fatigue state of the to-be-detected testee.
Optionally, the EEG raw data and corresponding labels of the plurality of historical subjects are obtained by a calibration procedure comprising the steps of: firstly, an individual needs to execute a calibration program under different fatigue degrees to complete a signal acquisition task, when the signal is acquired, the individual needs to wear a multichannel electroencephalogram cap to acquire EEG data and complete PVT objective reflecting capability test, and the fatigue degree of EEG signals in a corresponding time period is calibrated through response time obtained by PVT.
Optionally, the to-be-detected testee wears the multi-channel electroencephalogram cap to acquire EEG data, and the obtained original three-dimensional EEG data is a time sampling point multiplied by the number of electrode channels multiplied by an EEG voltage value.
Optionally, spectrum analysis is performed on the acquired EEG original data to obtain power values of alpha and beta frequency bands, all the electroencephalogram electrode channels of each tested person are ordered according to the value of alpha/beta, the largest four electrode channels are selected, and the final result is that each tested person obtains a self electroencephalogram channel set, wherein the number of all the electroencephalogram electrode channels is greater than four.
Optionally, before modeling based on the EEG raw data and the label of the selected historical subject, the method further comprises preprocessing the EEG raw data of the selected historical subject, wherein the specific preprocessing steps are as follows: filtering, de-artifacting algorithm and slicing the selected EEG original data of the historical testee into EEG samples with the time length of 1s, wherein the format is as follows: 1s time sampling point x electrode channel number x EEG voltage value; after preprocessing, a machine learning algorithm is selected to build a fatigue prediction model from EEG samples of time length 1s and corresponding tags.
In another aspect, there is provided an electroencephalogram signal-based fatigue degree prediction system, including:
The system comprises an acquisition module, a detection module and a detection module, wherein the acquisition module acquires EEG (electronic test) original data and labels of a plurality of historical testees and acquires EEG original data of testees to be detected;
the frequency spectrum analysis module is used for respectively carrying out frequency spectrum analysis on the EEG original data of the plurality of historical testees and the EEG original data of the testees to be detected, and selecting an electroencephalogram acquisition channel to obtain respective electroencephalogram channel sets and target electroencephalogram channel sets of the historical testees;
the screening module is used for selecting a historical tested person which is the same as the target electroencephalogram channel set of the tested person to be detected from the historical tested person;
The fatigue prediction model building module is used for modeling based on the EEG original data and the label of the selected historical testee to obtain a fatigue prediction model;
And the result output module predicts the tested person to be detected by using the fatigue prediction model to obtain the fatigue state of the tested person to be detected.
Optionally, the system further comprises a preprocessing module, wherein the preprocessing module is used for filtering, removing artifact algorithm and slicing the selected EEG original data of the historical testee into EEG samples with the time length of 1s, and the format is as follows: 1s time sampling points x electrode channel number x EEG voltage value.
Compared with the prior art, the fatigue degree prediction method and system based on the electroencephalogram signals provided by the invention have the advantages that the individuals to be predicted are matched with the existing individuals through the specific electroencephalogram electrode channels, so that the individuals to be predicted can maintain higher prediction accuracy without complicated calibration procedures, the individuals are more convenient to use, and the acceptability is high. In addition, the invention has higher time resolution, and reduces the dimension of EEG (brain wave) data and the interference of information irrelevant to the fatigue degree on a model by extracting the electrode channel most relevant to the fatigue degree, so that the EEG data and the fatigue degree label are more easy to correspond. The data slicing work converts continuous EEG (brain wave) signals in the time domain into a 1s sample time window, and the model is built according to the 1s sample time window during modeling, so that in an actual fatigue degree real-time prediction task, fatigue degree prediction can be performed with a time resolution of 1s, and higher prediction accuracy is ensured while higher efficiency is ensured.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the method provided by the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
On the one hand, the embodiment of the invention discloses a fatigue degree prediction method based on an electroencephalogram signal, which is shown in fig. 1 and comprises the following steps:
Collecting EEG raw data E1, E2, E3...en and labels L1, L2, L3..ln of a plurality of historical subjects S1, S2, S3..sn;
Performing spectrum analysis on EEG original data of a plurality of historical testees, and selecting an electroencephalogram acquisition channel to obtain respective electroencephalogram channel sets C1, C2, C3.;
acquiring EEG original data Et of a to-be-detected testee St, performing spectrum analysis and selecting an electroencephalogram acquisition channel to obtain a target electroencephalogram channel set Ct;
selecting historical testees Sx, sy and Sz which are the same as the target electroencephalogram channel set of the testee to be detected from the historical testees;
Modeling based on EEG original data Ex, ey, ez of the selected historical testees Sx, sy, sz and labels Lx, ly, lz to obtain a fatigue prediction model;
And predicting the to-be-detected testee by using the fatigue prediction model to obtain the fatigue state of the to-be-detected testee.
In a specific embodiment, the EEG raw data and corresponding labels of a plurality of historical subjects are obtained by a calibration procedure that collects EEG data of the subjects in different fatigue states prior to performing specific predictive tasks, thereby creating a database of EEG data and corresponding fatigue state labels for training of the fatigue detection model. The method comprises the following specific steps: firstly, an individual needs to execute a calibration program under different fatigue degrees to complete a signal acquisition task, when the signal is acquired, the individual needs to wear a multichannel electroencephalogram cap to acquire EEG data and complete PVT objective reflecting capability test, and the fatigue degree of EEG signals in a corresponding time period is calibrated through response time obtained by PVT.
In a specific embodiment, a to-be-detected person wears a multi-channel electroencephalogram cap to collect EEG data, and the obtained original three-dimensional EEG data is a time sampling point multiplied by the number of electrode channels multiplied by an EEG voltage value.
In a specific embodiment, spectrum analysis is performed on the acquired EEG original data to obtain power values of alpha and beta frequency bands, all the electroencephalogram electrode channels of each tested person are ordered according to the value of alpha/beta, the largest four electrode channels are selected, and the final result is that each tested person obtains a self electroencephalogram channel set Cn, wherein the number of all the electroencephalogram electrode channels is greater than four.
In a specific embodiment, the method further comprises preprocessing the selected EEG raw data of the historical subject prior to modeling based on the selected EEG raw data and the tag of the historical subject, the specific preprocessing steps being: filtering, de-artifacting algorithm and slicing the selected EEG original data of the historical testee into EEG samples with the time length of 1s, wherein the format is as follows: 1s time sampling point x electrode channel number x EEG voltage value; after preprocessing, a machine learning algorithm is selected to build a fatigue prediction model from EEG samples of time length 1s and corresponding tags. The specific process is as follows: EEG data Ex, ey, ez. (format: 1s time sampling point x number of electrode channels x EEG voltage value) and corresponding fatigue labels Lx, ly, lz., of which 90% is training set and 10% is test set, etc. of a historical subject Sx, sy, sz., etc. of a subject to be detected, having electrode channels, are simultaneously divided. And defining a loss function and an optimizer as a two-class cross entropy loss function and an Adam optimizer respectively, finally, sending a training set into a deep learning model EEGNet for iterative training and evaluating accuracy in a testing set, and storing the model for a subsequent fatigue detection task to be detected.
According to the invention, the individual to be predicted is matched with the existing individual through the specific electroencephalogram electrode channel, so that the individual to be predicted can keep higher prediction accuracy without complicated calibration procedures, the individual use is more convenient, and the acceptability is high.
The channels most relevant to fatigue are selected through electrode channels, so that EEG (brain wave) individual differences among different tested can be effectively reduced, and the channel matching method can select individuals Sx, sy, sz. which are most similar to the tested St to be monitored from the existing tested S1, S2, S3.
The invention has higher time resolution, can convert a sample time window of 1 second into corresponding fatigue degree, and can ensure relatively higher prediction efficiency.
The channel selection method extracts the electrode channel most relevant to the fatigue degree, reduces the dimension of EEG (brain wave) data, reduces the interference of information irrelevant to the fatigue degree on a model, and enables EEG data and a fatigue degree label to be more easily corresponding. The continuous EEG (brain wave) signals in the time domain are converted into a sample time window of 1s by the data slicing operation, and a model is built according to the sample time window of 1s during modeling, so that in an actual fatigue degree real-time prediction task, fatigue degree prediction can be performed with time resolution of 1s, and high prediction accuracy is ensured while high efficiency is ensured.
On the other hand, a real-time fatigue degree prediction system based on an electroencephalogram signal is disclosed, and the system comprises the following modules:
The system comprises an acquisition module, a detection module and a detection module, wherein the acquisition module acquires EEG (electronic test) original data and labels of a plurality of historical testees and acquires EEG original data of testees to be detected;
the frequency spectrum analysis module is used for respectively carrying out frequency spectrum analysis on EEG original data of a plurality of historical testees and EEG original data of testees to be detected, and selecting an electroencephalogram acquisition channel to obtain respective electroencephalogram channel sets and target electroencephalogram channel sets of the historical testees;
the screening module is used for selecting a historical tested person which is the same as the target electroencephalogram channel set of the tested person to be detected from the historical tested person;
The fatigue prediction model building module is used for modeling based on the EEG original data and the label of the selected historical testee to obtain a fatigue prediction model;
And the result output module predicts the tested person to be detected by using the fatigue prediction model to obtain the fatigue state of the tested person to be detected.
In a specific embodiment, the method further comprises a preprocessing module, wherein the preprocessing module is used for filtering, de-artifact algorithm and slicing the selected EEG original data of the historical testee into EEG samples with the time length of 1s, and the format is as follows: 1s time sampling points x electrode channel number x EEG voltage value.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (7)
1. The fatigue degree prediction method based on the electroencephalogram signals is characterized by comprising the following steps of:
collecting EEG original data and labels of a plurality of historical testees;
Performing spectrum analysis on the EEG original data of the plurality of historical testees, and selecting an electroencephalogram acquisition channel to obtain respective electroencephalogram channel sets of the historical testees;
acquiring EEG original data of a to-be-detected tested person, performing frequency spectrum analysis and selecting an electroencephalogram acquisition channel to obtain a target electroencephalogram channel set;
Selecting a historical testee which is the same as the target electroencephalogram channel set of the testee to be detected from the historical testees;
modeling is carried out based on the selected EEG original data and the label of the historical testee, so as to obtain a fatigue prediction model;
And predicting the to-be-detected testee by using the fatigue prediction model to obtain the fatigue state of the to-be-detected testee.
2. The method for predicting the fatigue degree based on the electroencephalogram signals according to claim 1, wherein the EEG raw data and the corresponding labels of a plurality of historical testees are obtained through a calibration procedure, comprising the following specific steps: firstly, an individual needs to execute a calibration program under different fatigue degrees to complete a signal acquisition task, when the signal is acquired, the individual needs to wear a multichannel electroencephalogram cap to acquire EEG data and complete PVT objective reflecting capability test, and the fatigue degree of EEG signals in a corresponding time period is calibrated through response time obtained by PVT.
3. The fatigue degree prediction method based on the electroencephalogram signals according to claim 2, wherein a to-be-detected person wears a multi-channel electroencephalogram cap to collect EEG data, and the obtained original three-dimensional EEG data is a time sampling point multiplied by the number of electrode channels multiplied by an EEG voltage value.
4. The fatigue degree prediction method based on the electroencephalogram signals according to claim 1, wherein the acquired EEG original data is subjected to frequency spectrum analysis to obtain power values of alpha and beta frequency bands, all electroencephalogram electrode channels of each tested person are ordered according to the value of alpha/beta, the largest four electrode channels are selected, and the final result is that each tested person obtains an own electroencephalogram channel set, wherein the number of all electroencephalogram electrode channels is greater than four.
5. The method for predicting the fatigue level based on the electroencephalogram according to claim 1, wherein before modeling based on the EEG raw data and the labels of the selected historical subjects, further comprising preprocessing the EEG raw data of the selected historical subjects, wherein the preprocessing step is specifically as follows: filtering, de-artifacting algorithm and slicing the selected EEG original data of the historical testee into EEG samples with the time length of 1s, wherein the format is as follows: 1s time sampling point x electrode channel number x EEG voltage value; after pretreatment, EEG samples of time length 1s were modeled as fatigue predictions with the corresponding tags.
6. An electroencephalogram signal-based fatigue degree prediction system is characterized by comprising the following modules:
The system comprises an acquisition module, a detection module and a detection module, wherein the acquisition module acquires EEG (electronic test) original data and labels of a plurality of historical testees and acquires EEG original data of testees to be detected;
the frequency spectrum analysis module is used for respectively carrying out frequency spectrum analysis on the EEG original data of the plurality of historical testees and the EEG original data of the testees to be detected, and selecting an electroencephalogram acquisition channel to obtain respective electroencephalogram channel sets and target electroencephalogram channel sets of the historical testees;
the screening module is used for selecting a historical tested person which is the same as the target electroencephalogram channel set of the tested person to be detected from the historical tested person;
The fatigue prediction model building module is used for modeling based on the EEG original data and the label of the selected historical testee to obtain a fatigue prediction model;
And the result output module predicts the tested person to be detected by using the fatigue prediction model to obtain the fatigue state of the tested person to be detected.
7. The electroencephalogram signal based fatigue prediction system according to claim 6, further comprising a preprocessing module for filtering, de-artifacting, slicing selected EEG raw data of a historical subject into EEG samples of time length 1s in the format of: 1s time sampling points x electrode channel number x EEG voltage value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410351984.1A CN118177835A (en) | 2024-03-26 | 2024-03-26 | Fatigue degree prediction method and system based on electroencephalogram signals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410351984.1A CN118177835A (en) | 2024-03-26 | 2024-03-26 | Fatigue degree prediction method and system based on electroencephalogram signals |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118177835A true CN118177835A (en) | 2024-06-14 |
Family
ID=91406381
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410351984.1A Pending CN118177835A (en) | 2024-03-26 | 2024-03-26 | Fatigue degree prediction method and system based on electroencephalogram signals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118177835A (en) |
-
2024
- 2024-03-26 CN CN202410351984.1A patent/CN118177835A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107203692B (en) | Electrocardio data digital signal processing method based on deep convolutional neural network | |
| CN105841961A (en) | Bearing fault diagnosis method based on Morlet wavelet transformation and convolutional neural network | |
| CN105956623A (en) | Epilepsy electroencephalogram signal classification method based on fuzzy entropy | |
| CN111207875B (en) | Electromyographic signal-torque matching method based on multi-granularity parallel CNN model | |
| CN111134664B (en) | Epileptic discharge identification method and system based on capsule network and storage medium | |
| CN102085095A (en) | Method, system and electrocardioscanner for detecting ST segment in electrocardiogram | |
| CN117289067B (en) | Transformer running state on-line monitoring system | |
| CN107080545A (en) | A kind of lie detection system based on brain electricity | |
| CN112674782A (en) | Device and method for detecting epileptic-like electrical activity of epileptic during inter-seizure period | |
| CN113069117A (en) | Electroencephalogram emotion recognition method and system based on time convolution neural network | |
| CN116842361A (en) | Epileptic brain electrical signal identification method based on time-frequency attention mixing depth network | |
| CN113533945A (en) | Analog circuit fault diagnosis method based on two-dimensional convolutional neural network | |
| CN117782198B (en) | Highway electromechanical equipment operation monitoring method and system based on cloud edge architecture | |
| CN111025100A (en) | Transformer ultrahigh frequency partial discharge signal mode identification method and device | |
| CN118177835A (en) | Fatigue degree prediction method and system based on electroencephalogram signals | |
| CN117556359A (en) | Power distribution equipment health state evaluation system and method | |
| CN117204865A (en) | Steady-state visual evoked potential visual fatigue quantification method based on underdamped second-order stochastic resonance | |
| CN116725486A (en) | Dynamic electrocardiosignal atrial fibrillation detection method and device based on double phases | |
| CN114492146B (en) | Bolt group loosening positioning and quantitative analysis method and system based on transfer learning | |
| CN114224341B (en) | Wearable forehead electroencephalogram-based depression rapid diagnosis and screening system and method | |
| CN114841200A (en) | Steel rail damage monitoring method based on wavelet transform and semi-supervised convolution self-encoder | |
| CN101254098B (en) | Brain-electric ultra-slow mode wave testing analysis processing apparatus | |
| CN100427029C (en) | Multipoint sphygmus information confluence method based on sphygmus entropy | |
| Li et al. | Efficient ECG classification method for arrhythmia using MODWPT and adaptive incremental broad learning | |
| CN112244873B (en) | Electroencephalogram space-time feature learning and emotion classification method based on hybrid neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |