CN101159086B - Calling device based on brain electric information detection - Google Patents
Calling device based on brain electric information detection Download PDFInfo
- Publication number
- CN101159086B CN101159086B CN200710036195A CN200710036195A CN101159086B CN 101159086 B CN101159086 B CN 101159086B CN 200710036195 A CN200710036195 A CN 200710036195A CN 200710036195 A CN200710036195 A CN 200710036195A CN 101159086 B CN101159086 B CN 101159086B
- Authority
- CN
- China
- Prior art keywords
- service
- max
- calling
- module
- brain
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a calling device based on electrical brain information demodulation, which is to provide the calling device based on stable state visual induction electric potential demodulation for a ward calling system to substitute for a keypad type calling machine. The calling device consists of a service sign screen, an electrical brain collector, an electrical brain amplifier, an A/D converter and a calling extension machine; wherein the service sign screen is a LED display screen. The electric brain collector consists of 3-5 active electrodes, is fixed on the scalp on the position of occipital lobe of hind brain of a patient and is connected with the electric brain amplifier. Multi-lead electric brain signals are amplified by the electric brain amplifier and are transmitted to the A/D converter. The A/D converter transforms simulation signals into digital signals and transmits the digital signals to the calling extension machine. LED display control software and electric brain processing and analyzing software are arranged on the calling extension machine. The LED display control software controls the display of the service sign screen. The electric brain processing and analyzing software processes and analyzes the electric brain data in real time and transmits calling information to a host machine. The invention can substitute non-contact and non-action type fixation operation for keypad type operation.
Description
Technical field
The present invention relates to brain-Computer Interface Technology, especially a kind of can be applicable to the ward based on the calling device in contactless, the non-action type calling system of brain electric information demodulation.
Background technology
The ward calling system is that must set in the present hospital ward is equipped with, and is used for patient or accompanies and attends to and send calling Xiang the doctor when needed, thereby in time obtain doctor's concern, and it all has very big value for patient nursing and ward management.The ward calling system generally is made up of called host and calling set (or address is extension set), and calling set generally is one on each sick bed, generally hangs or is fixed on the wall of sick bed top, is used for sending call signal to called host; Called host is a computing machine, generally is positioned at the duty room at doctor place, is used for receiving calling signal, give the alarm Xiang the doctor, and the indicating call berth.Exist the ward calling system of a lot of manufacturer production in the market, its pattern is had nothing in common with each other, and function is difference to some extent also.For example, the communication between main frame and calling set, the employing RS485 bus that has, the employing wireless telecommunications that have; The information of storable call history quantity of information of main frame and demonstration is had nothing in common with each other; The information content that calling set sends has to have more to be lacked or the like.But these calling systems have a common feature, and promptly calling set is button or touch, and their use must be to talk with the doctor after directly sending call contents or start intercommunication function by button by button.
Therefore above-mentioned touch call device requires to go touch key-press to use by the motion of arm and finger, and for the patient of following several situations, such calling set is just inapplicable: the patient who 1) carries out arm operation or arm amputation; 2) other arm motion limited or because of extremely feeble can't movable arm the patient; 3) part or total loss the motorius disease patient of locomitivity, as half body/total paralysis, notochord lateral sclerosis (Amyotrophic Lateral Sclerosis, ALS) patient, their arm can't move or can't do the motion under the brain domination fully.What is more, has quite a few patient also to lose linguistic function, oneself idea beyond expression of words.To these patients, it all is unpractical allowing them use touch-tone calling system to remove to send signal or converse with the doctor.
For this class patient's treatment, generally be under medical personnel's meticulous nurse and treatment at present,, progressively recover its motion or language function by some training and rehabilitation maneuver.Yet can not obtain gratifying effect under many circumstances.This mainly be because, at first, the recovery of motion or language function is a process that complexity is very long, and need professional's guidance and the operation of craftsmenship and patient's close fit, yet sustain damage owing to the brain district sometimes or reason such as patient's economic bearing, this or even in the cards never.In addition, the patient such as can't move after these paralysis, paralysis, the operation, owing to can't oneself utilize calling system, when being arranged, needs express the idea of oneself effectively to the external world, so the medical care person of going into must be arranged or the relatives that stay in the ward to look after the patient night and day accompany constantly about, by observing and experiencing and make judgement, this work is wasted time and energy and is needed a lot of experiences, if improper measures just causes unnecessary damages to patient's psychology and health possibly.
(Brain-Computer Interface is to utilize EEG signals to realize human brain and extraneous a kind of new effective information reception and registration and the communication modes that exchanges and control BCI) to brain-Computer Interface Technology.It discerns people's intention by the spatiotemporal mode that corresponding different brain activity embodied in detection and the differentiation EEG signals.The research original intention of brain-Computer Interface Technology is to serve the motorius disease patient, and it does not need the participation of people's limb activity or language performance, is non-action, contactless.Its ultimate principle is: human brain is when carrying out different thinking activities or be subjected to different outside stimulus (as vision, the sense of hearing or sense of touch), its nervous activity meeting embodies different patterns in the EEG signals of correspondence, by real-time collection and extract these patterns, just can judge the patient and carry out which type of conscious activity at present or be subjected to which type of stimulation.
General non-intrusion type, the undamaged electrode that contacts with scalp that use in brain-computer interface system.Traditional need on scalp, prepare in advance when being used by moving electrode, the special conductive paste of needs reduces the impedance between electrode and the skin, its effect often depends on operator's experience and technology, and spended time is more, can cause unnecessary psychological pressure to the patient.In addition, along with acquisition time is elongated, conductive paste can become dry, and makes the signal quality variation.The active electrode that grew up in recent years, carry out the signal adjustment placing prefilter, improve signal performance, special advantages is arranged at aspects such as anti-interference, stable, securities from the very near place of scalp, and do not need scalp is done special pre-service, can significantly reduce setup time.
Because EEG signals is picked up from scalp, the electrical activity that interested brain district takes place is delivered to the electrode place and has been subjected to a lot of interference, comprising Nonlinear Superposition of the electrical activity of the influence of inhomogeneous mediums such as skull and eubolism (as heartbeat, breathing) etc., so under the normal condition, EEG signals is rambling, must realize effective communication by experimental design cleverly and advanced signal processing technology.In experimental design, use more be event-related brain potential (Event related potential, ERP), i.e. the potential change that environmental stimuli or specific brain activity brought out, to it research comparatively extensively and ripe.(Visual evoked potential is to use more a kind of event-related brain potential in brain-computer interface VEP) to VEP, and it is the reflection of brain to the electrophysiological mechanism of Vision information processing processing.When the frequency of outside visual stimulus is higher than 6Hz, just can in EEG signals, cause periodic response, by forming with the harmonic components that stimulates same frequency and integral multiple, be called the stable state vision inducting current potential (Steady-statevisual evoked potential, SSVEP).By detecting the variation of corresponding frequencies composition in the EEG signals, can judge the frequency of current stimulation.Utilize this point, can design stimulation target, when fixation object, detect the variation of EEG signals medium frequency composition, know that promptly current which target what watch attentively be, thereby realize different selection tasks with different frequency flicker.Rise in the eighties, VEP just is applied in early stage brain-computer interface system, and 1992, Sutter developed the brain-computer interface system based on VEP.At present about the signal characteristic of VEP and disposal route comparative maturity all, by appropriate design frequency carrier pattern, recognition correct rate, recognition speed and reliability based on the brain-computer interface system of VEP can both reach higher level, therefore, VEP also becomes the best that is used for brain-computing machine communication at present and brings out current potential.In addition according to achievement in research, the space distribution of VEP on cerebral cortex mainly concentrates on corticocerebral occipital lobe position, therefore as long as place some electrodes in this zone, need not gather at whole head, in actual applications, these characteristics can be simplified collecting device greatly and improve processing speed.
Be to detect in real time the variation of EEG signals medium frequency composition to generally comprise two steps: 1) extract the feature relevant based on the key of the brain-computer interface system of VEP with the frequency content of being paid close attention to; Whether 2) the conspicuousness of measurement frequecy characteristic promptly significantly to enough thinking owing to watching the increase that respective objects has caused the corresponding frequencies composition attentively, is exactly that the design category device is classified from the angle of pattern-recognition, and makes judgement.About first step feature extraction, present operable method has:
1) fast Fourier transform (FFT);
2) spectrum method of estimation;
3) time filtering+spatial filtering: earlier by the non-frequency of stimulation composition of electrode filtering, the spatial alternation that supervision is arranged then, as public space pattern (Common spatial pattern, CSP), canonical correlation analysis (Canonical correlation analysis, CCA), offset minimum binary (Partial least squares, PLS) etc.
Judge that about the second step conspicuousness the main sorter that uses has:
1) linear classifier, as linear discriminant analysis (Linear discriminant analysis, LDA);
2) non-linear sorter comprises: artificial neural network (Artificial neural networks, ANN), support vector machine (Support vector machine, SVM), and various nuclear learning method (Kernelmethod) etc.
Though the research in brain-computer interface field has obtained a lot of soul-stirring achievements, but present research also is in the laboratory exploratory stage mostly, really the EEG signals feature is used to serve the auxiliary product of physical disabilities and rehabilitation also for number is very few, yet there are no open report as for the calling system based on the detection of stable state vision inducting current potential of various technology more than the integrated use.
Summary of the invention
The technical problem to be solved in the present invention is to replace calling set for the ward calling system designs a kind of novel ward calling device based on the detection of stable state vision inducting current potential, replaces present button operation with the operation of watching attentively contactless, non-action type.
The present invention is made up of service mark screen, brain wave acquisition device, eeg amplifier, A/D converter, calling extension set.The service mark screen is connected with the calling extension set by netting twine, the electrode of brain wave acquisition device is fixed in patient's scalp, electrode links to each other with the input end of eeg amplifier by lead, the output terminal of eeg amplifier links to each other with the input end of A/D converter, and the output terminal of A/D converter links to each other by the USB mouth with the calling extension set.Calling out extension set is a computing machine, be connected to the called host that is positioned at duty room (be with the identical called host of the described ward of background technology calling system) by bus, call out service mark screen subsidiary LED demonstration Control Software and brain electric treatment and analysis software are installed on the extension set, LED shows that Control Software has the video library of being made up of the video file about patients ' demand that shows on the service mark screen, brain electric treatment and analysis software are handled in real time and are analyzed eeg data, send call information to called host.
The service mark screen is LED display, size is more than or equal to 1.5m * 0.2m, hang on the wall or ceiling that patient can watch attentively, a ward can be placed polylith according to circumstances, but require not to be placed on patient's sight line under state of nature and be easy to the place staying or be easy to watch attentively, as the wall of sick bed right opposite or the ceiling directly over the sick bed, because dropping on these positions easily, patient's sight line causes erroneous judgement, promptly takeing for patient needs certain service and sends calling, therefore to depart from out these positions to a certain extent, not be difficult to the corner seeing or be blocked but also be placed on patient.Each piece LED display all is connected with the calling extension set by netting twine, and shows the Control Software playing video file by calling out the subsidiary LED of extension set operation LED display, and the demonstration of LED display is controlled.
Video file is to represent the sign image of service content and the background image data presented stream that circulates continuously, is stored in and calls out on the extension set, shows that by LED Control Software plays, and shows synchronously in real time on the service mark screen.Every kind of video file has specific meanings and special display mode, is representing a kind of stimulation mode to patient's vision.Video file designs according to dissimilar patients' needs, method for designing is: every LED display is evenly divided into n piece (n generally gets 4～7), every zone shows the literal of the different service content of representative or corresponding picture (as feed, analgesia, first aid, change dressings etc.) with different desired display frequency f respectively, and every zone is all with the demonstration that circulates of the mode of sign (literal or picture) and background switching.The span of desired display frequency f is 6～15Hz, and each regional desired display frequency has nothing in common with each other.The frame per second of setting video file is Z, promptly shows the Z two field picture p.s., and every two field picture is combined side by side by the image in n zone, the image that each is regional or be the sign picture of representative service content, or be background picture.Bidding will picture shows once that every the d frame then Dui Ying desired display frequency is the frame per second/d=Z/d of f=video file.The value of f, Z, d is determined by following conditions: when Z gets certain round values, d gets n (n generally gets 4～7) different round values d1, d1 ..., during dn, obtain n different desired display frequency f 1 by f=Z/d, f1 ..., fn should satisfy 6≤min{f
i}＜max{f
i}≤15, and the interval between each fi is greater than 0.5Hz.For example, when Z gets 60, d got 4～10 o'clock, and corresponding desired display frequency is respectively 15Hz, 12Hz, 10Hz, 8.57Hz, 7.5Hz, 6.67Hz, 6Hz.Different desired display frequencies is selected in each zone from these frequencies, thereby shows according to each self-corresponding d frame period.If each zone of first frame video image is the show label picture all, then to be sign picture or background picture by separately d calculate in each zone in the every two field picture in back obtains, thereby designs corresponding video file.Adopt this method for designing, again according to various disease patient's different needs, even consider the special requirement of individual patients, can design different video files targetedly, can have the quantity (corresponding to the number difference of zoning) of different COS, different service content (corresponding to show label picture difference) such as them, or the like.These video files form the video library that LED shows Control Software, show Control Software selection use for LED.
The brain wave acquisition device is made up of 3～5 active electrodes, with the auxiliary patient's hindbrain occipital lobe position (if the patient wear neck brace also can utilize neck brace auxiliary fixing, can not bring extra misery to the patient) of being fixed in of headband.Each electrode is all drawn by lead, converges into a branch ofly, receives the input end of eeg amplifier.Eeg amplifier amplifies the multi-lead EEG signals of coming from a plurality of active electrodes, and amplified signal is sent into A/D converter, and A/D converter is a digital signal with analog signal conversion, again by USB interface incoming call extension set.
Brain electric treatment and analysis software are made up of impedance detection module, primary module, interface and control module and visualization model.
The impedance detection module is measured the impedance that active electrode contacts with scalp, and is suitable thereby whether definite electrode is worn, and could begin to gather EEG signals.This module and visualization model cooperate, and utilize visualization technique, demonstrate position and the title of each electrode on scalp, show with different signs respectively according to the impedance magnitude of electrode, thereby show the scope of impedance intuitively.This module is moved before the acquired signal after wearing electrode at every turn, and idiographic flow is as follows:
1) sends the impedance detection instruction to the brain wave acquisition device.
2) from calling out extension set USB interface reading of data, the data of this moment are not EEG signals, and represent the resistance value of counter electrode.
3) resistance value with each electrode sends to visualization model.
4) after the resistance value of all electrodes is all normal, withdraw from this module.
Primary module is finished the obtaining of EEG signals, signal filtering, feature extraction and pattern discrimination, is made of signal acquisition module, signal filtering module, feature extraction and pattern discrimination module.Its flow process is as follows:
Signal acquisition module reads the digitized original EEG signals that sends from A/D converter from the USB interface of calling out extension set, carries out data parsing, and is stored in internal memory, and store into simultaneously on the disk, is provided with the back off-line analysis and uses.Idiographic flow is as follows:
1) sends brain wave acquisition pattern steering order to the brain wave acquisition device, the transmission of beginning eeg data.
2) (10ms～80ms), read the data that collect in this time interval at regular intervals at interval from calling out the extension set USB interface.
3) carry out data parsing, promptly according to sample frequency, data length, number of poles continuous acquisition to data separate by electrode, be stored in the internal memory.
4) the deposit data after resolving to disk.
The signal filtering module adopts classical finite impulse response (FIR) (FIR) or infinite impulse response (IIR) wave filter that the EEG signals in the internal memory is carried out bandpass filtering by electrode, the filter transmission band scope should cover the service mark screen and go up the desired display frequency of each viewing area, be defaulted as 4～20Hz, also can be configured this parameter by menu.
At filtered EEG signals, feature extraction and pattern discrimination module select for use a kind of feature extracting method to extract needed feature, and select for use a kind of sorter to judge current electrode wearer and whether watching the service mark screen attentively, if watching attentively, which piece zone what then further judgement was watched attentively is, promptly need which kind of service, the zone of then whether corresponding results promptly being watched attentively the service mark screen and watching attentively sends to interface and control module, if differentiate the result for not watching the service mark screen attentively, just send zero, if watch attentively, just send the code name of respective service.Idiographic flow is as follows:
1) selected feature extracting method, comprise that a kind of in fast fourier transform, spectrum method of estimation, public space pattern, canonical correlation analysis, the offset minimum binary method carries out conversion on frequency and the space to filtered data, on the frequency spectrum after the conversion, the spectral magnitude that extracts each regional desired display frequency place of employed video file is as feature then.For the video file that is divided into n zone, establish its corresponding desired display frequency and be respectively f
1, f
2..., f
nCorresponding f
i(i=1 ..., the spectral magnitude of n) locating is P
i, the n dimensional feature vector that then extracts is F=[P
1, P
2..., P
n].
2) selected sorter (in linear discriminant analysis, artificial neural network, the support vector machine a kind of), F classifies to proper vector.Sorter is to obtain via the features training to a large amount of crowds before system uses.With proper vector F input category device, F is to frequency f in sorter output
i(i=1 ..., significance S n)
i(i=1 ..., n), its meaning is that F is by f
iThe possibility size that is excited.Get maximal value max_S=max{S
i, corresponding frequency is f
Max_iMax_S and excitation threshold thd_S (obtaining via the training to sorter) are compared.If max_S＜thd_S, show that F can't help any f
iInstitute excites, and this moment, the service mark screen was without any being watched attentively in the zone; If max_S 〉=thd_S then is judged as F by f
Max_iInstitute excites, and promptly max_i piece zone is watched attentively.
3) differentiation that sends sorter to interface and control module decision as a result.If F can't help any f
iInstitute excites, and then decision is 0; If F is by f
Max_iInstitute excites, and then decision is region labeling max_i.
Interface and control module are done corresponding processing after receiving the differentiation result that feature extraction and pattern discrimination module send in the primary module, and call information is sent to the called host that is positioned at duty room by bus.Interface and control module inside are provided with the current differentiation result's of record service state variable service and the timer timer of record persistent call time, and the time threshold thd_T (2s～5s is defaulted as 3s) of the expression call duration upper limit.Idiographic flow is as follows:
1) if feature extraction and pattern discrimination module send and differentiate as a result that decision is zero (show and do not watch any service mark attentively), with the service zero clearing, if timer starts then stops.
2) if the differentiation that feature extraction and pattern discrimination module send as a result decision be that service mark shields certain regional label max_i, if service is zero, then putting service is max_i, timer begins to start timing simultaneously; If service is non-vanishing, then compare service and max_i, if service ≠ max_i, then putting service is max_i, timer restarts timing simultaneously; If service==max_i, then timer continues timing, and compare with thd_T, if timer＞thd_T, then send information (call time, call out disease room number, berth, the service that needs etc.) to called host by bus, and the prompting patient that sounds calls out success, then with the service zero clearing and stop timer.
Visualization model is responsible for calling out some information of demonstration on the extension set according to the ruuning situation of impedance detection module, primary module, interface and control module, is specially:
1) when the impedance detection module is moved, position, title, the resistance value information of show electrode on cortex, and it is big or small with different sign show electrodes to press resistance value, if there is the bigger electrode of resistance value, the prompting user reexamines the situation of wearing of this electrode, till its resistance value drops to normally.
2) when primary module moves, show the EEG signals waveform that collects, filtered waveform, differentiate the result according to user's selection (operating) by menu.
3) when the operation of interface and control module, shows current differentiation result, the time status of timer, text prompt to the situation of called host transmission information, after sending successfully.
The process of using the present invention to call out is:
At first connect the power supply of service mark screen, start and call out extension set.The serial ports of calling out extension set is connected to (bus links to each other with the called host that is positioned at outside duty room) on the bus.Call out extension set and start LED demonstration Control Software, in video library, select to play the video of required demonstration.Be patient wear brain wave acquisition device, and the bundle conductor of electrode received in the eeg amplifier, the output terminal of eeg amplifier is linked to each other with the input end of A/D converter that the output terminal of A/D converter is connected with the USB interface of calling out extension set.Call out extension set and start brain electric treatment and analysis software, operation impedance detection module is adjusted wearing of brain wave acquisition device, and the impedance that makes each electrode determines that EEG signals can normally gather in normal.Move primary module then, interface and control module be operation automatically thereupon.The patient watches certain sign on the service mark screen attentively continuously when calling out, keep the stable of head in this process as far as possible, till hearing that calling out extension set sends the auditory tone cues of access success.Behind the access success, system automatically resets, and waits for next time and calling out.
Adopt the present invention can reach following technique effect:
1) EEG signals is gathered in the not damaged mode, to not injury of patient body.The brain wave acquisition device adopts active electrode, makes that the setup time of gathering is short, and anti-interference, security are good.
2) especially at dyskinesia patient design, need not to start or open one's mouth,, can realize calling corresponding sign representative content as long as the show label of required service is carried out watching attentively continuously of certain hour, easy to use, need not training.It is worthy of note that the present invention is suitable for too to other patient, and can complement one another with the ordinary call system.
3) call contents classification is many, and content and display mode are various, can consider different patients' special requirement, and the hommization degree is higher.
4) the present invention is simple in structure, easy to assembly, volume is moderate, is convenient to move.
Description of drawings
Fig. 1: each parts connection diagram of the present invention;
Fig. 2: the present invention uses schematic layout pattern;
Fig. 3: a kind of service mark screen video shows synoptic diagram;
Fig. 4: brain electric treatment of the present invention and analysis software basic flow sheet;
Fig. 5: the interface of brain electric treatment of the present invention and analysis software and control module process flow diagram.
Embodiment
Fig. 1 is a connection diagram of the present invention, and the present invention is made up of service mark screen 1, brain wave acquisition device 2, eeg amplifier 3, A/D converter 4, calling extension set 5.Service mark screen 1 is connected with calling extension set 5 by netting twine 6, the electrode of brain wave acquisition device 2 utilizes headband to be fixed in patient's head, electrode links to each other with the input end of eeg amplifier 3 by lead 7, the output terminal of eeg amplifier 3 links to each other with the input end of A/D converter 4, and the output terminal of A/D converter 4 links to each other by the USB mouth with calling extension set 5.Call out extension set operation service logo screen 1 subsidiary LED and show Control Software, play the video file that on service mark screen 1, shows, move brain electric treatment and analysis software simultaneously, eeg data is carried out real-time processing and analysis.Call out extension set 5 and be connected to called host by bus 8.
Fig. 2 is each component layouts synoptic diagram of the present invention, and service mark screen 1 hangs on the ceiling of sick bed opposite or side wall or top, and a ward can be placed a plurality of according to circumstances.Its position can be seen the patient and is got final product, and is easy to the place staying or be easy to watch attentively but should depart from sight line under patient's normal condition, to avoid erroneous judgement.Patient lies on the sick bed, is wearing brain wave acquisition device 2, and lead 7 is connected to eeg amplifier 3, the A/D converter 4 that places sick bed one side and calls out extension set 5.Call out extension set 5 operation LED and show the video that the Control Software broadcast is set, service mark screen 1 is shown control, move brain electric treatment and analysis software simultaneously, eeg data is handled in real time and analyzed.
Fig. 3 is that a kind of service mark screen video file shows synoptic diagram.The displaying contents of service mark screen is to show that by the LED that calls out on the extension set 5 video file that Control Software is play determines.Design according to video file, every LED display is evenly divided into several zones (among the figure be example with four), and every zone shows the literal of representing different service content and picture that accordingly can its meaning of visual representation (as feed, analgesia, first aid, change dressings etc.) respectively.Wherein, word segment (comprising Chinese and English) always shows always, the picture part then with and the demonstration that circulates of the mode switched of background (promptly without any showing).The display frequency difference of zones of different picture.For making the patient can clearly be seen that the picture of demonstration, in a cycle period (time that promptly shows a picture and a background), the time ratio of accounting for of picture and background is 2: 1～4: 1.
Fig. 4 is brain electric treatment and an analysis software basic flow sheet of calling out operation on the extension set 5.This software is made up of with control module and visualization model impedance detection module, primary module (comprising signal acquisition module, signal filtering module, feature extraction and pattern discrimination module), interface.Its flow process is:
Behind the software startup, at first the menu by software interface is provided with the operational factor of primary module, specifically comprises:
1) the logical scope of the band of filtering (should comprise the desired display frequency range of selected video, can Use Defaults 4～20Hz);
2) feature extracting methods (fast fourier transform, spectrum method of estimation, public space pattern, canonical correlation analysis, offset minimum binary method);
3) method of pattern discrimination (linear discriminant analysis, artificial neural network, support vector machine);
4) content of visualization display.
Above-mentioned parameter is also unified to save as configuration file, is put in the corresponding catalogue of software, directly loads during use to get final product.
After parameter setting finishes, start the impedance detection module, send the impedance detection instruction, read the electrode impedance Value Data from calling out the extension set USB interface, and do visualization display to the brain wave acquisition device.Check to adjust the situation of wearing of big impedance electrodes, till the impedance of all electrodes is all in normal range.
Withdraw from the impedance detection module, the operation primary module carries out the real-time collection and the processing of brain electricity, and according to setting the information that shows when pre-treatment.Flow process is as follows:
1) signal acquisition module sends brain wave acquisition pattern steering order to the brain wave acquisition device, the transmission of beginning eeg data.
2) signal acquisition module (10ms～80ms), read the data that collect in this time interval at regular intervals at interval from calling out the extension set USB interface.
3) signal acquisition module carries out data parsing, promptly according to sample frequency, data length, port number parameter continuous acquisition to data separate by passage, be stored in the internal memory.According to parameter configuration, whether select the deposit data after resolving to disk.
4) signal filtering module adopts classical finite impulse response (FIR) (FIR) or infinite impulse response (IIR) wave filter that the EEG signals in the internal memory is carried out bandpass filtering by passage.
5) selected feature extracting method carries out conversion on frequency and the space to filtered data in the configuration of feature extraction and pattern discrimination module operation parameter, on the frequency spectrum after the conversion, the frequency band energy that extracts each regional desired display frequency place of employed video file is as feature then.For the video file that is divided into n zone, establish its corresponding desired display frequency and be respectively f1, f2 ..., fn.Corresponding fi (i=1 ..., the spectral magnitude of n) locating is Pi, the n dimensional feature vector that then extracts is F=[P
1, P
2..., P
n].
6) selected sorter is classified to the feature F that extracts previously in the configuration of feature extraction and pattern discrimination module operation parameter.With feature F input category device, sorter output characteristic F to frequency f i (i=1 ..., significance Si n) (i=1 ..., n), its meaning is F by the possibility size that fi excited.Get maximal value max_S=max{Si}, corresponding frequency is fmax_i.Max_S and excitation threshold thd_S are compared.If max_S＜thd_S, show that F can't help any fi and excites, service mark shielded and was watched attentively without any the zone this moment; If max_S 〉=thd_S then is judged as F and is excited by fmax_i, promptly max_i piece zone is watched attentively.
7) feature extraction and pattern discrimination module send the judged result decision of sorter to interface and control module.If F can't help any fi and excites, then decision is 0; If F is excited by fmax_i, then decision is region labeling max_i.
Fig. 5 calls out the brain electric treatment of operation on the extension set 5 and the interface and the control module process flow diagram of analysis software.The meaning of three variablees of inside modules setting is:
Service: the service state variable that writes down current differentiation result
Timer: the timer of record persistent call time
Thd_T: the time threshold of the expression call duration upper limit
The flow process of this module is as follows:
1) receives differentiation decision as a result from feature extraction and pattern discrimination module.
2) judge whether decision is zero.If,, withdraw from module if timer starts then stops with the service zero clearing.
3) if decision is non-vanishing, also be that decision is that service mark shields certain regional label max_i, whether judgment variable service is zero.If putting service is max_i, timer begins to start timing, withdraws from module.
4), judge whether service and max_i equate if service is non-vanishing.If not, putting service is max_i, and timer restarts timing simultaneously, withdraws from module.
5) if service and max_i equate the timer timing that adds up.
6) judge whether timer＞thd_T.If send information (call time, call out disease room number, berth, the service that needs etc.) by bus to called host, and the prompting patient that sounds calls out success; With the service zero clearing, stop timer.Withdraw from module.
In the middle of above-mentioned each module operational process, the displaying contents that visualization model is set in being provided with according to parameter demonstrates content corresponding in real time on the screen of calling out extension set 5.
Claims (1)
1. calling device based on brain electric information demodulation, it is made up of service mark screen (1), brain wave acquisition device (2), eeg amplifier (3), A/D converter (4) and calling extension set (5); Service mark screen (1) be a LED display, and size is more than or equal to 1.5m * 0.2m, hangs on the wall or ceiling that patient can watch attentively, is connected with calling extension set (5) by netting twine (6); Brain wave acquisition device (2) is made up of 3～5 active electrodes, electrode is fixed in the scalp of patient's hindbrain occipital lobe position, electrode links to each other with the input end of eeg amplifier (3) by lead (7), polylith can be placed according to circumstances in (1) ward of service mark screen, is easy to the place staying or be easy to watch attentively but require not to be placed on patient's sight line under state of nature; The output terminal of eeg amplifier (3) links to each other with the input end of A/D converter (4), and eeg amplifier (3) amplifies the multi-lead EEG signals of coming from a plurality of active electrodes, and amplified signal is sent into A/D converter (4); The output terminal of A/D converter (4) links to each other by the USB mouth with calling extension set (5), and A/D converter (4) is that digital signal is passed through USB interface incoming call extension set (5) with analog signal conversion; Calling out extension set (5) is a computing machine, is connected to the calling exchange by bus (8), calls out service mark screen (1) subsidiary LED demonstration Control Software and brain electric treatment and analysis software are installed on the extension set (5); LED shows that Control Software has the video library of being made up of video file, by playing video file the demonstration of service mark screen (1) is controlled; Video file is to represent the sign image of service content and the background image data presented stream that circulates continuously, is stored in and calls out on the extension set (5), shows that by LED Control Software plays, and shows synchronously in real time on service mark screen (1); Every kind of a kind of stimulation mode of video file representative to patient's vision; Video file designs according to dissimilar patients' needs, method for designing is: every LED display is evenly divided into the n piece, every zone shows the literal or the corresponding picture of the different service content of representative respectively with different desired display frequency f, and every zone is all with the demonstration that circulates of the mode of sign and background switching; The every two field picture of video file is combined side by side by the image in n zone, the image that each is regional or be the sign picture of representative service content, or be background picture; Bidding will picture shows once that every the d frame then Dui Ying desired display frequency is f=Z/d, and Z is the frame per second of video file; Different desired display frequencies is selected in each zone, thereby shows according to each self-corresponding d frame period; If each zone of first frame video image is the show label picture all, then to be sign picture or background picture by separately d calculate in each zone in the every two field picture in back obtains, thereby designs corresponding video file; N gets 4～7, and f is 6～15Hz, and the value of Z and d is determined by following condition: when Z gets certain round values, d gets n different round values d
1, d
1..., d
nThe time, obtain n different desired display frequency f by f=Z/d
1, f
1..., f
n, should satisfy 6≤min{f
i}＜max{f
i}≤15, and each f
iBetween the interval greater than 0.5Hz; Z gets 60, and d gets 4～10, and corresponding desired display frequency is respectively 15Hz, 12Hz, 10Hz, 8.57Hz, 7.5Hz, 6.67Hz, 6Hz; It is characterized in that described brain electric treatment and analysis software handle in real time and analyze eeg data, send call information to called host, brain electric treatment and analysis software are made up of impedance detection module, primary module, interface and control module and visualization model:
1.1 the impedance detection module is measured the impedance that active electrode contacts with scalp, it is suitable thereby whether definite electrode is worn, could begin to gather EEG signals, it and visualization model cooperate, utilize visualization technique, demonstrate position and the title of each electrode on scalp, show with different signs respectively, thereby show the scope of impedance intuitively according to the impedance magnitude of electrode; This module is moved before the acquired signal after wearing electrode at every turn, and flow process is as follows:
1) sends the impedance detection instruction to brain wave acquisition device (2);
2) from calling out extension set (5) USB interface reading of data, the data of this moment are not EEG signals, and represent the resistance value of counter electrode;
3) resistance value with each electrode sends to visualization model;
4) after the resistance value of all electrodes is all normal, withdraw from this module;
1.2 primary module is made of signal acquisition module, signal filtering module, feature extraction and pattern discrimination module:
1.2.1 signal acquisition module reads the digitized original EEG signals that sends from A/D converter (4) from the USB interface of calling out extension set (5), carry out data parsing, and be stored in internal memory, and store on the disk simultaneously, be provided with the back off-line analysis and use, idiographic flow is as follows:
1) sends brain wave acquisition pattern steering order to brain wave acquisition device (2), the transmission of beginning eeg data;
2) read the data that collect in this time interval every 10ms～80ms from calling out extension set (5) USB interface;
3) carry out data parsing, promptly according to sample frequency, data length, number of poles continuous acquisition to data separate by electrode, be stored in the internal memory;
4) the deposit data after resolving to disk;
1.2.2 the signal filtering module adopts finite pulse response FIR or infinite impulse response iir filter that the EEG signals in the internal memory is carried out bandpass filtering by electrode, the filter transmission band area requirement covers the desired display frequency that service mark screen (1) is gone up each viewing area, is defaulted as 4～20Hz;
1.2.3 feature extraction and pattern discrimination module select for use a kind of feature extracting method to extract needed feature from filtered EEG signals, and select for use a kind of sorter to judge current electrode wearer and whether watching service mark screen (1) attentively, if watching attentively, which piece zone what then further judgement was watched attentively is, promptly need which kind of service, the zone of then whether corresponding results promptly being watched attentively service mark screen (1) and watching attentively sends to interface and control module, if differentiate the result for not watching service mark screen (1) attentively, just send zero, if watch attentively, just send the code name of respective service; Idiographic flow is as follows:
1) selected feature extracting method, comprise that a kind of in fast fourier transform, spectrum method of estimation, public space pattern, canonical correlation analysis, the offset minimum binary method carries out conversion on frequency and the space to filtered data, on the frequency spectrum after the conversion, the spectral magnitude that extracts each regional desired display frequency place of employed video file is as feature then; For the video file that is divided into n zone, establish its corresponding desired display frequency and be respectively f
1, f
2..., f
n, corresponding f
i(i=1 ..., the spectral magnitude of n) locating is P
i, the n dimensional feature vector that then extracts is F=[P
1, P
2..., P
n].
2) selected sorter is linear discriminant analysis or artificial neural network or support vector machine, and F classifies to proper vector: with proper vector F input category device, F is to frequency f in sorter output
i(i=1 ..., significance S n)
i(i=1 ..., n), its meaning is that F is by f
iThe possibility size that is excited; Get maximal value max_S=max{S
i, corresponding frequency is f
Max_iMax_S and excitation threshold thd_S are compared, and thd_S is obtained by the training to sorter; If max_S＜thd_S, show that F can't help any fi and excites, service mark shielded (1) and was watched attentively without any the zone this moment; If max_S 〉=thd_S then is judged as F by f
Max_iInstitute excites, and promptly max_i piece zone is watched attentively;
3) differentiation that sends sorter to interface and control module decision as a result is if F can't help any f
iInstitute excites, and then decision is 0; If F is by f
Max_iInstitute excites, and then decision is region labeling max_i;
1.3 interface and control module are done corresponding processing after receiving the differentiation result that feature extraction and pattern discrimination module send in the primary module, and call information is sent to called host by bus (8); Idiographic flow is as follows:
1) if feature extraction and pattern discrimination module send and differentiate as a result that decision is zero, with the service zero clearing, if the timer timer of record persistent call time starts then stops;
2) if the differentiation that feature extraction and pattern discrimination module send as a result decision be service mark screen (1) certain regional label max_i, if service is zero, then putting service is max_i, timer begins to start timing simultaneously; If it is non-vanishing to write down current differentiation result's service state variable service, then compare service and max_i, if service ≠ max_i, then putting service is max_i, timer restarts timing simultaneously; If service==max_i, then timer continues timing, and compare with the time threshold thd_T of the expression call duration upper limit, thd_T is 2s～5s, if timer＞thd_T, then by bus (8) to called host send call time, call out the disease room number, berth, the such call information of service that needs, and the prompting patient that sounds calls out success, then with the service zero clearing and stop timer;
1.4 visualization model according to the ruuning situation of impedance detection module, primary module, interface and control module, is responsible for calling out upward some information of demonstration of extension set (5), is specially:
1) when the impedance detection module is moved, position, title, the resistance value information of show electrode on cortex, and it is big or small with different sign show electrodes to press resistance value, if there is the bigger electrode of resistance value, the prompting user reexamines the situation of wearing of this electrode, till its resistance value drops to normally;
2) when primary module moves, show the EEG signals waveform that collects, filtered waveform, differentiate the result according to user's selection;
3) when the operation of interface and control module, shows current differentiation result, the time status of timer, text prompt to the situation of called host transmission information, after sending successfully.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200710036195A CN101159086B (en) | 2007-11-22 | 2007-11-22 | Calling device based on brain electric information detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200710036195A CN101159086B (en) | 2007-11-22 | 2007-11-22 | Calling device based on brain electric information detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101159086A CN101159086A (en) | 2008-04-09 |
| CN101159086B true CN101159086B (en) | 2010-05-19 |
Family
ID=39307164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200710036195A Expired - Fee Related CN101159086B (en) | 2007-11-22 | 2007-11-22 | Calling device based on brain electric information detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101159086B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101887307B (en) * | 2010-06-03 | 2011-12-07 | 西安交通大学 | Multi-frequency time sequence combined steady-stage visual evoked potential brain-computer interface method |
| KR101668249B1 (en) * | 2010-08-24 | 2016-10-21 | 엘지전자 주식회사 | Mobile terminal and operation control method thereof |
| CN102871657B (en) * | 2012-10-16 | 2014-04-02 | 中国人民解放军国防科学技术大学 | System and method for collecting electroencephalogram signal on basis of impedance self-adaption |
| CN103093085B (en) * | 2012-12-31 | 2016-01-20 | 清华大学 | Based on the analytical approach of the steady-state induced current potential of canonical correlation analysis |
| CN103607519A (en) * | 2013-10-17 | 2014-02-26 | 南昌大学 | Brain-computer interface-based telephone system for double-upper limb disabled people |
| US10394330B2 (en) | 2014-03-10 | 2019-08-27 | Qualcomm Incorporated | Devices and methods for facilitating wireless communications based on implicit user cues |
| CN105022488B (en) | 2015-07-20 | 2017-11-21 | 北京工业大学 | Wireless BCI input systems based on SSVEP brain electric potentials |
| CN106994689B (en) * | 2016-01-23 | 2020-07-28 | 鸿富锦精密工业(武汉)有限公司 | Intelligent robot system and method based on electroencephalogram signal control |
| CN108470182B (en) * | 2018-01-23 | 2022-04-29 | 天津大学 | Brain-computer interface method for enhancing and identifying asymmetric electroencephalogram characteristics |
| CN108388846B (en) * | 2018-02-05 | 2021-06-08 | 西安电子科技大学 | Electroencephalogram alpha wave detection and identification method based on canonical correlation analysis |
| CN108965625B (en) * | 2018-07-02 | 2020-07-31 | 昆明理工大学 | Automatic calling device based on brain-computer interface and training method thereof |
| CN110101953B (en) * | 2019-05-24 | 2021-07-30 | 易念科技（深圳）有限公司 | Psychological decompression method and equipment |
| CN111526632A (en) * | 2020-04-01 | 2020-08-11 | 成都千立网络科技有限公司 | Device and method for controlling flickering of network interface LED lamp |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2525600Y (en) * | 2002-01-18 | 2002-12-11 | 东北大学 | Environment controller for the disabled based on brain electric signal |
| CN1742673A (en) * | 2005-11-18 | 2006-03-08 | 高春平 | Electroencephalo safety monitoring prompting device |
| CN2770008Y (en) * | 2005-03-04 | 2006-04-05 | 香港理工大学 | Dozing detection alarm |
| CN100366215C (en) * | 1999-10-29 | 2008-02-06 | 清华大学 | Control method and system and sense organs test method and system based on electrical steady induced response |
-
2007
- 2007-11-22 CN CN200710036195A patent/CN101159086B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100366215C (en) * | 1999-10-29 | 2008-02-06 | 清华大学 | Control method and system and sense organs test method and system based on electrical steady induced response |
| CN2525600Y (en) * | 2002-01-18 | 2002-12-11 | 东北大学 | Environment controller for the disabled based on brain electric signal |
| CN2770008Y (en) * | 2005-03-04 | 2006-04-05 | 香港理工大学 | Dozing detection alarm |
| CN1742673A (en) * | 2005-11-18 | 2006-03-08 | 高春平 | Electroencephalo safety monitoring prompting device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101159086A (en) | 2008-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101159086B (en) | Calling device based on brain electric information detection | |
| CN103793058B (en) | A kind of active brain-computer interactive system Mental imagery classification of task method and device | |
| Rao et al. | Brain-computer interfacing [in the spotlight] | |
| CN110765920A (en) | Motor imagery classification method based on convolutional neural network | |
| CN103699216B (en) | A kind of based on Mental imagery and the E-mail communication system of vision attention mixing brain-computer interface and method | |
| CN206045144U (en) | A kind of novel intelligent sleeping and the device for waking up naturally | |
| CN105022488A (en) | Wireless BCI (Brain Computer Interface) input system based on SSVEP (Steady-State Visual Evoked Potentials) brain electric potential | |
| CN105942974A (en) | Sleep analysis method and system based on low frequency electroencephalogram | |
| CN110262658B (en) | Brain-computer interface character input system based on enhanced attention and implementation method | |
| CN102708288A (en) | Brain-computer interface based doctor-patient interaction method | |
| CN109582131A (en) | The asynchronous mixing brain-machine interface method of one kind and system | |
| CN106909226A (en) | A kind of polymorphic brain machine interface system | |
| CN106510702B (en) | The extraction of sense of hearing attention characteristics, identifying system and method based on Middle latency auditory evoked potential | |
| CN106708273B (en) | EOG-based switching device and switching key implementation method | |
| CN102799267A (en) | Multi-brain-computer interface method for three characteristics of SSVEP (Steady State Visual Evoked Potential), blocking and P300 | |
| CN206285117U (en) | Intelligence hearing terminal | |
| CN116400800B (en) | ALS patient human-computer interaction system and method based on brain-computer interface and artificial intelligence algorithm | |
| CN108304074A (en) | Display control method and related product | |
| CN108836327A (en) | Intelligent outlet terminal and EEG signal identification method based on brain-computer interface | |
| CN106344007A (en) | Monitoring system for lecture state | |
| CN113855052B (en) | Nerve feedback intervention system and method based on positive idea meditation | |
| CN101339413B (en) | Switching control method based on brain electric activity human face recognition specific wave | |
| CN112578682A (en) | Intelligent obstacle-assisting home system based on electro-oculogram control | |
| CN107802263A (en) | A kind of wearable sound feedback system based on Electroencephalo signal | |
| CN206563944U (en) | A kind of switching device based on EOG |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100519 Termination date: 20111122 |