CN104921723B - A kind of state of consciousness detecting system based on multi-mode brain-computer interface - Google Patents
A kind of state of consciousness detecting system based on multi-mode brain-computer interface Download PDFInfo
- Publication number
- CN104921723B CN104921723B CN201510253168.8A CN201510253168A CN104921723B CN 104921723 B CN104921723 B CN 104921723B CN 201510253168 A CN201510253168 A CN 201510253168A CN 104921723 B CN104921723 B CN 104921723B
- Authority
- CN
- China
- Prior art keywords
- ssvep
- module
- hit
- consciousness
- patient
- 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]
- A61B5/377—Electroencephalography [EEG] using evoked responses
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4029—Detecting, measuring or recording for evaluating the nervous system for evaluating the peripheral nervous systems
- A61B5/4041—Evaluating nerves condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6802—Sensor mounted on worn items
- A61B5/6803—Head-worn items, e.g. helmets, masks, headphones or goggles
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Veterinary Medicine (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Neurology (AREA)
- Public Health (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Neurosurgery (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of state of consciousness detecting system based on multi-mode brain-computer interface, including brain wave acquisition cap, portable amplifier, P300 potentiometric detections module, SSVEP detection modules, decision-making module and evaluation module, wherein brain wave acquisition cap is placed in user's head, the scalp EEG signals of collection are transmitted separately to P300 potentiometric detections module, SSVEP detection modules after portable amplifier amplifies, P300 potentiometric detections module, the output end of SSVEP detection modules are connected with decision-making module respectively, and the output end of decision-making module is connected with evaluation module.The detecting system of the present invention, determines whether he can identify the target specified (facial photo or others' facial photo of oneself) by detecting the P300 and SSVEP of patient, can effectively identify whether the patient is conscious.
Description
Technical field
The present invention relates to brain-computer interface research field, more particularly to a kind of state of consciousness inspection based on multi-mode brain-computer interface
Examining system.
Background technology
Consciousness is the concept of a rich connotation, is divided into two major parts：Awakening (arousal) and awareness
(awareness).For serious cerebral injury patient, they will enter the recovery rank of stable disease after acute phase is spent
Section.Relative to acute stage, the condition assessment needs in this stage are comprehensive as far as possible, and want to react prognosis and lapse to.
According to the horizontal difference of awareness, such patient can be divided into multiple clinical states.Some patients may enter vegetative state
(vegetative state, VS), they have the cycle of Sleep-Wake, but completely lose the awareness to itself and external environment condition
Ability.Some patients may then return to minimally conscious state (minimally consciousstate, MCS), that is, exist micro-
Awareness ability that is weak but determining.As can be seen here, MCS and VS differentiation is whether evidence suggests patient has awareness ability.
On the other hand, block comprehensive disease (locked-in syndrome, LIS) patient is awakening and has intact awareness energy
Power, but body is paralysed completely.So it is also possible to cause to obscure with vegetative state or minimally conscious state.In general, no
Patient with disturbance of consciousness degree needs different therapeutic schemes, and therefore, the accurate disturbance of consciousness level for judging patient seems different
It is often important.
At present, the disturbance of consciousness degree for clinically evaluating patient relies primarily on scale (for example, Glasgow Glasgow is confused
Fan's scale, JFK stupors recover scale etc.) and clinical experience, by checking caused by eyes, speech and the aspect stimulation of motion three
Reaction carry out overall merit.Scale (JFK is recovered with the stupor of JFK medical centers Giacino of the U.S. in 2004 et al. modifications
Coma Recovery Scale-Revised, CRS-R) exemplified by, it is divided into 6 projects：The sense of hearing, vision, motion, verbal response, friendship
Stream and arousal level.When being evaluated using stupor recovery scale CRS-R to the state of consciousness of patient, if some project of patient
Score value be less than or equal to standards of grading (sense of hearing 2 divides, and vision 1 is divided, move 2 points, verbal response 2 divides, exchange 0 point, wake-up degree is not
Scoring only refers to), then its state of consciousness is rated as VS；If some project is MCS higher than standards of grading；If sports events reaches
To 6 points or exchange project reaches 2 points to depart from MCS.This kind of method is simple and easy, simple and direct effective to being gone into a coma after acute brain injury,
But then seem excessively coarse for VS and MCS patients.Sometimes it is not high to the susceptibility of change of illness state, it is impossible to which that definite reflection is clinical real
Border situation, therefore behavior is judged often with certain subjectivity and unpredictability.In addition, for serious disturbance of consciousness patient,
They there may be different degrees of damage to the motion expression system exchanged, while its arousal level has limitation and warp
It is often unstable, therefore the accuracy of behavior judgement operationally in itself is not easy to ensure.Existing document is confirmed to the VS disturbances of consciousness
The misdiagnosis rate that patient carries out conventional behavior diagnosis is up to 37%-43%.
Specifically, the shortcomings that prior art and prior art are present is as follows：
First, the technical scheme of prior art one
At present, the disturbance of consciousness degree for clinically evaluating patient relies primarily on scale (for example, Glasgow Glasgow is confused
Fan's scale, JFK stupors recover scale etc.) and clinical experience, by checking caused by eyes, speech and the aspect stimulation of motion three
Reaction carry out overall merit.
2nd, the shortcomings that prior art one
This kind of method is simple and easy, simple and direct effective to being gone into a coma after acute brain injury, but then seems for VS and MCS patients
In coarse.Sometimes it is not high to the susceptibility of change of illness state, it is impossible to definite reflection clinical practice situation, therefore behavior judges often band
There are certain subjectivity and unpredictability.In addition, for serious disturbance of consciousness patient, they are to the motion expression system that exchanges
It there may be different degrees of damage, while its arousal level has limitation and often unstable, therefore behavior judgement is in itself
Accuracy operationally is not easy to ensure.Existing more documents confirm to carry out conventional behavior diagnosis to VS disturbances of consciousness patient
Misdiagnosis rate is up to 37%-43%.
3rd, the technical scheme of prior art two
Recently, some scholars utilize FMRI (functional magnetic resonance
Imaging, fMRI) carry out the brain function that research consciousness impaired patients remain.Their purpose is that base is detected in fMRI
In the change of particular command, and provide the evidence of the awareness ability not against motion expression system.It is imaged using functional neurosurgery, it is preceding
The research of people such as calls out name, objective weighbridge amount to patient MCS using the stimulation of self instruction acoustically carry out being familiar with sound
The ability of brain in patients processing.In single case research [4], Owen et al. requires VS patient's root during fMRI is carried out
The imagination task of " playing tennis " and " in oneself room walk " is completed according to prompting.In two tasks, the patient show with just
The similar brain activation phenomenon of normal subject.After some months, the patient is diagnosed as MCS by the scoring of behavior expression.
4th, the shortcomings that prior art two
Although scholars propose much methods based on fMRI, fMRI some limitations include high cost, no
Portable (can only be carried out in the nuclear magnetic resonance room of large hospital) and the requirement of strict body (can not be forbidden to move with metallic support
Leave many restrictions such as body) limit to its extensive use in consciousness impaired patients.Different from fMRI, brain-computer interface is due to its phase
To cheap price and good portability, thus can be applied to very well clinically.
5th, the technical scheme of prior art three
Currently, brain-computer interface also begins to the detection for carrying out state of consciousness to disturbance of consciousness patient.Lule et al. [5] is right
13 MCS patients, 3 VS patients and 2 LIS patients have carried out the 4 classification brain-computer interface tests based on sense of hearing P300.Training
After stage terminates, each patient must be asked by paying attention in sound sequence the answer 10 that is repeated to of " YES " or " NO "
Topic.One patient LIS achieves 60% correct response, and another LIS patient only achieve 20% accuracy rate and can not make
Exchanged with BCI.In addition, BCI can be used to carry out online exchange without any MCS or VS patients.
6th, the shortcomings that prior art three
Brain-machine interface method currently used for detecting state of consciousness is all realized using based on acoustic stimuli.Sense of hearing brain machine
One common fault of interface is exactly that accuracy rate is low.At present from the point of view of the service condition of normal person, vision brain-computer interface is than sense of hearing brain
The accuracy rate of machine interface is far better.What the present invention designed is exactly the meaning that disturbance of consciousness patient is carried out using vision brain-computer interface
Know detection.
In summary, using the more objective and method of science come to detect the state of consciousness of patient be very necessary.
The content of the invention
The shortcomings that it is an object of the invention to overcome prior art and deficiency, there is provided a kind of based on multi-mode brain-computer interface
State of consciousness detecting system.
The purpose of the present invention is realized by following technical scheme：
A kind of state of consciousness detecting system based on multi-mode brain-computer interface, including brain wave acquisition cap, portable amplifier,
P300 potentiometric detections module, SSVEP detection modules, decision-making module and evaluation module, wherein brain wave acquisition cap are placed in user's head,
The scalp EEG signals of collection are transmitted separately to P300 potentiometric detections module, SSVEP detection moulds after portable amplifier amplifies
Block, P300 potentiometric detections module, the output end of SSVEP detection modules are connected with decision-making module respectively, the output end of decision-making module
It is connected with evaluation module；Wherein
The brain wave acquisition cap, two photos are shown on a graphical user interface, when user wants to select a kind of photograph, use
Family watches corresponding photograph and the number of silent number photo frame flicker attentively；SSVEP is produced by the flicker of photograph, at the same time, P300 electricity
After position is produced by the flicker of photo frame, scalp EEG signals are gathered；
The portable amplifier, record scalp EEG signals；Then scalp EEG signals are copied into two parts, passed respectively
Transport to P300 potentiometric detections module, SSVEP detection modules；
The P300 potentiometric detections module, its course of work are as follows：
(1) scalp EEG signals carry out bandpass filtering in 0.1-10Hz frequency ranges, and carry out 1/5 down-sampling；For
The several sample value samplings in one sample sequence interval once, so obtain the down-sampling that new sequence is exactly former sequence；1/5 down-sampling is just
Refer to sample 1 time at interval of the sample value of 5 EEG signals；
(2) vector of 10 passages is connected, and by being averaged out 5 round flicker, so as to construct corresponding each phase
The characteristic vector of frame；Mark, generated by these training set datas corresponding every using corresponding to these characteristic vectors and they
SVMs (SVM) grader of patient；The SVM classifier of generation will use in following online P300 detections；
(3) in online synchronized algorithm, P300 detections are to carry out once for every 10 seconds, the sudden strain of a muscle of corresponding 5 round photo frames
It is bright；Similarly, by being averaging to 5 round vectors corresponding to 5 flickers, so as to obtain characteristic vector；These features to
Amount is input in the support vector machine classifier above generated, is corresponded to 2 fractions of 2 photo frames respectively；
The SSVEP detection modules, its course of work are as follows：
(1) bandpass filtering is carried out to EEG signals in the range of 4-20Hz；Secondly, choose in 8 passages and flash the stage
10 second datas vector, and the method combined using least energy produces new signal vector；Using Fourier transform, calculate new
The power density spectrum of the signal vector of definition；
(2) SSVEP energy is calculated by integrating the power density spectrum of flicker frequency and its harmonic wave；
(3) calculate per the narrow band energy of sheet photo and the ratio of wide band energy；
The decision-making module, the ratio detected using the rule of addition come the fraction with reference to above P300 detections and SSVEP,
And the target of detection is determined by finding the maximum index of additive value, feedback is finally used as using the photograph of detection；
The evaluation module, accuracy rate are obtained by the trial correctly responded number divided by whole trial number
Come；In order to weigh whether accuracy rate is notable, by counting hit (hit) and being not hit by the number of observation (observed of (miss)
Frequencies) and theoretical number (expected frequencies), and according to following equation Chi-square statistic is carried out：
Wherein, foiAnd feiBeing i-th respectively, (i=1,2 ..., k) number of observation of individual classification and theoretical number, observe item
Mesh is divided into hit, is not hit by two classes, fo1And fo2It is hit and the number of observation that is not hit by respectively, fe1And fe2It is hit respectively
With the theoretical number being not hit by, and the free degree be classification item number k subtract 1, i.e. df=1；Sentence when having carried out 50 times using two class BCI
Regularly, should be that to hit (hit) and be not hit by the number of (miss) all should be 25 in the case where full-probability is equal；In order to
Show the conspicuousness (p≤0.05) of number of observation, test statistics χ2(df=1) need to be more than 3.84, i.e., in 50 trial
In, patient needs correctly to respond in >=32 trial, now judges that user's state of consciousness is normal.
Described 2 fractions for corresponding to 2 photo frames respectively represent credibility of two photo frames in P300 detections, are used for
Decision phase below.
8 described passages are specially " P7 ", " P3 ", " Pz ", " P4 ", " P8 ", " O1 ", " Oz " and " O2 ".
Described arrowband and the scope in broadband are each set to ± 0.1Hz and ± 1Hz, described narrow band energy and broadband
The ratio of energy represents credibility of two sheet photos in SSVEP detections respectively, with the decision phase later.
Described 10 second datas vector includes 2500 data points.
The present invention compared with prior art, has the following advantages that and beneficial effect：
The clinical detection that disturbance of consciousness patient retains consciousness is very challenging, and correct diagnosis to patient, shield
Reason strategy and quality of life are all abnormal important.In the present invention, we introduce the multi-mode brain-computer interface based on P300 and SSVEP
Cognitive function and subliminal possibility are retained to improve its detection.In each trial of experiment, two kinds of photographs, a trouble
The photograph of person oneself and others' photograph, random display is on the GUI left side and the right.A photo frame is embedded in per sheet photo
In.Two sheet photos are respectively with 6Hz and 7.5Hz frequency scintillation so as to producing SSVEP.In the same time, the photo frame of two sheet photos
Flashed with random order so as to produce P300 current potentials.Patient is required to watch the photograph of oneself or others' photograph attentively, and writes from memory
The flashing times of the corresponding photo frame of number.Brain-computer interface determines which sheet photo patient is look at by P300 and SSVEP detections.
Experimental duties are completed, this needs many cognitive abilities, such as the ability to understand speech to instruction, two sheet photos
Object Selection ability, remember to do what working memory and lasting notice (removing fixation object photograph 10 seconds).It is any
A kind of missing of cognitive ability is likely to the failure for causing the task.In addition, often there is false negative in brain-computer interface research
As a result.Therefore, for that could not show the patient of the ability of toeing the line in this experiment, we can not directly think that they are exactly not
It is conscious.However, positive result can be shown that these patients have cognitive function really, so as to show that they are conscious
's.
Some existing researchs on P300 and SSVEP can support our conclusion from another point of view.On the one hand, P300
It is always treated as a cognitive potential for depending on notice and working memory.P300 modulation be by it is conscious perception come
Complete, such as stimulate masking, pay attention to formation and anaesthesia, these all protrude its effect as consciousness mark.On the other hand,
Although current research can't explain the specific underlying mechanisms of SSVEP well, there are some researchs to point out that SSVEP is
The reaction of some cognitive variables (as paid attention to, stimulating classification and memory search etc.).In addition, P300 and SSVEP are directed to cortex
A series of activation in area, and depend on cognitive ability.In the present invention, patient can produce in two Run experiment
P300 (such as Fig. 3,4) and SSVEP (such as Fig. 5,6) responses.This shows that they can voluntarily adjust the brain activity of oneself, and this reacts
Some cognitive abilities.These EEG evidences can be led to the same conclusion, and patient is conscious.
Meanwhile the subsequent rehabilitation situation of these patients demonstrates the validity of the system.Participating in 8 trouble of this research
In person, there is 1 patient (VS1) to be identified as being completely in vegetative state after behavior evaluation repeatedly.But it have been found that she
The multi-mode brain-computer interface based on P300 and SSVEP can be used come selection target photograph, and it is horizontal to reach acceptable operation.
This means the patient that some behavior evaluations are VS has the cognitive function and or even self consciousness of residual.It is furthermore interesting that 1
MCS patient (MCS1) can use our multi-mode brain-computer interface in an experiment.After this 2 months, he was recovered simultaneously
Departing from MCS.And other patient clinicals still keep original state constant.Do not exchange and surpass with the external world in view of the patient
1 year is spent, and does not always show the clinical manifestation to toe the line, this experimental result is extremely long-pending for patient and its household
Pole.By to before distributing new dispatchs (10 months after experiment), the health of the patient has obtained further rehabilitation, and can use mouth
Head expression is simply exchanged.Therefore, our multi-mode brain-computer interface can in the level of understanding of identification disturbance of consciousness patient
Can be sensitiveer than some clinical means.
Brief description of the drawings
Fig. 1 is the workflow diagram of the state of consciousness detecting system of the present invention based on multi-mode brain-computer interface；
Fig. 2 is the normal form schematic diagram of consciousness detection；
Fig. 3 is the average P300 oscillograms of patient's passage " Pz " in Run 1 is tested；Wherein, solid line correspond to include
P300 the target button, dotted line correspond to the non-targeted button not comprising P300；
Fig. 4 is the average P300 oscillograms of patient's passage " Pz " in Run 2 is tested；Wherein, solid line correspond to include
P300 the target button, dotted line correspond to the non-targeted button not comprising P300；
Fig. 5 be patient under 6Hz target frequency 8 selected passages (" P7 ", " P3 ", " Pz ", " P4 ", " P8 ", " O1 ",
" Oz " and " O2 ") EEG signal average power spectral density figure；Wherein, the deeper point of color represents spot photograph flicker frequency
Spectrum energy, and the shallower point of color represents the spectrum energy of non-targeted photograph flicker frequency；
Fig. 6 be patient under 7.5Hz target frequency 8 selected passages (" P7 ", " P3 ", " Pz ", " P4 ", " P8 ",
" O1 ", " Oz " and " O2 ") EEG signal average power spectral density figure；Wherein, the deeper point of color represents spot photograph flicker
The spectrum energy of frequency, and the spectrum energy for putting the non-targeted photograph flicker frequency of expression that color is shallower.
Embodiment
With reference to embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are unlimited
In this.
As Fig. 1, a kind of state of consciousness detecting system based on multi-mode brain-computer interface are specific to include with lower part：
First, interface and controlling mechanism:
In order to cause the attention of patient, we use face's photograph as stimulating, in this, as graphic user interface
(GUI).The front of two full faces is shone to be randomly selected out from Liang Ge faces photo album (each photo album has 10 sheet photos) is inner, and
It is randomly placed at the GUI left side and the right.One be patient oneself face's photograph, another is with sex but trouble with patient
The unfamiliar face's photograph of person.Face's photograph of each patient is obtained from their family members.In all photographs, do not have
The situation that appearance patient's wear a pair of spectacles or face are blocked by hair.We trim to photograph, remove unnecessary background, but protect
The profile of face is stayed to include the difference of hair style.In addition, we use Photoshop 7.0 (Adobe, the San of Adobe companies
Jose, Calif) photograph is uniformly processed so that the global brightness and contrast of all photographs is consistent in white background.
Every face's photograph (size：6.6 9 centimetres of cm x) it is placed on a photo frame (size：8.6 11 centimetres of cm x, border is wide
Degree：1 centimetre) in.Horizontal range between two photo frames is 4 centimetres.Size (area) ratio of photograph, photo frame and GUI is solid
It is scheduled on 0.07:0.1:1.
The face's photograph and the unfamiliar face's photograph of a patient of one patient oneself are randomly placed at the GUI left side
And the right.Two sheet photos are flashed with 6.0Hz and 7.5Hz respectively by the form from occurring disappearance, and flicker frequency corresponds to GUI's
The left side and the right.Meanwhile two photo frames strengthen (become white) with random order respectively by becoming the form of white.
Two sheet photos flash from occurring disappearance form, and its flicker frequency corresponds to the GUI left side and the right is respectively 6.0Hz
And 7.5Hz.Meanwhile two photo frames flash (become white) respectively in a random order, per the increasing for including 200ms between flashing twice
Strong and 800ms interval.Therefore, a round includes the flicker of two photo frames, needs to continue a 2000ms (round quilt altogether
It is defined as the complete cycle of each photo frame flicker once).When user wants to select a kind of photograph, he/her needs to watch attentively accordingly
Photograph and the number of silent number photo frame flicker.By such interface, SSVEP is produced by the flicker of photograph, at the same time, P300 electricity
Position is produced by the flicker of photo frame.
Detection process such as Fig. 2 elaborates the normal form of consciousness detection.Each patient carries out two Run test.At first
Run, patient requests watch the photograph of oneself, and the flashing times of silent oneself photo frame of number attentively.In second Run, patient requests watch attentively
Others' photograph, and the flashing times of silent others' photo frame of number.Each stage is made up of 5 block.One is only carried out daily
Block test.Each block has 10 trial.Each trial at first, above screen test by system
The instruction of task, and play the instruction of family members' recording.Instruction is：" watch the photograph of oneself (others) attentively, number oneself of writing from memory
The flashing times of (others) photo frame." simultaneously, two sheet photos, a face's photograph and others' face's photograph of oneself, point
It is not illustrated in the GUI left side and the right randomly.After 6 seconds, two sheet photos carry out the flicker in cycle, and two corresponding photo frames are carried out
Random flicker.After 10 seconds, photograph stops flicker and photo frame stops enhancing.System is in as feedback result using the photograph detected
Present GUI center.If testing result is correct, the applause for continuing 4 seconds will be used for increasing the enthusiasm of patient as feedback.
According to the degree of fatigue of patient, there is the of short duration time of having a rest between each trial, until patient is ready to.
2nd, eeg signal acquisition
We use the NuAmps portable amplifiers (Nuamps 7181, Compumedics of Compumedics companies
USA, Charotte, NC) record scalp EEG signals.In signal acquisition process, patient wears the brain wave acquisition of LT37 types
Cap, it is sitting on the wheelchair of comfortable Armrest, and is accompanied by a doctor and a family members.The EEG signals of all passages with
Auris dextra is dashed forward as reference, wherein " HEOG " represents eye movement with " VEOG " two passages, therefore is left out herein.This experiment
Only with " Fz ", " Cz ", " P7 ", " P3 ", " Pz ", " P4 ", " P8 ", " O1 ", " Oz " Yu " O2 " this 10 passages EEG signals.
During brain wave acquisition, the impedance value of all electrodes is all below 5k Ω.EEG signals are with 250Hz frequency sampling, and 0.1
Bandpass filtering is carried out in the range of to 30Hz.
3rd, data processing and algorithm:
P300 and SSVEP detection is to separate design.Eeg data is duplicated into two parts, then has respectively entered simultaneously
In two testing processes.Fig. 3 shows the data handling procedure of P300 and SSVEP detections.Specifically used analysis method and algorithm
As described below:
1) P300 potentiometric detections：First, EEG signals carry out bandpass filtering in 0.1-10Hz frequency ranges, and carry out 1/
5 down-samplings.The signal that each passage gathers is divided into data cell, each data cell (40 data points) is from one group
Represented in vector form in millisecond time from 0 to 800 after button flicker.If patient watches a certain sheet photo attentively, the vector
P300 waveform can be included.Then, we connect the vector of 10 passages, and by being averaged out 5 round sudden strain of a muscle
It is bright, so as to construct the characteristic vector of corresponding each photo frame.Marked using corresponding to these characteristic vectors and they, pass through these instructions
Practice SVMs (SVM) grader of the corresponding every patient of collection data generation.The SVM classifier of generation is by following online
Used in P300 detections.
In online synchronized algorithm, P300 detections are to carry out once for every 10 seconds, the flicker of corresponding 5 round photo frames.Together
Sample, by being averaging to 5 round vectors corresponding to 5 flickers, so as to obtain characteristic vector.These characteristic vectors are defeated
Enter into the SVM classifier above generated, corresponded to 2 fractions of 2 photo frames respectively.This 2 fractions represent two photo frames
Credibility in P300 detections uses the decision phase later.
2) SSVEP is detected：First, bandpass filtering is carried out to EEG signals in the range of 4Hz to 20Hz.Next, we
Choose the 10 second datas vector in flicker stage in 8 passages (" P7 ", " P3 ", " Pz ", " P4 ", " P8 ", " O1 ", " Oz " and " O2 ")
(2500 data points), and the method combined using least energy produces new signal vector.Use Fourier transform, Wo Menji
Calculate the power density spectrum of the signal vector newly defined.Then, we are by integrating the power density spectrum of flicker frequency and its harmonic wave
To calculate SSVEP energy.Finally, we calculate the narrow band energy of every sheet photo and the ratio of wide band energy.Herein, arrowband
Scope and broadband range are each set to ± 0.1Hz and ± 1Hz.This 2 ratios represent two sheet photos and detected in SSVEP respectively
In credibility with decision phase later.
3) decision phase：We are using the rule being added come the fraction with reference to above P300 detections and the ratio of SSVEP detections
Rate, and the target of detection is determined by finding the maximum index of additive value, feedback is finally used as using the photograph of detection.
4) evaluation criteria：Accuracy rate is got by the trial correctly responded number divided by whole trial number.
In order to weigh whether accuracy rate is notable, we are by counting hit (hit) and being not hit by the number of observation (observed of (miss)
Frequencies) and theoretical number (expected frequencies), and according to following equation Chi-square statistic is carried out：
Wherein, foiAnd feiThe number of observation of (i=1,2 ..., k) the individual classification that is i-th respectively and theoretical number.Herein,
Observation item is divided into two classes (hit and miss), fo1And fo2It is hit and miss number of observation respectively, fe1And fe2It is respectively
Hits and misses theoretical number, and the free degree are that classification item number k subtracts 1 (i.e. df=1).Carried out when using two class BCI
Should be to hit (hit) and be not hit by the number of (miss) all should be in the case where full-probability is equal when judging for 50 times
25.In order to show the conspicuousness of number of observation (p≤0.05), test statistics χ2(df=1) need to be more than 3.84.I.e. at 50
In trial, patient needs correctly to respond in >=32 trial.
Above-described embodiment is the preferable embodiment of the present invention, but embodiments of the present invention are not by above-described embodiment
Limitation, other any Spirit Essences without departing from the present invention with made under principle change, modification, replacement, combine, simplification,
Equivalent substitute mode is should be, is included within protection scope of the present invention.
Claims (5)
1. a kind of state of consciousness detecting system based on multi-mode brain-computer interface, it is characterised in that including brain wave acquisition cap, portable
Formula amplifier, P300 potentiometric detections module, SSVEP detection modules, decision-making module and evaluation module, wherein brain wave acquisition cap are placed in
User's head, the scalp EEG signals of collection be transmitted separately to after portable amplifier amplifies P300 potentiometric detections module,
SSVEP detection modules, P300 potentiometric detections module, the output end of SSVEP detection modules are connected with decision-making module respectively, decision model
The output end of block is connected with evaluation module；Wherein
The brain wave acquisition cap, two photos are shown on a graphical user interface, when user wants to select a kind of photograph, user's note
Depending on corresponding photograph and the number of silent number photo frame flicker；SSVEP is produced by the flicker of photograph, at the same time, P300 current potentials by
After the flicker of photo frame produces, collection cap collection scalp EEG signals；
The portable amplifier, record scalp EEG signals；Then scalp EEG signals are copied into two parts, be transmitted separately to
P300 potentiometric detections module, SSVEP detection modules；
The P300 potentiometric detections module, its course of work are as follows：
(1) scalp EEG signals carry out bandpass filtering in 0.1-10Hz frequency ranges, and carry out 1/5 down-sampling；
(2) vector of 10 passages is connected, and by being averaged out 5 round flicker, so as to construct corresponding each photo frame
Characteristic vector；Marked using corresponding to these characteristic vectors and they, corresponding every patient is generated by these training set datas
Support vector machine classifier；
(3) by being averaging to 5 round vectors corresponding to 5 flickers, so as to obtain characteristic vector；These characteristic vectors
It is input in the support vector machine classifier above generated, is corresponded to 2 fractions of 2 photo frames respectively；
The SSVEP detection modules, its course of work are as follows：
(1) bandpass filtering is carried out to EEG signals in the range of 4-20Hz；Secondly, choose and the 10 of the stage is flashed in 8 passages
Second data vector, and the method combined using least energy produces new signal vector；Using Fourier transform, new definition is calculated
Signal vector power density spectrum；
(2) SSVEP energy is calculated by integrating the power density spectrum of flicker frequency and its harmonic wave；
(3) calculate per the narrow band energy of sheet photo and the ratio of wide band energy；
The decision-making module, using the rule of addition come the fraction with reference to above P300 detections and the ratio of SSVEP detections, and lead to
Cross and find the maximum index of additive value to determine the target of detection, feedback is finally used as using the photograph of detection；
The evaluation module, by counting hit and the number of observation being not hit by and theoretical number, and carried out according to following equation
Chi-square statistic：
<mrow>
<msup>
<mi>&chi;</mi>
<mn>2</mn>
</msup>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>k</mi>
</munderover>
<mfrac>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>fo</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>fe</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mrow>
<msub>
<mi>fe</mi>
<mi>i</mi>
</msub>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
Wherein, foiAnd feiBe i-th respectively (i=1,2 ..., k) number of observation of individual classification and theoretical number, observation item point
To hit, being not hit by two classes, fo1And fo2It is hit and the number of observation that is not hit by respectively, fe1And fe2Respectively be hit and not
The theoretical number of hit, and the free degree are that classification item number k subtracts 1, i.e. df=1；Test statistics χ2(df=1) need to be more than
3.84, now judge that user's state of consciousness is normal.
2. the state of consciousness detecting system according to claim 1 based on multi-mode brain-computer interface, it is characterised in that described
Respectively correspond to 2 photo frames 2 fractions represent two photo frames P300 detection in credibility, with decision-making rank later
Section.
3. the state of consciousness detecting system according to claim 1 based on multi-mode brain-computer interface, it is characterised in that described
8 passages be specially " P7 ", " P3 ", " Pz ", " P4 ", " P8 ", " O1 ", " Oz " and " O2 ".
4. the state of consciousness detecting system according to claim 1 based on multi-mode brain-computer interface, it is characterised in that described
Arrowband and the scope in broadband be each set to ± 0.1Hz and ± 1Hz, the ratio point of described narrow band energy and wide band energy
Credibility of two sheet photos in SSVEP detections is not represented.
5. the state of consciousness detecting system according to claim 1 based on multi-mode brain-computer interface, it is characterised in that described
10 second datas vector include 2500 data points.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510253168.8A CN104921723B (en) | 2015-05-15 | 2015-05-15 | A kind of state of consciousness detecting system based on multi-mode brain-computer interface |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510253168.8A CN104921723B (en) | 2015-05-15 | 2015-05-15 | A kind of state of consciousness detecting system based on multi-mode brain-computer interface |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104921723A CN104921723A (en) | 2015-09-23 |
| CN104921723B true CN104921723B (en) | 2017-12-01 |
Family
ID=54109409
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510253168.8A Active CN104921723B (en) | 2015-05-15 | 2015-05-15 | A kind of state of consciousness detecting system based on multi-mode brain-computer interface |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104921723B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105266805B (en) * | 2015-10-23 | 2018-10-09 | 华南理工大学 | A kind of state of consciousness detection method based on audio visual brain-computer interface |
| CN105411580B (en) * | 2015-10-27 | 2018-05-11 | 中国航天员科研训练中心 | A kind of brain control wheelchair system based on tactile auditory evoked potential |
| CN108335728A (en) * | 2018-02-07 | 2018-07-27 | 广州贝方医疗设备有限公司 | A kind of name of calling out for disturbance of consciousness patient promotees system of waking up |
| EP3756542A1 (en) | 2019-06-25 | 2020-12-30 | Koninklijke Philips N.V. | Apparatus for patient sedation monitoring |
| CN110363242B (en) * | 2019-07-11 | 2021-07-23 | 中南大学 | Brain consciousness multi-classification method and system based on support vector machine |
| WO2021237429A1 (en) * | 2020-05-25 | 2021-12-02 | Center For Excellence In Brain Science And Intelligence Technology, Chinese Academy Of Sciences | A systematic device and scheme to assess the level of consciousness disorder by using language related brain activity |
| CN113197586B (en) * | 2021-04-13 | 2022-08-23 | 复旦大学 | Self-consciousness detection method based on visual electroencephalogram signals |
| CN113778580B (en) * | 2021-07-28 | 2023-12-08 | 赤子城网络技术（北京）有限公司 | Modal user interface display method, electronic device and storage medium |
| CN115424108B (en) * | 2022-11-08 | 2023-03-28 | 四川大学 | Cognitive dysfunction evaluation method based on audio-visual fusion perception |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1538823A (en) * | 2001-06-13 | 2004-10-20 | Method and apparatus for monitoring consciousness | |
| CN104434311A (en) * | 2013-09-13 | 2015-03-25 | 深圳迈瑞生物医疗电子股份有限公司 | Physiological parameter processing method and system and monitoring and nursing equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7689274B2 (en) * | 2007-11-30 | 2010-03-30 | Palo Alto Research Center Incorporated | Brain-wave aware sleep management |
-
2015
- 2015-05-15 CN CN201510253168.8A patent/CN104921723B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1538823A (en) * | 2001-06-13 | 2004-10-20 | Method and apparatus for monitoring consciousness | |
| CN104434311A (en) * | 2013-09-13 | 2015-03-25 | 深圳迈瑞生物医疗电子股份有限公司 | Physiological parameter processing method and system and monitoring and nursing equipment |
Non-Patent Citations (2)
| Title |
|---|
| Noninvasive BCIs:Multiway Signal-Processing Array Decompositions;Andrzej Cichocki et.al.;《Computer》;20081031;第34-42页 * |
| 脑机接口技术在意识障碍领域应用的前景展望;李远清;《中华神经创伤外科电子杂志》;20150430;第1卷(第2期);第60-61页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104921723A (en) | 2015-09-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104921723B (en) | A kind of state of consciousness detecting system based on multi-mode brain-computer interface | |
| US12042288B2 (en) | Systems and methods for assessing and improving sustained attention | |
| KR102477327B1 (en) | Processor-implemented systems and methods for measuring cognitive ability | |
| Chung et al. | Involuntary attention restoration during exposure to mobile-based 360 virtual nature in healthy adults with different levels of restorative experience: Event-related potential study | |
| JP6755356B2 (en) | Detection device | |
| CN105496404A (en) | Assessment method of auxiliary CRS-R scale auditory shock based on brain-computer interface | |
| CN105266805A (en) | Visuoauditory brain-computer interface-based consciousness state detecting method | |
| CN109645994A (en) | A method of based on brain-computer interface system aided assessment vision positioning | |
| Pandian | Sleep pattern analysis and improvement using artificial intelligence and music therapy | |
| CN113192600A (en) | Cognitive assessment and correction training system based on virtual reality and eye movement tracking | |
| Xiao et al. | Visual fixation assessment in patients with disorders of consciousness based on brain-computer interface | |
| Wallace et al. | EEG/ERP: Within episodic assessment framework for cognition | |
| Tan et al. | Virtual classroom: An ADHD assessment and diagnosis system based on virtual reality | |
| CN107080521A (en) | It is a kind of that the method that the exchange of CRS R scales is assessed is aided in based on brain-computer interface | |
| Wan et al. | An attention level monitoring and alarming system for the driver fatigue in the pervasive environment | |
| US20190090802A1 (en) | Virtual reality device for diagnosis of nerve disorder, system and method for thereof | |
| TWI793457B (en) | Virtual reality based automatic attention-deficit diagnosis method and system | |
| Massoz | Non-invasive, automatic, and real-time characterization of drowsiness based on eye closure dynamics | |
| Anand et al. | Eye-tracker based test for assessing cognition | |
| Ji et al. | Diagnostic and rehabilitation system for Alzheimer’s disease based on virtual reality technology in rehabilitation rooms | |
| Kenny | A comprehensive investigation of repetitive head impacts (heading) on brain activity and biomechanics in varsity women's soccer | |
| Galvan et al. | Using Cognitive Dual-Task to Evaluate Individuals With Mild Traumatic Brain Injuries: A Pilot Study | |
| South et al. | Neuroimaging in autism spectrum disorders | |
| Harrison | Psychophysiological Biomarkers of Concussion Recovery | |
| Rizvi et al. | Analyzing neural activity under prolonged mask usage through EEG |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190712 Address after: 510640 Tianhe District, Guangdong, No. five road, No. 381, Co-patentee after: Guangzhou South China University of Technology Asset Management Co., Ltd. Patentee after: Li Yuanqing Address before: 510640 Tianhe District, Guangdong, No. five road, No. 381, Patentee before: South China University of Technology |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190820 Address after: 510670 Room 201, Building 72, Nanxiang Second Road, Huangpu District, Guangzhou City, Guangdong Province Patentee after: South China Brain Control (Guangdong) Intelligent Technology Co., Ltd. Address before: 510640 Tianhe District, Guangdong, No. five road, No. 381, Co-patentee before: Guangzhou South China University of Technology Asset Management Co., Ltd. Patentee before: Li Yuanqing |
|
| TR01 | Transfer of patent right |