CN106333675B - The mask method and system of EEG signals data type under waking state - Google Patents
The mask method and system of EEG signals data type under waking state Download PDFInfo
- Publication number
- CN106333675B CN106333675B CN201610839406.8A CN201610839406A CN106333675B CN 106333675 B CN106333675 B CN 106333675B CN 201610839406 A CN201610839406 A CN 201610839406A CN 106333675 B CN106333675 B CN 106333675B
- Authority
- CN
- China
- Prior art keywords
- eeg signals
- electro
- sample
- spike
- ocular signal
- 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.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1103—Detecting eye twinkling
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/398—Electrooculography [EOG], e.g. detecting nystagmus; Electroretinography [ERG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4806—Sleep evaluation
- A61B5/4809—Sleep detection, i.e. determining whether a subject is asleep or not
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4806—Sleep evaluation
- A61B5/4812—Detecting sleep stages or cycles
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4806—Sleep evaluation
- A61B5/4815—Sleep quality
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Artificial Intelligence (AREA)
- Psychiatry (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Ophthalmology & Optometry (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Anesthesiology (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
- Eye Examination Apparatus (AREA)
Abstract
The present invention relates to the mask methods and system of the EEG signals data type under a kind of waking state, the method comprise the steps that acquiring the real-time electro-ocular signal and EEG signals sample of user after user starts sleep procedure；Wavelet decomposition is carried out to the real-time electro-ocular signal and EEG signals sample respectively, and signal reconstruction is carried out according to the wavelet coefficient of setting low-frequency range and obtains electro-ocular signal and EEG signals；According to the feature of the correlation and eye electrical waveform of blinking of synchronization EEG signals and electro-ocular signal, blink activity is detected on electro-ocular signal；When detecting blink activity, the signal type of the EEG signals sample is labeled as waking state.The present invention can be interfered to avoid EEG signals, accurately detect the waking state of EEG signals, and carry out effective data type mark, so that using the mark EEG signals sample training come out personal classifier recognition accuracy it is higher, also improve the later period to the reliability of personal sleep state testing result.
Description
Technical field
The present invention relates to assisting sleep technical fields, more particularly to the EEG signals data type under a kind of waking state
Mask method and system.
Background technique
There are some ancillary equipments on the market at present to carry out auxiliary people's sleep, i.e. assisting sleep, to improve user
Sleep quality.Sleep state analysis is the important means that ancillary equipment understands user's sleep quality, and in the process, it needs
User's sleep state is detected, accurately to know that user is awake or asleep state, then can be carried out corresponding
Intervening measure.
Polysomnogram (Polysomnography, PSG), also known as sleep electroencephalogram are clinically to examine at present for sleeping
It is disconnected and analysis " goldstandard ".Polysomnogram analyzes sleep using a variety of vital signs, in these signs,
Brain electricity is in core status；Utilize 4 species rhythm of brain wave: δ wave (1-3Hz), θ wave (4-7Hz), α wave (8-12Hz), β wave (14-
Frequency characteristic 30Hz).According to the brain wave of the different rhythm and pace of moving things and eye movement feature, other than the awake stage, sleep can be with
It is divided into non-dynamic sleep (No Rapid Eye Movement Sleep, NREM sleep) and the dynamic sleep (Rapid that is sharp-eyed of being sharp-eyed
Eye Movement Sleep, REM sleep) period.Wherein non-dynamic sleep of being sharp-eyed can be divided into 4 periods again: the S1 phase is (completely
Regain consciousness to the transition stage between sleep), the S2 phase (shallowly sleeps the stage), the S3 phase (middle deep sleep), the S4 phase (sound sleep phase).
Under normal circumstances, whether detection user is in waking state, is the signal (δ by utilizing brain wave in 4 frequency ranges
Wave frequency section, θ wave frequency section, α wave frequency section and β wave frequency section) training waking state identification model (classifier) come to EEG signals into
Row identification, these identification models are often the universal identification model being trained using other people brain wave, but due to brain electricity
A human specific of signal is very strong, and the intensity of brain electricity is very weak (brain electricity is microvolt rank, and electrocardio is millivolt rank), in signal
It is easily interfered by outer signals when acquisition.
In the process, when needing training individual's classifier, it is necessary to the number of the personal EEG signals sample of acquisition
It is labeled according to type, just self study and test can be carried out to the data of marking types in this way, trained and be more applicable for
The personal classifier of people, and using universal identification model come to EEG signals carry out detection mark when, as previously described, due to brain
The intensity of electric signal is very weak to be easy to be interfered, and the type of eeg signal sample is marked using universal identification model, is easy
It is mixed into interference component, causes the recognition accuracy for training the personal classifier come lower, affects the later period to individual's sleep shape
The reliability of state testing result.
Summary of the invention
Based on this, it is necessary in view of the above-mentioned problems, a kind of mark of the EEG signals data type under providing waking state
Method and system can accurately detect the EEG signals under waking state, and carry out the mark of data type.
A kind of mask method of EEG signals data type under waking state, comprising:
After user starts sleep procedure, the real-time electro-ocular signal and EEG signals sample of user are acquired；
Wavelet decomposition is carried out to the real-time electro-ocular signal and EEG signals sample respectively, and according to the small of setting low-frequency range
Wave system number carries out signal reconstruction and obtains electro-ocular signal and EEG signals；
According to the feature of the correlation and eye electrical waveform of blinking of synchronization EEG signals and electro-ocular signal, in eye telecommunications
Blink activity is detected on number；
When detecting blink activity, the signal type of the EEG signals sample is labeled as waking state.
A kind of labeling system of EEG signals data type under waking state, comprising:
Signal acquisition module, for after user starts sleep procedure, acquiring the real-time electro-ocular signal and brain telecommunications of user
Number sample；
Signal reconstruction module, for carrying out wavelet decomposition to the real-time electro-ocular signal and EEG signals sample respectively, and
Signal reconstruction, which is carried out, according to the wavelet coefficient of setting low-frequency range obtains electro-ocular signal and EEG signals；
Blink detection module, for the correlation and blink eye electric wave according to synchronization EEG signals and electro-ocular signal
The feature of shape detects blink activity on electro-ocular signal；
Sample labeling module, for when detecting blink activity, the signal type of the EEG signals sample to be marked
For waking state.
The mask method and system of EEG signals data type under above-mentioned waking state start sleep procedure in user
Afterwards, the real-time electro-ocular signal and EEG signals sample of user are acquired, wavelet decomposition is carried out and low-frequency range rebuilds to obtain electro-ocular signal
And EEG signals；According to the feature of the correlation and eye electrical waveform of blinking of synchronization EEG signals and electro-ocular signal, in eye
Blink activity is detected on electric signal to which the signal type of the EEG signals sample is labeled as waking state.With this solution
The type of eeg signal sample is marked, can be interfered to avoid EEG signals, the awake of EEG signals is accurately detected
State, and effective data type mark is carried out, so that the personal classification come out using the EEG signals sample training of the mark
The recognition accuracy of device is higher, also improves the later period to the reliability of personal sleep state testing result.
Detailed description of the invention
Fig. 1 is the flow chart of the mask method of the EEG signals data type under the waking state of one embodiment；
Fig. 2 is the EEG signals and electro-ocular signal schematic diagram in one common awake period；
Fig. 3 is electro-ocular signal waveform spike area schematic diagram in sliding window；
Fig. 4 is to detect the movable result schematic diagram of blink；
Fig. 5 is the waking state detector concept figure realized based on mask method of the invention；
Fig. 6 is the labeling system structural schematic diagram of the EEG signals data type under the waking state of one embodiment.
Specific embodiment
The mask method and system of the EEG signals data type under waking state of the invention are illustrated with reference to the accompanying drawing
Embodiment.
Refering to what is shown in Fig. 1, Fig. 1 be one embodiment waking state under EEG signals data types mask method
Flow chart, comprising:
Step S101 acquires the real-time electro-ocular signal and EEG signals sample of user after user starts sleep procedure；
In this step, in this step, it can be and assisting sleep is being carried out to user, when training individual's identification model, true
In the state that warranty family is awake, start to carry out EEG signals sample collection to user, related brain fax sense is worn by user
Equipment, the EEG signals that acquisition user generates in sleep procedure when acquiring EEG signals sample, while wearing phase using user
Eye electricity sensing equipment is closed to be acquired the real-time electro-ocular signal of user.
It can be that a frame is acquired with 30s, every frame EEG signals are subsequent right as a sample when acquiring signal
Every frame electro-ocular signal and EEG signals are analyzed and processed.
Generally, it is contemplated being 10~15 minutes to the normal time for falling asleep of people, time for falling asleep is very when tired/tired out
It can extremely shorten, therefore, start the EEG signals for acquiring user under the wide-awake state of user, it is believed that starting to acquire
In a set period of time T afterwards, it is believed that user is in waking state, as embodiment, the present invention select when
Between section be 300 seconds (5 minutes), i.e., user start acquire EEG signals after 300 seconds in EEG signals sample be judged as clearly
The state of waking up；
As described above, EEG signals sample is with 30s for a frame, then in the data type of preceding 10 brains electricity sample
Direct Mark is waking state.
Step S102 carries out wavelet decomposition to the real-time electro-ocular signal and EEG signals sample respectively, and according to setting
The wavelet coefficient of low-frequency range carries out signal reconstruction and obtains electro-ocular signal and EEG signals；
Wavelet decomposition is carried out to real-time electro-ocular signal and EEG signals sample respectively first, and according to the small of setting low-frequency range
Wave coefficient reconstruction electro-ocular signal and EEG signals, in order to avoid high-frequency noise interference simultaneously stick signal essential information, I
EEG signals are analyzed on compared with low-frequency range.For the convenience of calculating, the upper frequency limit (0~8Hz) of θ wave can choose
Carry out wavelet decomposition and reconstruction.
For the convenience of calculating, the upper frequency limit that can choose the θ wave (mainly 4-7Hz) of EEG signals is rebuild,
That is 0~8Hz.
Step S103, according to the spy of the correlation and eye electrical waveform of blinking of synchronization EEG signals and electro-ocular signal
Sign detects blink activity on electro-ocular signal；
In the sleep cycle of normal person, blink is an activity specific to lucid interval, electro-ocular signal when due to blink
Amplitude it is higher, EEG signals can be interfered.
It is the EEG signals and electro-ocular signal schematic diagram in one common awake period with reference to Fig. 2, Fig. 2；Solid line is in figure
EEG signals, dotted line are electro-ocular signal.It is all generated on EEG signals and electro-ocular signal by can be seen that blink activity in figure
Downward spike, blink activity show as the waveform with the high peak of short time on electroencephalogram, this is also blink eye electricity
The feature of waveform.
In above-mentioned detection process, using the correlation between the EEG signals and electro-ocular signal of synchronization, in conjunction with blinking
The wave character of eye electro-ocular signal is judged.
In one embodiment, the movable method of detection blink, be can be such that on electro-ocular signal
(1) the low frequency electro-ocular signal is intercepted using the sliding window with setting signal amplitude and time span；
Similarity degree of the present invention using signal amplitude, electro-ocular signal and EEG signals in a sliding window, spike
Acuity and spike duration to detect whether have blink activity in sliding window, when which verifies electroencephalogram
Between slide on axis, intercept electro-ocular signal waveform.
Since the time of blink is generally less than 0.4 second, can blink in the interior detection of a slightly larger sliding window electric
Signal, such as the sampling time length of 0.6 times of setting, i.e. n=0.6fs, n are sliding window length, and fs is electro-ocular signal
Sample rate.
The amplitude of electro-ocular signal can subtract minimum value (p by the maximum value of electro-ocular signal in sliding windowmax-pmin)
Mode is found out, and under normal circumstances, the amplitude of sliding window can be set to 75 microvolts between 300 microvolts.
(2) related coefficient of the waveform of electro-ocular signal and synchronization EEG signals in sliding window, sliding are calculated separately
The duration of the acuity parameter of electro-ocular signal waveform spike and spike in window；
Here acuity parameter be characterize feature high peak of short time when meeting blink of spike waveform it is strong and weak
Parameter.
In one embodiment, calculate sliding window in electro-ocular signal waveform spike acuity parameter, may include
It is as follows:
(a) upper area area and lower area area of the electro-ocular signal waveform in sliding window are calculated separately, is calculated
Formula is as follows:
In formula, piFor the electro-ocular signal in sliding window, pmaxFor the maximum value of electro-ocular signal in sliding window, pminFor cunning
The minimum value of electro-ocular signal, area in dynamic windowupIndicate upper area area, areadownIndicate lower area area；
(b) area of the electro-ocular signal waveform spike according to the upper area area and lower area areal calculation,
Calculation formula is as follows:
In formula, blinkareaIndicate the area of spike, if expression meets condition；
Refering to what is shown in Fig. 3, Fig. 3 is electro-ocular signal waveform spike area schematic diagram in sliding window, the spike in two kinds of directions
For upper and lower part region area as shown, left figure spike direction is upward, the spike direction of right figure is downward.
(c) according to spike areal calculation acuity parameter, calculation formula is as follows:
blinkratio=blinkarea/in-blinkarea
In formula, blinkratioIndicate acuity parameter, in-blinkareaIndicate the area of non-peak portion, it is sharp here
It is ratio between upper area area and lower area area that sharp extent index, which can also be converted into,.
The method for calculating the duration of electro-ocular signal waveform spike in sliding window, may include as follows:
(d) direction of the electro-ocular signal waveform spike according to the upper area area and lower area areal calculation,
Calculation formula is as follows:
In formula, blinkdirectionIndicate spike direction, labeled as 1 indicate spike it is downward, labeled as -1 indicate spike to
On；
(e) when spike is downward, the duration of spike is calculated according to the Local modulus maxima of spike two sides；Or in point
When peak is upward, the duration of spike is calculated according to the local local minizing point of spike two sides；Calculation formula is as follows:
blinkts=vertexright-vertexleft,
In formula, blinktsIndicate the duration of spike, vertexrightIt indicates the right side vertex moment of spike, indicates point
The leftmost vertices moment at peak, right_min_loca indicate the local minizing point on the right side of spike, and right_max_loca is indicated
Local modulus maxima on the right side of spike, left_min_loca indicate the local minizing point on the left of spike, left_max_loca
Indicate that the Local modulus maxima on the left of spike, s.t. indicate constraint condition.
(3) if the related coefficient, acuity parameter and duration meet respectively preset correlation coefficient threshold,
Acuity parameter threshold and duration threshold judge that there are blink activities for electro-ocular signal in the sliding window；
Specifically, the area of the similarity degree of signal amplitude, electro-ocular signal and EEG signals in the sliding window, spike
When all meeting condition with the duration, that is, thinking current sliding window mouth, there is blink activities；
For correlation coefficient threshold, it is considered that, two vectors of the related coefficient greater than 0.7 substantially may be considered just
It is relevant, it is contemplated that interference of the electro-ocular signal to EEG signals when blink, phase when blink between electro-ocular signal and EEG signals
Closing coefficient threshold can be set to 0.9；For acuity parameter threshold, 0.3 generally can be set to；For spike it is lasting when
Between threshold value, be usually 0.3-0.4 seconds according to wink time, therefore, duration threshold can be set to 0.3 second.
The signal type of the EEG signals sample is labeled as awake shape when detecting blink activity by step S104
State；
In the sleep cycle of normal person, blink is an activity specific to lucid interval, and therefore, sleep state was analyzed
Cheng Zhong, if having detected blink activity, i.e., it is believed that current EEG signals sample is to belong to awake shape within certain a period of time
State.
With reference to Fig. 4, Fig. 4 is to detect the movable result schematic diagram of blink, and solid line and dotted line are respectively to pass through small echo in figure
EEG signals and electro-ocular signal after transform reconstruction.On electro-ocular signal, circle has marked blink activity and has been formed by spike, can
With discovery, although the blink amount of activity detected is less, false detection rate is extremely low.
Further, in practical applications, in order to avoid erroneous detection goes out mistake when active belt of blinking comes to data type mark
Accidentally, by taking 30 seconds electro-ocular signals of a frame as an example, when at least to detect 2 or 2 or more blink activities in a frame signal,
Confirm the data type, the signal type of EEG signals sample is labeled as waking state, this makes it possible to mark is greatly improved
Accuracy.
The mask method of EEG signals data type under the waking state of the embodiment of the present invention utilizes brain telecommunications in many cases
In number being tested, waking state has all obtained accurate detection, it is seen that mask method of the invention accuracy with higher.
In order to further increase mark accuracy rate, on the basis of activity scheme is blinked in the detection based on electro-ocular signal, this
Invention is directed to the EEG signals of acquisition, and the data class of detection EEG signals can also be carried out by Sample Entropy manner of comparison
Type.
The technical solution of use can be such that
The Sample Entropy of the EEG signals is calculated first, then carries out the Sample Entropy with the sample entropy threshold precalculated
Compare, if the Sample Entropy is greater than the sample entropy threshold, the signal type of the EEG signals sample is labeled as regaining consciousness
State.
Sample Entropy is a kind of measurement of time series complexity, is widely used in the detection of epilepsy.In sleep cycle
6 stages in, regain consciousness the stage EEG signals Sample Entropy highest, the present invention using the size of the Sample Entropy of EEG signals come
Judge whether user is waking state, by setting sample entropy threshold, is compared with the Sample Entropy of EEG signals.
Further, in above-mentioned comparison procedure, the selection of sample entropy threshold is also a vital ring.Insane at present
The empirical data that detection of epilepsy etc. obtains is not appropriate for being used in the accurate judgement in sleep state analysis to waking state.
As in the prior embodiments, in the setting time T after starting acquisition signal, user is construed as regaining consciousness
, therefore, in the case where ensuring user's waking state, the EEG signals that just can use the user acquired in this period calculate sample
This entropy threshold.
In one embodiment, sample entropy threshold can be calculated with the following method, comprising:
The EEG signals in setting time are divided into multiple samples first, and calculate separately the sample of each EEG signals sample
This entropy obtains Sample Entropy set；In the process, it is assumed that using 30s be a frame, 300 seconds EEG signals of acquisition are handled, that
Here just there are 10 samples, the Sample Entropy set including 10 Sample Entropies can be calculated at this time.
Then sample entropy threshold is calculated according to the Sample Entropy set；As embodiment, the calculation formula of sample entropy threshold
It can be such that
sampen_vali=sampen (y [p_start:p_end])
P_start=(i-1) * time_length*fs+1
P_end=t_start+time_length*fs-1
P_end < Tfs
In formula, wherein sampen_thre is sample entropy threshold, sampen_valiFor i-th sample in Sample Entropy set
Sample Entropy, sampen are the operation for seeking Sample Entropy, and input y [p_start:p_end] is EEG signals y in pth _ start point
Start the part until pth _ end point, time_length is the time span for calculating each sample of Sample Entropy, and fs is brain
The sample rate of electric signal, T are to start to acquire the setting time after EEG signals, and v is setup parameter.
In above-mentioned numerical procedure, the value of parameter v is extremely important, can control recognition accuracy by parameter v；Cause
This can calculate parameter v value to improve recognition accuracy by following formula:
Assuming that the set X of Sample Entropy when the awake stage obeys standardized normal distribution, i-th of element representation in set X are as follows:
Wherein,
At this point, x=v
According to the integral of standard normal distribution function:
Wherein, P (X≤x) indicates that the value in the set X of Sample Entropy is less than the probability of x, it is possible thereby to calculate, with T=
For 300s, time_length=30s, as parameter v=2.58, probability of the value less than x of Sample Entropy set X is 99.5%.
Finally, by by the Sample Entropy and Sample Entropy threshold value comparison of the EEG signals calculating of present frame, if Sample Entropy is big
In sample entropy threshold, the signal type of the EEG signals sample can be labeled as waking state.
In this step, it is contemplated that the Sample Entropy of the EEG signals in awake stage is maximum, the sample calculated based on the above embodiment
This entropy threshold sampen_thre, it is known that the EEG signals by Sample Entropy greater than sampen_thre are judged as that waking state can obtain
To compared with high-accuracy.
The scheme of above-described embodiment, detection blink it is movable on the basis of, further judged by EEG signals Sample Entropy,
To judge the type of EEG signals sample, mark accuracy rate can be further increased.Technology of the invention for part interference compared with
Serious signal may will affect recall rate, but not interfere with accuracy rate, can be adapted in sleep state analysis,
When training individual's identification model, the EEG signals data type under waking state is labeled as detector.
Refering to what is shown in Fig. 5, Fig. 5 is the waking state detector concept figure realized based on mask method of the invention, simultaneously
EEG signals sample and real-time electro-ocular signal are acquired, after electro-ocular signal acquisition, by detecting the blink activity in electro-ocular signal, from
And it can be to blink situation that may be present during Sample Entropy threshold decision, if there is the EEG signals sample just to present frame
Originally it is labeled, in addition EEG signals then will do it with the calculating and threshold decision of Sample Entropy, if Sample Entropy is more than threshold value,
The mark of data type is then carried out to EEG signals sample, and thus the mark accuracy rate further marked.With EEG signals and
For mono- frame of electro-ocular signal 30s, if in electro-ocular signal, detecting at least two blink activity, and if the same period the frame brain
Electric signal is not judged to waking state, then can carry out supplement mark to it after Sample Entropy threshold decision, reach reinspection and
The function of leak repairing.
The mask method of EEG signals data type under the waking state of the embodiment of the present invention utilizes brain telecommunications in many cases
Number carry out waking state test experience in, the type of EEG signals sample has obtained accurate judgement, passes through the standard of waking state
Really judgement, can form waking state detector, if detector output result is "Yes", that is, be labeled as waking state, such as
Fruit output result then abandons the sample due to neither waking state but can not be considered sleep state for "No".
Refering to what is shown in Fig. 6, labeling system knot of the Fig. 6 for the EEG signals data type under the waking state of one embodiment
Structure schematic diagram, comprising:
Signal acquisition module 101, for after user starts sleep procedure, acquiring the real-time electro-ocular signal and brain electricity of user
Sample of signal；
Signal reconstruction module 102, for carrying out wavelet decomposition to the real-time electro-ocular signal and EEG signals sample respectively,
And signal reconstruction is carried out according to the wavelet coefficient of setting low-frequency range and obtains electro-ocular signal and EEG signals；
Blink detection module 103, for the correlation and blink eye according to synchronization EEG signals and electro-ocular signal
The feature of electrical waveform detects blink activity on electro-ocular signal；
Sample labeling module 104, for when detecting blink activity, by the signal type mark of the EEG signals sample
Note is waking state.
Under the labeling system of EEG signals data type under waking state of the invention and waking state of the invention
The mask method of EEG signals data type corresponds, the mark side of the EEG signals data type under above-mentioned waking state
Technical characteristic and its advantages that the embodiment of method illustrates are suitable for the mark of the EEG signals data type under waking state
In the embodiment of injection system, hereby give notice that.
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (8)
1. a kind of mask method of the EEG signals data type under waking state characterized by comprising
After user starts sleep procedure, the real-time electro-ocular signal and EEG signals sample of user are acquired；
Wavelet decomposition is carried out to the real-time electro-ocular signal and EEG signals sample respectively, and according to the wavelet systems of setting low-frequency range
Number carries out signal reconstruction and obtains electro-ocular signal and EEG signals；
According to the feature of the correlation and eye electrical waveform of blinking of synchronization EEG signals and electro-ocular signal, on electro-ocular signal
Detect blink activity；
When detecting blink activity, the Sample Entropy of the EEG signals is calculated, by the Sample Entropy and the Sample Entropy precalculated
Threshold value is compared, if the Sample Entropy is greater than the sample entropy threshold, the signal type of the EEG signals sample is marked
For waking state；
The calculation formula of the sample entropy threshold is as follows:
sampen_vali=sampen (y [p_start:p_end])
P_start=(i-1) * time_length*fs+1
P_end=p_start+time_length*fs-1
P_end < Tfs
In formula, wherein sampen_thre is sample entropy threshold, sampen_valiFor the sample of i-th of sample in Sample Entropy set
Entropy, sampen are the operation for seeking Sample Entropy, and input y [p_start:p_end] is that EEG signals y starts in pth _ start point
Part until pth _ end point, time_length are the time span for calculating each sample of Sample Entropy, and fs is brain telecommunications
Number sample rate, T be start acquire EEG signals after setting time, v is setup parameter.
2. the mask method of the EEG signals data type under waking state according to claim 1, which is characterized in that institute
The feature for stating the correlation according to synchronization EEG signals and electro-ocular signal and eye electrical waveform of blinking, is examined on electro-ocular signal
Survey blink activity step include:
Low frequency electro-ocular signal is intercepted using the sliding window with setting signal amplitude and time span；
Calculate separately the related coefficient of the waveform of electro-ocular signal and synchronization EEG signals in sliding window, eye in sliding window
The acuity parameter of electric signal waveform spike and the duration of spike；
If the related coefficient, acuity parameter and duration meet preset correlation coefficient threshold, sharp journey respectively
Parameter threshold and duration threshold are spent, judges that there are blink activities for electro-ocular signal in the sliding window.
3. the mask method of the EEG signals data type under waking state according to claim 2, which is characterized in that institute
Stating the step of calculating the acuity parameter of electro-ocular signal waveform spike in sliding window includes:
Upper area area and lower area area of the electro-ocular signal waveform in sliding window are calculated separately, calculation formula is such as
Under:
In formula, piFor the electro-ocular signal in sliding window, pmaxFor the maximum value of electro-ocular signal in sliding window, pminFor sliding window
The minimum value of electro-ocular signal, area in mouthfulupIndicate upper area area, areadownIndicate lower area area；
According to the area of electro-ocular signal waveform spike described in the upper area area and lower area areal calculation, calculation formula
It is as follows:
In formula, blinkareaIndicate the area of spike, if expression meets condition；
According to spike areal calculation acuity parameter, calculation formula is as follows:
blinkratio=blinkarea/in-blinkarea
In formula, blinkratioIndicate acuity parameter, in-blinkareaIndicate the area of non-peak portion.
4. the mask method of the EEG signals data type under waking state according to claim 3, which is characterized in that institute
Stating the step of calculating the duration of electro-ocular signal waveform spike in sliding window includes:
According to the direction of electro-ocular signal waveform spike described in the upper area area and lower area areal calculation, calculation formula
It is as follows:
In formula, blinkdirectionIt indicates spike direction, indicates that spike is downward labeled as 1, indicate that spike is upward labeled as -1；
When spike is downward, the duration of spike is calculated according to the Local modulus maxima of spike two sides；Or it is upward in spike
When, the duration of spike is calculated according to the local minizing point of spike two sides；Calculation formula is as follows:
blinkts=vertexright-vertexleft,
In formula, blinktsIndicate the duration of spike, vertexrightAt the right side vertex moment for indicating spike, indicate spike
Leftmost vertices moment, right_min_loca indicate the local minizing point on the right side of spike, and right_max_loca indicates spike
The Local modulus maxima on right side, left_min_loca indicate the local minizing point on the left of spike, and left_max_loca is indicated
Local modulus maxima on the left of spike, s.t. indicate constraint condition.
5. the mask method of the EEG signals data type under waking state according to claim 4, which is characterized in that institute
The amplitude for stating sliding window is 75 microvolts to 300 microvolts, the related coefficient when blink between electro-ocular signal and EEG signals
Threshold value is 0.9, and the acuity parameter threshold is 0.3, and the duration threshold is 0.3 second.
6. the mask method of the EEG signals data type under waking state according to any one of claims 1 to 5, special
Sign is, further includes:
In setting time T after starting acquisition EEG signals sample and real-time electro-ocular signal, by all EEG signals sample marks
Note is waking state.
7. the mask method of the EEG signals data type under waking state according to claim 1, which is characterized in that institute
The value for stating parameter v is as follows:
Wherein,
X=v
In formula, X indicates Sample Entropy set, XiIndicate that i-th of element in Sample Entropy set X, Φ (x) indicate standardized normal distribution letter
Several integrals, P (X≤x) indicate that the value in the set X of Sample Entropy is less than the probability of x.
8. a kind of labeling system of the EEG signals data type under waking state characterized by comprising
Signal acquisition module, for after user starts sleep procedure, acquiring the real-time electro-ocular signal and EEG signals sample of user
This；
Signal reconstruction module is used for respectively to the real-time electro-ocular signal and EEG signals sample progress wavelet decomposition, and according to
The wavelet coefficient of setting low-frequency range carries out signal reconstruction and obtains electro-ocular signal and EEG signals；
Blink detection module, for according to the correlations of synchronization EEG signals and electro-ocular signal and blink eye electrical waveform
Feature detects blink activity on electro-ocular signal；
Sample labeling module, for calculating the Sample Entropy of the EEG signals when detecting blink activity, by the Sample Entropy with
The sample entropy threshold precalculated is compared, if the Sample Entropy is greater than the sample entropy threshold, by the EEG signals sample
This signal type is labeled as waking state；
The calculation formula of the sample entropy threshold is as follows:
sampen_vali=sampen (y [p_start:p_end])
P_start=(i-1) * time_length*fs+1
P_end=p_start+time_length*fs-1
P_end < Tfs
In formula, wherein sampen_thre is sample entropy threshold, sampen_valiFor the sample of i-th of sample in Sample Entropy set
Entropy, sampen are the operation for seeking Sample Entropy, and input y [p_start:p_end] is that EEG signals y starts in pth _ start point
Part until pth _ end point, time_length are the time span for calculating each sample of Sample Entropy, and fs is brain telecommunications
Number sample rate, T be start acquire EEG signals after setting time, v is setup parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610839406.8A CN106333675B (en) | 2016-09-21 | 2016-09-21 | The mask method and system of EEG signals data type under waking state |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610839406.8A CN106333675B (en) | 2016-09-21 | 2016-09-21 | The mask method and system of EEG signals data type under waking state |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106333675A CN106333675A (en) | 2017-01-18 |
| CN106333675B true CN106333675B (en) | 2019-06-11 |
Family
ID=57840186
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610839406.8A Active CN106333675B (en) | 2016-09-21 | 2016-09-21 | The mask method and system of EEG signals data type under waking state |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106333675B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107440714A (en) * | 2017-09-07 | 2017-12-08 | 哈尔滨理工大学 | A kind of focus and blink action extraction element and method |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105105714A (en) * | 2015-08-26 | 2015-12-02 | 吴建平 | Sleep period separating method and system |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020188216A1 (en) * | 2001-05-03 | 2002-12-12 | Kayyali Hani Akram | Head mounted medical device |
| DE10243937A1 (en) * | 2002-09-20 | 2004-04-01 | MCC Gesellschaft für Diagnosesysteme in Medizin und Technik mbH & Co. KG | Method and device for determining a sleep profile |
| US20080177197A1 (en) * | 2007-01-22 | 2008-07-24 | Lee Koohyoung | Method and apparatus for quantitatively evaluating mental states based on brain wave signal processing system |
| WO2011158198A2 (en) * | 2010-06-17 | 2011-12-22 | Card Guard Scientific Survival Ltd. | A method and a system for monitoring of sleep and other physiological conditions |
-
2016
- 2016-09-21 CN CN201610839406.8A patent/CN106333675B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105105714A (en) * | 2015-08-26 | 2015-12-02 | 吴建平 | Sleep period separating method and system |
Non-Patent Citations (4)
| Title |
|---|
| 《基于EEG与EOG信号的疲劳驾驶状态综合分析》;王福旺等;《东北大学学报》;20140228;正文第2-3页，表1 |
| 《基于可穿戴传感器的驾驶疲劳肌心电信号分析》;付荣荣等;《汽车工程》;20131231;正文第2-3页 |
| 《基于眼电的智能输入系统研究》;甘玉龙等;《中国生物医学工程学报》;20151231;正文第3页左栏第一段，第4页右栏至第5页左栏 |
| 基于样本熵的睡眠脑电分期;和卫星等;《江苏大学学报》;20090930;第30卷(第5期);正文第2-3页，表1 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106333675A (en) | 2017-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106175754B (en) | Waking state detection device in sleep state analysis | |
| US20090062679A1 (en) | Categorizing perceptual stimuli by detecting subconcious responses | |
| CN109199412B (en) | Abnormal emotion recognition method based on eye movement data analysis | |
| CN104173064B (en) | Lie detecting method based on analysis of heart rate variability and device of detecting a lie | |
| CN110135285B (en) | Electroencephalogram resting state identity authentication method and device using single-lead equipment | |
| CN103699226A (en) | Tri-modal serial brain-computer interface method based on multi-information fusion | |
| CN106333676B (en) | The annotation equipment of EEG signals data type under waking state | |
| CN109199411B (en) | Case-conscious person identification method based on model fusion | |
| KR20180124327A (en) | wearable device and method for measuring the degree of drowsiness or concentration of the user | |
| CN101382837B (en) | Computer mouse control device of compound motion mode | |
| CN114145717B (en) | Sleep state analysis method based on PPG heart rate characteristic parameters and exercise amount | |
| CN109009098A (en) | A kind of EEG signals characteristic recognition method under Mental imagery state | |
| CN106344008B (en) | Waking state detection method and system in sleep state analysis | |
| Jiao et al. | Detecting driver sleepiness from EEG alpha wave during daytime driving | |
| CN106361327B (en) | Waking state detection method and system in sleep state analysis | |
| Singh et al. | Emotion recognition using electroencephalography (EEG): a review | |
| Yosi et al. | Emotion recognition using electroencephalogram signal | |
| CN106175698B (en) | Sleep cycle detection device in sleep state analysis | |
| CN106333675B (en) | The mask method and system of EEG signals data type under waking state | |
| CN106388812B (en) | The mask method and system of EEG signals data type under waking state | |
| CN106333674B (en) | Sleep cycle detection method and system in sleep state analysis | |
| CN106333677A (en) | Blinking activity detection method and blinking activity detection system in sleep state analysis | |
| CN111714089A (en) | Drug addiction evaluation system based on multi-stimulus short video event related potential | |
| CN106333678A (en) | Method and system for detecting brain wave at sleep state in electroencephalogram | |
| CN106236084B (en) | The mask method and system of EEG signals data type under waking state |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |