CN107788976A - Sleep monitor system based on Amplitude integrated electroencephalogram - Google Patents
Sleep monitor system based on Amplitude integrated electroencephalogram Download PDFInfo
- Publication number
- CN107788976A CN107788976A CN201710864522.XA CN201710864522A CN107788976A CN 107788976 A CN107788976 A CN 107788976A CN 201710864522 A CN201710864522 A CN 201710864522A CN 107788976 A CN107788976 A CN 107788976A
- Authority
- CN
- China
- Prior art keywords
- sleep
- amplitude
- data
- module
- eeg signals
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/25—Bioelectric electrodes therefor
- A61B5/279—Bioelectric electrodes therefor specially adapted for particular uses
- A61B5/291—Bioelectric electrodes therefor specially adapted for particular uses for electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4806—Sleep evaluation
Abstract
The invention belongs to sleep monitor technical field, specially a kind of sleep monitor system based on Amplitude integrated electroencephalogram.The flexible electrode for encephalograms that present system is prepared using carbon sponge material obtains EEG signals, and by amplifying, filtering, analog-to-digital conversion and Bluetooth transmission, the EEG signals after processing are real-time transmitted into mobile device；On the mobile apparatus, pass through the conversion to EEG signals, obtain amplitude-integrated EEG signals, a variety of amplitude-integrated EEG signals characteristic informations are drawn in conjunction with the algorithm of feature extraction, and sleep stage classification is carried out to it by machine learning algorithm, so as to realize the analysis of sleep state and quality and monitoring；To finally obtained dormant data be analyzed and original EEG signals are dynamically uploaded in cloud platform in real time, tele-medicine auxiliary and clinical decision support are realized by knowledge reasoning machine.Present system is simple to operate, applied widely, is advantageous to user and understands the sleep quality of oneself in real time, reasonably adjusts itself sleep habit so as to reach the generation of prevention latent disease.
Description
Technical field
The invention belongs to sleep monitor technical field, and in particular to a kind of sleep monitor system based on Amplitude integrated electroencephalogram
System.
Background technology
People in life about 1/3 time spent in sleep, sleep is the necessary process of life, Ye Shiyi
Kind complicated physiology and action process.Good sleep is the basis of people's self-control and self-regeneration, but now more next
More people but endures the puzzlement of sleeping disorders to the fullest extent.Real-time sleep monitor, people can be helped reasonably to adjust itself sleep and practised
It is used to so as to reach the generation of prevention latent disease, while diagnosis and treatment sleeping disorders can also be aided in.
At present clinically, sleep monitor is mainly by leading hypnotic instrument (the Iber et al, 2007) that realizes more.It is more
Hypnotic instrument monitoring is led, is " goldstandard " at present clinically, it can monitor a variety of physiological signals, such as electroencephalogram, electromyogram, the heart
Electrograph, electromyogram etc..But lead hypnotic instrument installation and disassembling section complex steps, expensive more；For the physiology monitored
The deciphering of signal, professional requirement is high, it is big to understand difficulty, solution read time is tediously long.Therefore, result in hypnotic instrument is led so far not having also more
It is widely used.
In addition to leading hypnotic instrument, different sleep monitor systems is suggested in succession more.For example, it is based on Air Cushion
Sleep sublevel system (Watanabe et al, 2004), its by pressure sensing mat monitor heart rate, respiratory rate, snoring and
Body movement signal, realize sleep sublevel；Sleep sublevel system (Kortelainen et al, 2010) based on Bed Sensor, its
Sleep sublevel is realized by heart rate and body movement signal；Sleep based on Biomotion Sensor/awakening sleep monitor system (De
Chazal et al, 2011), it passes through the monitoring that the body movement signal of biomotion Sensor monitoring people realizes sleep and awakening；
Sleep Hunter (Gu et al, 2016) propose using the Sensor monitoring body built in smart mobile phone is dynamic and voice signal from
And realize sleep sublevel.But current sleep monitor systemic-function is single, majority is limited only to sublevel of sleeping, the essence of system
Degree and the degree of accuracy have to be hoisted, and real-time is inadequate, and the high in the clouds of data is stored it is expected to obtain available for clinical monitoring system
System etc. does not relate to.
It is short for sleeping problems and sleep monitor systematic research time in China, some shortcomings are still suffered from present：1) sleep
The popularity of dormancy monitoring instrument is low.2) in the independent research starting evening of sleep monitor instrument, research and development ability has to be hoisted.At present,
Most sleep monitor instruments are needed from off-shore purchase and introduction.3) sleep monitor instrument body product is huge, lacks portability.4)
Sleep monitor instrument function is single.At present, most sleep monitor instruments only provide data acquisition, and the analysis for data stills need
Doctor artificially understands, and lacks automation, can not in real time monitor which results in most sleep monitor instruments and be slept with feedback user
Dormancy state and quality, lack real-time.
Except leading hypnotic instrument and different sleep monitor systems more, in recent years, Amplitude integrated electroencephalogram in clinic by
Increasing concern.Amplitude integrated electroencephalogram is the form after a kind of continuous recording simplification of electroencephalogram, its is simple to operate,
Can non-volatile recording, become vast neonate department, Neurology etc. be used for brain function situation objective evaluation standard.At present,
Amplitude integrated electroencephalogram equipment mainly provides the display of amplitude-integrated EEG signals, is not directed to signal analysis.Moreover, amplitude is whole
Syncerebrum electrograph is mainly used in the detection fainted from fear, and realizes that automatic, real-time sleep state and quality are commented using Amplitude integrated electroencephalogram
Estimate and still need to be explored and develop.
In summary, current sleep monitor system, it is cumbersome, automaticity is not high, it is impossible to realize that dynamic is real
When monitor.Real-time sleep monitor system proposed by the present invention based on Amplitude integrated electroencephalogram still belongs to blank in the field.
Bibliography：
De Chazal,P.,Fox,N.,O’HARE,E.M.E.R.,Heneghan,C.,Zaffaroni,A.,Boyle,
P.,...&McNicholas,W.T.(2011).Sleep/wake measurement using a non‐contact
biomotion sensor.Journal of sleep research,20(2),356-366.
Gu,W.,Shangguan,L.,Yang,Z.,&Liu,Y.(2016).Sleep hunter:Towards fine
grained sleep stage tracking with smartphones.IEEE Transactions on Mobile
Computing,15(6),1514-1527.
Iber,C.,Ancoli-Israel,S.,Chesson,A.,&Quan,S.F.(2007).The AASM manual
for the scoring of sleep and associated events:rules,terminology and
technical specifications(Vol.1).Westchester,IL:American Academy of Sleep
Medicine.
Kortelainen,J.M.,Mendez,M.O.,Bianchi,A.M.,Matteucci,M.,&Cerutti,S.
(2010).Sleep staging based on signals acquired through bed sensor.IEEE
Transactions on Information Technology in Biomedicine,14(3),776-785.
Watanabe,T.,&Watanabe,K.(2004).Noncontact method for sleep stage
estimation.IEEE Transactions on biomedical engineering,51(10),1735-1748.。
The content of the invention
In view of current sleep monitor system, such as leads hypnotic instrument more and installation and disassembling section complex steps, equipment body be present
Product is huge, expensive, and for the deciphering of the physiological signal monitored, professional requirement is high, it is big to understand difficulty, solution read time is superfluous
It is long；Other sleep monitor systemic-functions are single, most only to provide data acquisition and display, and the analysis for data lacks automatic
Change and real-time.The present invention proposes the sleep monitor system based on Amplitude integrated electroencephalogram, with overcome the deficiencies in the prior art.
Sleep monitor system proposed by the present invention based on Amplitude integrated electroencephalogram, has merged multi-disciplinary technology, including
Electronic information technology, advanced material technology, microelectronics, computer technology, signal transacting and clinical sleep analysis etc..This is
System obtains EEG signals first with the novel flexible electrode for encephalograms of carbon sponge material preparation, by amplifying, filtering, modulus turn
Change and Bluetooth transmission, the EEG signals after processing are real-time transmitted to mobile device；On the mobile apparatus, by EEG signals
Conversion, obtain amplitude-integrated EEG signals, a variety of amplitude-integrated EEG signals features drawn in conjunction with the algorithm of feature extraction
Information, and sleep stage classification is carried out to it by machine learning algorithm, so as to realize the analysis of sleep state and quality and prison
Survey；Final system will analyze obtained dormant data and original EEG signals are dynamically uploaded in cloud platform in real time, passes through knowledge
Inference machine realizes tele-medicine auxiliary and clinical decision support.
The present invention provides a kind of non-intrusion type, accurate, portable, safe sleep monitor system；It can realize for a long time, even
Continuous monitoring, while can also reflect the dormant change situation of people in time.
Complete, easy-operating, portable sleep monitor proposed by the present invention, it is to be based on amplitude-integrated EEG signals
, data acquisition, data analysis from EEG signals can realize " sleep state and quality evaluation " to data storage, " move in real time
State monitoring ", " tele-medicine ", " individual character medical treatment " etc..Meanwhile the real-time monitoring of the system also can provide dynamic number for subsequent medical
According to so as to provide foundation for the formulation of therapeutic scheme.The invention is simple to operate, applied widely, is advantageous to user and understands in real time
The sleep quality of oneself, itself sleep habit is reasonably adjusted so as to reach the generation of prevention latent disease；It simultaneously can be used for
Neonatal intensive care unit, neonatal sleep info is provided in real time for doctor and nurse, it is newborn so as to aid in doctor to explore
The Sleep architecture of youngster, is advantageous to doctor and nurse takes suitable therapeutic scheme, Tending and care program, helps neonate
Early stage establishes ripe circadian sleep cyclic pattern, improves neonatal sleep, to promote neonatal development.
Sleep monitor system provided by the invention based on Amplitude integrated electroencephalogram, it is mainly comprising hardware system, software
Three parts of analysis system and cloud platform, see Fig. 1 system schematics.
1st, hardware system
Hardware system is used to being acquired the EEG signals of user, amplifies, filters, analog-to-digital conversion, signal store and letter
Number send, hardware system is as shown in Fig. 2 it includes：Microcontroller, power supply module, signal acquisition module, signal processing module,
Impedance detection circuit, signal transmitting module, clock module, memory module etc..Wherein, signal acquisition module is to be placed in Double Tops bone
The a pair of electrodes at place, for gathering EEG signals, electrode position is according to the P3- in international standard 10-20 electrode placement systems
P4 leads, while Fz ground connection in the middle part of forehead, electrode for encephalograms take the mode of the dry electrode of condenser type.Signal processing module includes putting
Greatly, filtering and A/D converter circuit, EEG signals are by the amplification and filtering of signal processing module, then through analog-digital converter (A/
D converters) analog-to-digital conversion is carried out to analog signal, data signal is obtained, is sent into microcontroller；The numeral that microcontroller will obtain
Signal is put into memory module, and data signal is transmitted to mobile terminal (electricity signal transmitting module in a manner of being wirelessly transferred
Brain, mobile phone, tablet personal computer etc.).Impedance detection circuit is used for measuring electrode and scalp contact impedance size, feeds back to operator
Member.Impedance detection circuit can be realized to being gathered while EEG signals and contact impedance signal in monitoring process, be doctor
The exclusion noise of analysis EEG signals provides important evidence in real time, and artefact is removed when also being analyzed for EEG signal and provides foundation.Power supply
Module provides power supply for microcontroller, and clock module is that microcontroller sets clock.
2nd, software systems
Software systems to the EEG signals of user for being received, real-time display, while EEG signals are converted to and shaken
Width integrates EEG signals, then carries out feature extraction, then using grader, the sleep to user to amplitude-integrated EEG signals
Stage is assessed, and finally realizes that sleep state and quality are assessed using statistical calculations.Software system structure such as Fig. 8 institutes
Show.It mainly includes：User log-in block, data reception module, original eeg data display module, Data Format Transform mould
Block, data analysis module, result display module and the inquiry of historical data module.Wherein：
The user log-in block, for logging in system by user；Old user inputs username and password and can be used this soft
Part module, new user need that the software module just can be used after registration is complete.User enters after the software module, data
Receiving module just starts to start, and system immediately enters data receiver pattern；Data reception module mainly utilizes the indigo plant of mobile device
Tooth realizes data receiver.The data that system receives directly can be shown in the aobvious of mobile device using original eeg data display module
Show on device.Meanwhile the original eeg data received is changed into amplitude-integrated brain electricity by the data format conversion module in system
After figure, sleep quality assessment is carried out to converted amplitude-integrated eeg data using data analysis module, and related
Assessment result is sent to result display module and shown.In addition, sleep quality assessment result is transmitted and is stored in history number
According in enquiry module, it is easy to user to realize the inquiry of historical data.
3rd, cloud platform
Cloud platform is used for the real-time upload for realizing data, by original eeg data, the data (sleep state of mobile terminal analysis
And sleep quality) high in the clouds is uploaded in real time, so as to realize slept to user tracking and sleep associated with disease prediction.Meanwhile
Support that remotely data are called in storage to multiple client online, support tele-medicine auxiliary and clinical decision support, it is mainly realized
Framework is as shown in Figure 10, mainly includes：Cloud platform data memory module, medical knowledge library module and tele-medicine auxiliary and clinic
Support module.Cloud platform data memory module is mainly used in original eeg data and the mobile terminal analysis that storage uploads to cloud platform
Sleep state and sleep quality data；Medical knowledge library module is mainly included required for problem solving (clinical decision reasoning)
Knowledge, it includes expertise, clinical knowledge, health knowledge and comprehensive knowledge etc.；Tele-medicine aids in and clinical support module
It is responsible for utilizing the data in data memory module and the knowledge in integrative medicine knowledge base goes to solve and reasoning sleep is related asks
Topic, such as sleep associated with disease etc., so as to provide auxiliary and support for tele-medicine and clinical decision.It is implemented as follows：Cloud
Related data and relevant knowledge are sent to inference machine by platform data memory module and medical knowledge library module respectively, and inference machine is born
Knowledge reasoning in data and knowledge base that duty is analyzed using brain electricity initial data, mobile terminal goes out result, for example, user whether
With related sleeping disorders, the recent sleep quality of user, and its reasoning results is sent to result composite part, then result
Evaluation portion is then assessed the result of result composite part transmission, and result is transferred to result feedback if result is correct
Part simultaneously exports, if the result of reasoning generation is incorrect, result is retransferred to inference machine, inference machine and then re-started
Reasoning, until the result of evaluation of result part evaluation generation is correct.
The features of the present invention：
1. the present invention with novel flexible carbon sponge material replaces traditional wet electrode, using carbon sponge self-conductive performance,
The characteristics such as mechanical performance (pliability, compressibility etc.), design and research and development meet ergonomics novel flexible electrode for encephalograms and
Eeg collection system, so as to obtain stable, sensitive, long-term EEG signals；
2. the present invention proposes a new sleep monitor system, wherein not only including information acquisition system, Ke Yishi
Existing long-time collection easy to operate, portable, non-intrusion type EEG signals；It further comprises real-time signal processing and analyzing simultaneously
System, it is possible to achieve the real-time processing of EEG signals and sleep state and quality evaluation；
3. the method for sleep state and quality evaluation proposed by the present invention based on Amplitude integrated electroencephalogram, it is possible to achieve right
The sleep sublevel of user, sleep quality such as assess at the function.Present invention also offers real-time electroencephalogram, amplitude-integrated brain electricity simultaneously
The simultaneous display function of figure, sleep state and quality.While sleep state is shown, electroencephalogram is shown, analysis result is done
Traced back to active；
4. invention introduces the theory of cloud platform long distance monitoring, by the EEG signals of user and sleep state and quality
Assess data and upload to high in the clouds in real time, so as to realize the tracking to the sleep state of user, while can by knowledge reasoning machine,
Tele-medicine auxiliary and personalized clinical decision support are provided for doctor.
The technology of the present invention effect
1. hardware components
(1) easy to operate, portable, non-intrusion type electro-physiological signals collections
The hardware system of the present invention mainly includes three carbon sponge electrodes, electrode cap, signal processing circuit module.The present invention
Middle pith is the research and development of electrode for encephalograms and the design of electrode cap.Traditional electrode for encephalograms uses wet electrode, and it is primarily present
Following shortcoming：(a) scalp impedance in order to reduce between brain scalp and electrode be present and obtain the preferable signal of quality, adopting
Need to carry out grinding process to the scalp of human body before collection signal and smear Signa Gel, therefore certain customers cause to feel sometimes
Dye；(b) contact scalp impedance of traditional wet electrode with skin deteriorates with the extension of acquisition time, and it is not suitable for for a long time
Sleep monitor；(c) easily there is the situation of electrode delamination or absent-mindedness in traditional wet electrode under sleep environment, so as to cause letter
Number disappear or degradation；(d) traditional brain electric system needs the assistance of medical professional to operate, and operation is more multiple
Miscellaneous, human cost is higher；
Therefore, the present invention proposes the dry electrode of a novel flexible.Electrode uses the carbon sponge with self supporting structure to make
For new flexible electrode material, it has self-supporting stereochemical structure, it is not necessary to is attached to flexible substrate；Meanwhile carbon sponge has
There is continuous tridimensional network, continuous structure can guarantee that electronics transmits rapidly；In addition, carbon sponge shows excellent machinery
Performance, it is flexible, it is compressible, it can closely be attached to any surface contacted.The dry electrode of novel flexible carbon sponge is not required to
Will to scalp carry out grinding process or smear Signa Gel, its have non-intrusion type, softness, it is comfortable the characteristics of, will not be to human body
The advantages that causing infection；
To ensure carbon sponge the electrode stability that uses and sensitivity for a long time, the present invention, which have also been devised, a meets human body
The electrode cap of engineering, for aiding in fixed electrode, and it is more applicable for measuring for a long time, meets the actual need of sleep monitor
Ask.Brain electricity cap is using comfortable, good permeability, easy cleaned two-sided knitting venetian cloth.Signal processing circuit module integrates
The module such as impedance detection, signal amplification, filtering, A/D conversions, processor, Bluetooth transmission.Electrode position is according to international standard
10-20 electrode placement systems, P3, P4, tri- positions of Fz are placed on, wherein P3, P4 are active electrode, and Fz is indifferent electrode (ginseng
Examine electrode)；
(2) guarded compared to conventional sleep, the lead number required for the present invention is less, and its is simple to operate, it is not necessary to clinical skill
Art personnel assist, and can greatly save manpower, make it possible Home Tele-monitoring System；The equipment has portable spy simultaneously
Point, suitable for being used under various measuring environments.
2. software section
(1) multi-user monitors
The present invention supports multi-user to monitor simultaneously, and software can receive the sleep cerebral electricity data of multiple users simultaneously.Also simultaneously
Monitored user can be managed, monitored user is added, deletes monitored user, checks monitored user's basic status etc.；
(2) wisdom of sleep state and quality is assessed
The present invention can provide the real-time assessment of sleep state and quality based on Amplitude integrated electroencephalogram, and it can be exactly
Realize the sleep stage assessment of user, the functions such as duration, sleep quality assessment of sleeping；
(3) dynamic monitoring
The present invention supports cloud platform data storage, and the correlation that sleep quality is tracked prediction sleep can be carried out to user
Disease, while be alternatively doctor and tele-medicine auxiliary and personalized clinical decision support are provided.
The new product that the present invention is formed, the following aspects will be embodied in the value of sleep monitor, intelligent medical treatment：
(1) data acquisition, data analysis, the integration of data storage of physiological signal, and the number of medical act are realized
Word, intellectuality, personalization.Product can be used for family, medical care both of which simultaneously, and doctor can monitor the letter of patient in real time
Breath, user can also understand the sleep state of oneself in real time by system；
(2) existing most sleep monitor systems only provide physiological signal collection function, and data analysis stills need medical practitioner
Artificial deciphering.It, which does not require nothing more than medical personnel, the ability and experience of stronger visual observations, while can also increase medical care work
Make the burden of personnel, influence the efficiency judged.Automated sleep analysis in the present invention, the work of medical personnel can be mitigated
Amount can also provide objective assessment simultaneously；
(3) cloud platform in the present invention, can the EEG signals of upload user and sleep evaluation result in real time, it can be square
Just medical personnel (or user) carry out prolonged sleep monitor to user and provide the prediction of sleep associated with disease, while also may be used
Support is provided with the clinical decision for medical personnel, promotes social sustainable development.
Brief description of the drawings
Fig. 1-system schematic.
Fig. 2-hardware module and functional diagram.
Fig. 3-brain electricity cap schematic diagram.
Fig. 4-electrode for encephalograms schematic diagram.
Fig. 5-electrode for encephalograms base schematic diagram.
Fig. 6-brain electricity cap and electrode for encephalograms pictorial diagram.
Fig. 7-electrode for encephalograms performance test comparing result figure.
Fig. 8-software module and functional diagram.
Fig. 9-data analysis algorithm flow chart.
Figure 10-clinical decision figure.
Label in figure：1- brain electricity cap overall schematics；2- electrode P4 riding positions；3- electrode P3 riding positions；4- electrodes
Fz riding position brain electricity caps；5- electrode for encephalograms overall schematics；6- mother-son mouth shape pin thread conductive metal sheets；7- conductive silver pastes；8-
Carbon sponge conductive electrode material；9- mother-son mouth shape box conductive metal sheets；10- conducting wires.
Embodiment
Hardware components
(1) flexible carbon sponge electrode for encephalograms and electrode cap
The present invention can carry out sleep monitor lasting for a long time, and hardware system is mainly used in eeg signal acquisition, whole hardware
System includes electrode, electrode cap, signal processing module, bluetooth module, processor etc..
What electrode part was taken is the flexible dry electrode of carbon sponge.The fine or not EEG signals quality to collecting of electrode design
Influence is vital, therefore preferably electrode design is the key component of hardware acquisition system.Electrode design mainly needs
The factor of consideration has brain scalp and interelectrode impedance magnitude, comfortableness, if is adapted to use for a long time, if safety etc..
The electrode system that the present invention designs is detachable electrode, and electrode structure is as shown in figure 4, electrodes base is as shown in Figure 5.Including：Son
Female mouth shape pin thread conductive metal sheet 6, conductive silver paste 7, carbon sponge conductive electrode material 8, mother-son mouth shape box conductive metal sheet 9 and
Conducting wire；Wherein, the button structure 5 that mother-son mouth shape pin thread conductive metal sheet 6 forms with mother-son mouth shape box conductive metal sheet 9,
Carbon sponge material 8 is connected by conductive silver paste 7 with mother-son mouth shape pin thread conductive metal sheet 6, and is further handled with advance signal
Circuit is connected, and can be very good to draw signal by the button structure, ensures signal quality simultaneously；The button structure makes simultaneously
Electrode can be dismantled easily, be easy to disposable, it is extra large that carbon contact with human body scalp need to be only changed before each use
Continuous material.Wherein, carbon sponge material 8 is the part being in contact with human body scalp, and such carbon sponge conductive material has big
Small size can adjust, be fixed easily, be soft, be safe and non-toxic, current potential is stable, it is small to human body excitant, to hair destroy less, cost it is low
The features such as.By our preliminary test, the carbon sponge material that the present invention uses has good electric conductivity, suitable for brain telecommunications
Number collection；The carbon sponge material has flexible well simultaneously, will not bring discomfort with human contact, be surveyed suitable for long-time
Amount；The carbon sponge material is pore structure, can be very good to be adapted with human hair, small to hair destructiveness, and is reduced
The interference that hair is brought to signal acquisition；Take the mode of dry electrode, it is not necessary to human body scalp is polished, therefore not deposited
Infection the problem of；Using softness the dry electrode of carbon sponge, have one it is relatively difficult the problem of be how electrode preferably
Connected with external circuit, so the present invention devises one by mother-son mouth shape pin thread conductive metal sheet 6, mother-son mouth shape box
The button structure 5 that conductive metal sheet 9 forms.
Electrode takes the mode of bipolar lead and preamplifier to be attached, because between two float electrodes P3, P4
It is distant, be adapted to access external circuit using the mode of bipolar lead, compared to unipolar lead, strong antijamming capability.This
The carbon sponge electrode used in invention is used to gather EEG signals and done performance test, by the electrode and traditional electrode for encephalograms simultaneously
For gathering the EEG signals of approximately the same position, resulting result is contrasted, its contrast test result such as Fig. 7 institutes
Show, can intuitively find out from result, the signal that two kinds of electrodes collect is similar, and the electrode material of this explanation present invention can be used for
Eeg signal acquisition.
In order to fix three electrode for encephalograms well, the present invention devises a brain electricity electricity for being adapted to collection lead and setting
Polar cap 1, its structural representation is as shown in figure 3, two float electrodes P3, P4 and reference electrode Fz are arranged on electrode for encephalograms cap
The correspondence position at top；The material of electrode for encephalograms cap can be very good to be adapted to using the two-sided knitting venetian cloth of elasticity
Different brain sizes are to obtain preferable signal quality and wear comfort.It can be very good to fix brain electricity electricity by the design
Pole, prevent in sleep procedure, due to electrode delamination or the blackout brought and ropy problem are relaxed, so as to protect
Card measures for a long time；Electrode for encephalograms and brain electricity cap material object display from all angles as shown in fig. 6, illustrate brain wave acquisition dress respectively
Put.
(2) hardware circuit
System signal processing hardware circuit entire block diagram is as shown in Fig. 2 main include amplifying to EEG signals for routine, filter
The processing such as ripple, AD conversion, Bluetooth transmission and impedance detection etc..
Wherein, preamplifier takes AD620, and it is a low cost, high-precision meter amplifier, it is only necessary to which one outer
Portion resistance sets gain.Including trapper, it is mainly used in filtering out 50Hz Hz noises, filter circuit uses anti-aliasing filter
Ripple device, for filtering out high-frequency noise, its cut-off frequency is arranged to 1000Hz.The device that wave filter in the present invention is chosen is
UAF42, UAF42 are a integrated filter chips, the shadow of the characteristic frequency and quality factor of this wave filter by outer meeting resistance
Ring very little.Requirements of the UAF42 to external discrete component is relatively low, contains 1000pf inside it, and precision is 5% electric capacity, outside nothing
Electric capacity need to be connect, only need to connect 0.1% resistance, UAF42 only needs external several resistance to be easily designed as high pass, low pass etc.
Wave filter.Trapper can be made up of high-pass filter and low pass filter summation.What digital-to-analogue dress mold changing block was chosen is 24 height
Performance A/D converter ADS1278, there is very high conversion accuracy, such device can realize 8 road signal synchronous collections, for not
Carry out passage extension to provide convenience.
Present invention additionally comprises an impedance detection circuit, for measuring electrode and scalp contact impedance size, feeds back to behaviour
Make personnel.When electrode and the insufficient contact of scalp will cause contact impedance higher, it is poor to will result in signal quality, meeting simultaneously
The mismatch of impedance between each lead is caused so as to introduce noise.In addition to gathering EEG signals, while measurement scalp impedance in real time
Signal, measuring electrode and scalp contact impedance size, and Real-time Feedback is to medical worker.Although many conventional electroencephalographs
Human body scalp and interelectrode scalp impedance can be measured before measurement starts, but the impedance can not be entered in gatherer process
Line trace measures, and the impedance detection circuit of the present invention can be realized in monitoring process to EEG signals and contact impedance signal
While gather, the exclusion noise for analyzing EEG signals in real time for doctor provides important evidence, is removed when also being analyzed for EEG signal
Artefact provides foundation.
In view of the power consumption of hardware system, and the comfortableness of whole hardware system, portability, the system use low-power consumption
Microprocessor MSP430, it is wirelessly transferred part and uses the low power consumption protocol of bluetooth 4.0, it is convenient so as to reduce hardware system overall power
Realize that extended sleep monitors.MSP430 series monolithics are that TI companies started the super low-power consumption of release in 1996, have and simplify
The 16bit mixed-signal processors of instruction set (RISC).Compared with other single-chip microcomputers, the series monolithic have super low-power consumption,
The advantages such as On-Chip peripheral is abundant.The system is from MSP430F5529 as control chip.In MSP430 family chips,
MSP430F5529 can not only meet functional requirement, and overall power consumption and cost are also relatively low.Bluetooth module is using business
Industry bluetooth module CC2564, CC2564 are the dual mode bluetooth chip of TI exploitations, while support bluetooth 3.0 and bluetooth 4.0
Transmission, equipment compatibility is good, is easy to establish connection with equipment such as computer, mobile phone, tablet personal computers, bluetooth module transfers data to
PC is stored behind end, the subsequent operation such as display operation and signal transacting.
Software section
The software function and module of system, as shown in figure 8, mainly including data receiver, the display of original brain electricity, data format
Conversion (data conversion of original brain electricity to amplitude-integrated EEG signals), data analysis are (based on sleeping for amplitude-integrated EEG signals
Dormancy state and quality evaluation), result show (sleep state and quality results are shown) and the inquiry of historical data etc., it is main to include using
Family login interface, data display interface, data analysis interface, the inquiry of historical data interface.The data point of specific software section
Algorithm flow chart is analysed, as shown in figure 9, original brain electricity therein is to the data conversion of amplitude-integrated EEG signals and based on amplitude
The sleep state of integration EEG signals and the method for quality evaluation are as follows：
(1) data conversion of the EEG signals to amplitude-integrated EEG signals
The data conversion of EEG signals to amplitude-integrated EEG signals mainly includes following four part：Filtering, end points carry
Take, the compression of amplitude compression and time.
Filtering：The main purpose of filtering be eliminate because sweating, muscle activity, artefact caused by electrical interference etc., while make not
Decay of the brain electricity of same frequency composition in transmitting procedure is accordingly compensated.2-15Hz's is carried out to the EEG signals received
Bandpass filtering, while design high performance finite impulse response digital filter and realize that the transmission of the non-rhythm and pace of moving things composition of EEG signals declines
The compensation subtracted.
End points extracts：The main purpose of end points extraction is by the fluctuating range reflection of EEG signals to amplitude-integrated brain electricity
On waveform.EEG signals are divided into segment, extracted per a bit of peak-to-peak value (maximum and minimum value), as amplitude
Integrate the upper extreme point and lower extreme point of the corresponding vertical line of brain electricity.
Amplitude compression：Amplitude compression main purpose is to reduce original EEG signals by the dynamic range of changes in amplitude while dash forward
Show the part of the low wave amplitude of brain electricity.If the peak-to-peak value (the vertical line endpoint value extracted) of original brain electricity is less than 6 μ V, erect
Line is still shown, for then carrying out log-compressed more than or equal to 6 μ V parts without log-compressed by linear relationship.
Time is compressed：Time compression main purpose is that prolonged EEG signals are reflected in very short amplitude-integrated brain
In the wave spectrum band of electricity, in order to which medical personnel observe patient on the whole.Each vertical line is corresponded to the original brain electricity of segment, it is real
Compressed between current.
(2) sleep state and quality evaluation based on amplitude-integrated EEG signals
Sleep state and quality evaluation based on amplitude-integrated EEG signals, mainly amplitude-integrated EEG signals are carried out
Analysis, sleep sublevel is realized, the sleep simultaneously for a period of time provides time for falling asleep, recovery time, always sleep duration, difference
Sleep stage distinguishes duration.It mainly includes pretreatment, feature extraction, the grader classification of amplitude-integrated EEG signals, result
Count part.
The pretreatment of amplitude-integrated EEG signals：It is mainly used for amplitude-integrated EEG signals with analyzing with sentencing
It is disconnected.It is abnormal to judge that it is belonging respectively to normal amplitude, amplitude mile abnormality and amplitude severe according to signal amplitude.
Feature extraction：Mainly to amplitude, normally amplitude-integrated EEG signals carry out characteristics extraction for it.
(1) aEEG essential characteristic extraction
AEEG essential characteristic extraction, the main maximum included to original aEEG signals, minimum amplitude, amplitude average are high
In 10uV scope amplitude ratios, less than 5uV scope amplitude ratios.
(2) aEEG morphological feature extraction
AEEG morphological feature extraction, main include are asked for the upper lower envelope of original aEEG signals.
(3) aEEG Nonlinear feature extraction
Nonlinear feature extraction, mainly include the extraction of the features such as correlation dimension, Lyapunov indexes and approximate entropy.
Correlation dimension (Correlation Dimension)
Correlation dimension mainly calculates the relevance before and after variable using correlation integral, is established rules really with this to describe signal
Rule and its degree.Uncertainty in signal is more, and correlation dimension is bigger；Certainty composition in signal is more, association
Dimension is smaller.Therefore, correlation dimension can be used for quantitative description aEEG nonlinear characteristic.
Calculate correlation dimension：
Comprise the following steps that：Assuming that initial data is { u (1), u (2) ..., u (N) } N number of point altogether, setting Embedded dimensions are
M, time delay τ, the phase space X reconstructed under the parameter can be obtainedi=(ui,ui+τ,...,ui+(m-1)τ), wherein i=1,
2,...,N-(m-1)τ.τ m dimensional vector of N- (m-1) is obtained after phase space reconfiguration.To the phase space vector X of reconstructi, calculate
Correlation integralH is Heaviside functions in formula, and it is defined asI, j represent vector numbers, and r represents distance between two vectors, | | Xi-Xj| | represent two phase point XiAnd Xj's
Distance.C (r) represents probability of the distance between two points less than r in phase space.Finally correlation dimension is
Lyapunov indexes
The size of Lyapunov indexes shows the index percent of close orbit averaging convergence or diverging in phase space.It is if maximum
Lyapunov indexes is just, then system have sensitivity to initial state, and its motion is chaos；If maximum Lyapunov indexes are
Zero, show that system is insensitive to initial value, periodic motion is presented；If maximum Lyapunov exponent be less than zero, system it is long-term
Behavior is unrelated with initial value, will converge to an equalization point.
Comprise the following steps that：Phase space reconstruction, find each point X on given trackiClosest point X'iInitial distance,
I.e.Wherein di(0) it is the i-th point of initial distance for arriving its nearest-neighbor.According to maximum
The definition of Lyapunov indexes, its value should be the average value of nearest-neighbor diverging speed, so there is di(k)=Cieλ1(kΔt), wherein
CiFor initial distance, di(k) it is i-th point of distance with its nearest-neighbor after k chronomere.Above-mentioned equation is taken the logarithm can
Obtain lndi(k)=lnCi+λ1(k Δ t), maximum Lyapunov exponent can then be obtained by least square fitting.
Approximate entropy (Approximate Entropy, ApEn)
Approximate entropy is a kind of complexity of metric sequence and the nonlinear kinetics parameter of statistic quantification, and it is non-negative with one
Number represents the complexity of a time series, and approximate entropy corresponding to more complicated time series is bigger.
Calculate approximate entropy：Wherein
Comprise the following steps that：Assuming that initial data is { u (1), u (2) ..., u (N) } N number of point altogether, it is continuous by sequence number
Order forms one group of m n dimensional vector n (from X (1) to X (N-m+1)), Xi=u (i), u (i+1) ..., and u (i+m-1) }, calculate any
Vectorial XiAnd XjThe distance between (j=1,2 ..., N-m+1, j ≠ i) dij=max | u (i+j)-u (j+k) |, k=0,1 ...,
The maximum of absolute difference is exactly the distance between two vectors between m-1, i.e. two vectorial corresponding elements.Then threshold is given
Value r, to each vectorial Xi, count dijLess than r number and this number and the ratio apart from total N-m+1, it is designated asWillNatural logrithm is taken, its average value to all i is then sought again, is designated as φm(r).Increase dimension, repeat the above steps,
φ can be obtainedm+1(r), so as to obtaining approximate entropy ApEn.
Grader is classified：It mainly classifies to the characteristic value of extraction, and sleep is divided into awakening, shallow slept and sound sleep three
The individual different stage.
SVMs (Support Vector Machine, SVM) is a kind of Novel machine based on Statistical Learning Theory
Device learning method, it can preferably solve the practical problem such as small sample, non-linear, it has also become the heat of technical field of intelligence research
Point, the numerous areas such as fault diagnosis, pattern-recognition are widely used at present.SVMs is by optimal during linear separability
Optimal Separating Hyperplane continues to develop what is come.For the classification problem in the case of Nonlinear Classification, the general thought of SVMs is,
Input space data are mapped to first with a nonlinear transformation characteristic vector space of a higher-dimension, then in this feature
Construct optimal separating hyper plane in space, carry out linear classification, finally map back to behind former space just into the input space
Nonlinear Classification.Classical algorithm of support vector machine only gives the algorithm of two classification, (awakens, shallow for sleep sublevel
Sleep and sound sleep) three classification problems, the present invention in employ one-to-one temporal voting strategy (one against one).By training set
The class of sample two, form to two classes 3 training sets, i.e., (awakening, shallow to sleep), (awakening, sound sleep) and (shallow to sleep, sound sleep) is trained and obtained
3 graders of SVM bis-：SVM1, SVM2 and SVM3.When test, test sample is sequentially sent to these three two graders,
The form of ballot is taken to obtain final classification results.Specific voting process is as follows：
Initialization：
Vote (awakening)=vote (shallow to sleep)=vote (sound sleep)=0
Voting process：
If the grader SVM1 obtained using training set (awakening, shallow to sleep), test sample is judged to awakening, then vote
(awakening)=vote (awakening)+1；Otherwise vote (shallow to sleep)=vote (shallow to sleep)+1；
If the grader SVM2 obtained using training set (awakening, sound sleep), test sample is judged to awakening, then vote
(awakening)=vote (awakening)+1；Otherwise vote (sound sleep)=vote (sound sleep)+1；
If the grader SVM3 obtained using training set (shallow to sleep, sound sleep), by test sample be determined as it is shallow sleep, then vote
(shallow to sleep)=vote (shallow to sleep)+1；Otherwise vote (sound sleep)=vote (sound sleep)+1；
Court verdict：
Max (vote (awakening), vote (shallow to sleep), vote (sound sleep)), the maximum for taking ballot is court verdict.
As a result count：It mainly to the full assessment of sleep stage, provides time for falling asleep, recovery time, always slept respectively
Dormancy duration, different sleep stages difference duration and Sleep efficiency.
Time on the Sleep efficiency=length of one's sleep/bed.
(3) cloud platform
The change of people's sleep state, often it is a kind of it is trickle, be not easy the process discovered, catch sleep state transformation, and
When to give related feedback information particularly important.The present invention is utilizing cloud platform, it is possible to achieve the record and prison of extended sleep information
Survey, in conjunction with knowledge reasoning machine, the transformation of the sleep quality, sleep state that can timely and effectively feed back and relevant disease it is pre-
Survey.The present invention is combined in cloud platform using sleep info, EEG signals with medical knowledge, utilizes different inference methods
(it is determined that, uncertain, fuzzy etc. different inference methods), obtains corresponding the reasoning results, then by main reference machine by whole reasoning
As a result comprehensive analysis is carried out, and the reasoning results are fed back to doctor so as to realize tele-medicine auxiliary and clinical decision support, or
The reasoning results are fed back to user so as to realize the autonomous sleep regulation of user by person.
Claims (7)
1. a kind of sleep monitor system based on Amplitude integrated electroencephalogram, it is characterised in that include hardware system, software analysis system
System and three parts of cloud platform；Wherein：
Hardware system is used to being acquired the EEG signals of user, amplifies, filters, the storage of analog-to-digital conversion, signal and signal are sent out
Send, it includes：Microcontroller, power supply module, signal acquisition module, signal processing module, impedance detection circuit, signal send mould
Block, clock module, memory module；Wherein, signal acquisition module is to be placed in a pair of electrodes at Double Tops bone, for gathering brain electricity
Signal；Signal processing module includes amplification, filtering and A/D converter circuit, the amplification that EEG signals pass through signal processing module
And filtering, then analog-to-digital conversion is carried out to analog signal through analog-digital converter, data signal is obtained, is sent into microcontroller；Microcontroller
Obtained data signal is put into memory module by device, and data signal is passed in a manner of being wirelessly transferred signal transmitting module
Transport to mobile terminal；Impedance detection circuit is used for measuring electrode and scalp contact impedance size, feeds back to operating personnel；Power supply module
Power supply is provided for microcontroller；Clock module is that microcontroller sets clock；
Software systems are used to receiving the EEG signals of user, real-time display, while it is whole that EEG signals are converted into amplitude
Syncerebrum electric signal, feature extraction then is carried out to amplitude-integrated EEG signals, then using grader, to the sleep stage of user
Assessed, finally realize that sleep state and quality are assessed using statistical calculations；It includes：User log-in block, number
According to receiving module, original eeg data display module, data format conversion module, data analysis module, result display module and
The inquiry of historical data module；Wherein：
The user log-in block, for logging in system by user；User enters after the software module, data reception BOB(beginning of block)
Start, system enters data receiver pattern；Data reception module mainly realizes data receiver using the bluetooth of mobile device；System
The data of reception are directly shown on the display of the mobile device using original eeg data display module；Meanwhile in system
After the original eeg data received is changed into Amplitude integrated electroencephalogram by data format conversion module, data analysis mould is utilized
Block carries out sleep quality assessment to converted amplitude-integrated eeg data, and dependent evaluation result is sent to result and shown
Module is shown；Sleep quality assessment result is transmitted and is stored in the inquiry of historical data module, is easy to user to realize and is gone through
History data query；
Cloud platform is used for the real-time upload for realizing data, and by original eeg data, the data of mobile terminal analysis include sleep state
And sleep quality, high in the clouds is uploaded in real time, so as to realize slept to user tracking and sleep associated with disease prediction；Meanwhile support
Remotely data are called in storage to multiple client online, support tele-medicine auxiliary and clinical decision；Cloud platform is mainly put down comprising cloud
Platform data memory module, medical knowledge library module and tele-medicine auxiliary and clinical support module；Wherein, cloud platform data storage
Module is mainly used in storage and uploads to the original eeg data of cloud platform and the sleep state and sleep quality number of mobile terminal analysis
The knowledge required for problem solving is included according to, medical knowledge library module, includes expertise, clinical knowledge, health knowledge and comprehensive
Close knowledge；Tele-medicine aids in and clinical support module is responsible for utilizing the data and integrative medicine knowledge base in data memory module
In knowledge go to solve the problem of related to reasoning sleep, aided in and support so as to be provided for tele-medicine and clinical decision.
2. the sleep monitor system according to claim 1 based on Amplitude integrated electroencephalogram, it is characterised in that described letter
In number acquisition module, electrode system is demountable structure, including：Mother-son mouth shape pin thread conductive metal sheet (6), conductive silver paste (7),
Carbon sponge conductive electrode material (8), mother-son mouth shape box conductive metal sheet (9) and conducting wire；Wherein, mother-son mouth shape pin thread is conductive
The button structure (5) of sheet metal (6) and mother-son mouth shape box conductive metal sheet (9) composition, carbon sponge material (8) pass through conductive silver
Slurry (7) is connected with mother-son mouth shape pin thread conductive metal sheet (6), and is further connected with advance signal process circuit.
3. the sleep monitor system according to claim 2 based on Amplitude integrated electroencephalogram, it is characterised in that described letter
In number acquisition module, the electrode system is arranged on an electrode for encephalograms cap, two float electrodes P3, P4 and reference electrode
Fz is arranged on the correspondence position at the top of electrode for encephalograms cap；The material of electrode for encephalograms cap is using elastic two-sided knitting venetian cloth
Material, to be adapted to different brain sizes to obtain preferable signal quality and wear comfort.
4. the sleep monitor system according to claim 2 based on Amplitude integrated electroencephalogram, it is characterised in that software section
Data reception module, original brain electricity display module, data format conversion module, realize that original brain electricity arrives amplitude-integrated brain telecommunications
Number data conversion；Mainly include：Filtering, end points extraction, amplitude compression and time compression；
Filtering：For eliminating because sweating, muscle activity, artefact caused by electrical interference, while the brain electricity of different frequency composition is existed
Decay in transmitting procedure is accordingly compensated；
End points extracts：It is by the fluctuating range reflection of EEG signals to the waveform of amplitude-integrated brain electricity；EEG signals are divided into
Segment, extract per a bit of peak-to-peak value, as the upper extreme point and lower extreme point of the corresponding vertical line of amplitude-integrated brain electricity；
Amplitude compression：For reducing original EEG signals by the dynamic range of changes in amplitude while highlighting the portion of the low wave amplitude of brain electricity
Point；If the peak-to-peak value of original brain electricity is that the vertical line endpoint value extracted is less than 6 μ V, vertical line is still pressed without log-compressed
Linear relationship is shown, for then carrying out log-compressed more than or equal to 6 μ V parts；
Time is compressed：It is to be reflected in prolonged EEG signals in the wave spectrum band of very short amplitude-integrated brain electricity, in order to
Medical personnel observe patient on the whole；Each vertical line is corresponded to the original brain electricity of segment, realizes that the time is compressed.
5. the sleep monitor system according to claim 2 based on Amplitude integrated electroencephalogram, it is characterised in that software section
Data analysis module realize sleep state based on amplitude-integrated EEG signals and quality evaluation；Include amplitude-integrated brain telecommunications
Number pretreatment, feature extraction, grader classification, result statistics；
The pretreatment of amplitude-integrated EEG signals：For to amplitude-integrated EEG signals and analysis and judgement；According to signal width
It is abnormal that value judges that it is belonging respectively to normal amplitude, amplitude mile abnormality and amplitude severe；
Feature extraction：It is that normally amplitude-integrated EEG signals carry out characteristics extraction to amplitude；Including：
(1) aEEG essential characteristic extraction, includes the maximum to original aEEG signals, minimum amplitude, amplitude average, higher than 10uV
Scope amplitude ratio, less than 5uV scope amplitude ratios；
(2) aEEG morphological feature extraction, asking for for the upper lower envelope to original aEEG signals is included；
(3) aEEG Nonlinear feature extraction, the extraction of correlation dimension, Lyapunov indexes and approximate entropy feature is included；
Grader is classified：It is that the characteristic value of extraction is classified, sleep is divided into awakening, shallow slept and three different ranks of sound sleep
Section；
As a result count：It is the full assessment to sleep stage, provides time for falling asleep, recovery time, always sleep duration, difference respectively
Sleep stage distinguishes duration and Sleep efficiency.
6. the sleep monitor system according to claim 5 based on Amplitude integrated electroencephalogram, it is characterised in that aEEG's is non-
In Linear feature extraction：
The correlation dimension is to calculate the relevance before and after variable using correlation integral, is established rules really rule with this to describe signal
And its degree；The step of correlation dimension, is as follows：Assuming that initial data is { u (1), u (2) ..., u (N) } N number of point altogether, set embedding
It is m, time delay τ to enter dimension, obtains the phase space X reconstructed under the parameteri=(ui,ui+τ,...,ui+(m-1)τ), wherein i=
1,2,...,N-(m-1)τ；τ m dimensional vector of N- (m-1) is obtained after phase space reconfiguration；To the phase space vector X of reconstructi, meter
Calculate correlation integral：
H is Heaviside functions in formula, and it is defined asI, j represent vector numbers, and r represents two
Distance between vector, | | Xi-Xj| | represent two phase point XiAnd XjDistance, C (r) represents distance between two points in phase space less than r
Probability；Finally correlation dimension is
The size of the Lyapunov indexes shows the index percent of close orbit averaging convergence or diverging in phase space；It is specific to calculate
Step is as follows：Phase space reconstruction, find each point X on given trackiClosest point X 'iInitial distance, i.e.,Wherein di(0) it is the i-th point of initial distance for arriving its nearest-neighbor；According to maximum Lyapunov
The definition of index, its value is the average value that nearest-neighbor dissipates speed, so there is di(k)=Cieλ1(kΔt), wherein CiFor initially away from
From di(k) it is i-th point of distance with its nearest-neighbor after k chronomere；Above-mentioned equation is taken the logarithm, obtains ln di(k)=
lnCi+λ1(k Δ t), maximum Lyapunov exponent are obtained by least square fitting；
The approximate entropy is a kind of complexity of metric sequence and the nonlinear kinetics parameter of statistic quantification, and it is non-negative with one
Number represents the complexity of a time series, and approximate entropy corresponding to more complicated time series is bigger；The calculating formula of approximate entropy
For：Wherein
Comprise the following steps that：Assuming that initial data is { u (1), u (2) ..., u (N) } N number of point altogether, it is pressed into sequence number consecutive order
One group of m n dimensional vector n is formed, from X (1) to X (N-m+1), Xi=u (i), u (i+1) ..., and u (i+m-1) }, calculate any vectorial Xi
And XjThe distance between (j=1,2 ..., N-m+1, j ≠ i) dij=max | u (i+j)-u (j+k) |, k=0,1 ..., m-1, i.e.,
The maximum of absolute difference is exactly the distance between two vectors between two vectorial corresponding elements；Then given threshold value r is right
Each vectorial Xi, count dijLess than r number and this number and the ratio apart from total N-m+1, it is designated asWillTake
Natural logrithm, its average value to all i is then sought again, is designated as φm(r)；Increase dimension, repeat the above steps, obtain φm+1
(r), so as to obtaining approximate entropy ApEn.
7. the sleep monitor system according to claim 5 based on Amplitude integrated electroencephalogram, it is characterised in that the classification
Device uses SVMs, input space data is mapped to first with a nonlinear transformation characteristic vector of a higher-dimension
Space, optimal separating hyper plane is then constructed in this feature space, linear classification is carried out, after finally mapping back to former space
Just into the Nonlinear Classification in the input space；One-to-one temporal voting strategy is wherein used, i.e., by the class of sample two, two classes of training set
Ground forms 3 training sets, i.e.,：Awakening, shallow sleep；Awakening, sound sleep；It is shallow sleep, sound sleep；Training obtains 3 graders of SVM bis-：SVM1,
SVM2 and SVM3；When test, test sample is sequentially sent to these three two graders, takes the form of ballot to obtain most
Whole classification results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710864522.XA CN107788976A (en) | 2017-09-22 | 2017-09-22 | Sleep monitor system based on Amplitude integrated electroencephalogram |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710864522.XA CN107788976A (en) | 2017-09-22 | 2017-09-22 | Sleep monitor system based on Amplitude integrated electroencephalogram |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107788976A true CN107788976A (en) | 2018-03-13 |
Family
ID=61532165
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710864522.XA Pending CN107788976A (en) | 2017-09-22 | 2017-09-22 | Sleep monitor system based on Amplitude integrated electroencephalogram |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107788976A (en) |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108703753A (en) * | 2018-05-08 | 2018-10-26 | 南京伟思医疗科技股份有限公司 | A kind of neonatal sleep wake cycle detection method based on dynamic electroencephalogram |
| CN108784690A (en) * | 2018-06-19 | 2018-11-13 | 苏州修普诺斯医疗器械有限公司 | Mobile EEG signals Transmission system and method for transmitting signals |
| CN108830865A (en) * | 2018-05-08 | 2018-11-16 | 南京伟思医疗科技股份有限公司 | A kind of determination method of the stabilization up-and-down boundary for Ambulatory EEG image |
| CN109497988A (en) * | 2018-12-13 | 2019-03-22 | 复旦大学 | Multi-modal electro-physiological signals monitor system for domestic sleeping monitoring field |
| CN109512390A (en) * | 2018-10-26 | 2019-03-26 | 武汉大学 | Sleep stage method and wearable device based on EEG time domain various dimensions feature and M-WSVM |
| CN109925587A (en) * | 2019-03-18 | 2019-06-25 | 西安电子科技大学 | A kind of deep sleep improvement detection system and method based on biological low noise |
| CN110074778A (en) * | 2019-05-29 | 2019-08-02 | 北京脑陆科技有限公司 | A kind of extensive brain electrosleep monitoring system based on EEG equipment |
| CN110179457A (en) * | 2019-05-31 | 2019-08-30 | 电子科技大学 | A kind of more physiological signal detecting devices of flexible wearable |
| CN110327040A (en) * | 2019-04-24 | 2019-10-15 | 武汉理工大学 | Sleep stage method and system based on cloud platform |
| CN110575165A (en) * | 2019-09-25 | 2019-12-17 | 北京脑陆科技有限公司 | APP used for brain monitoring and intervention in cooperation with EEG equipment |
| CN111150372A (en) * | 2020-02-13 | 2020-05-15 | 云南大学 | Sleep stage staging system combining rapid representation learning and semantic learning |
| CN111513675A (en) * | 2020-04-07 | 2020-08-11 | 南京伟思医疗科技股份有限公司 | Amplitude-integrated electroencephalogram-based neonatal sleep-wake cycle detection method |
| CN111603135A (en) * | 2020-05-11 | 2020-09-01 | 江南大学 | Low-power-consumption epilepsy detection circuit based on master-slave support vector machine |
| CN112115856A (en) * | 2020-09-17 | 2020-12-22 | 浙江迈联医疗科技有限公司 | Electroencephalogram quality evaluation method, storage medium and system |
| CN112156319A (en) * | 2020-09-22 | 2021-01-01 | 浙江工业大学 | Bone conduction music helps dormancy system |
| CN112545535A (en) * | 2020-12-07 | 2021-03-26 | 杭州沃维医疗科技有限公司 | Sleep-wake cycle analysis method based on amplitude integrated electroencephalogram |
| CN113208615A (en) * | 2021-06-07 | 2021-08-06 | 山东大学 | Continuous electroencephalogram monitoring and feedback system and method for cardio-pulmonary resuscitation instrument |
| EP3900624A1 (en) * | 2020-04-20 | 2021-10-27 | Universitat Rovira I Virgili | An electrode device for amplitude-integrated electroencephalography |
| CN114052747A (en) * | 2020-08-10 | 2022-02-18 | 西安领跑网络传媒科技股份有限公司 | Brain state adjusting method and device, computer equipment and storage medium |
| CN115040754A (en) * | 2022-07-05 | 2022-09-13 | 上海全澜科技有限公司 | Sleep enhancement system based on brain activity detection |
| US11547350B2 (en) | 2018-12-26 | 2023-01-10 | Industrial Technology Research Institute | Personalized parameter learning method, sleep-aid device and non-transitory computer readable medium |
| CN116491909A (en) * | 2023-06-27 | 2023-07-28 | 北京理工大学 | Sleep dynamics characterization method based on amplitude modulation multi-scale entropy |
| CN117349684A (en) * | 2023-12-04 | 2024-01-05 | 南京邮电大学 | Sleep electroencephalogram analysis method based on vector arrangement maximum distance |
| CN117481667A (en) * | 2023-10-24 | 2024-02-02 | 沈阳工业大学 | Electroencephalogram signal acquisition system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101627909A (en) * | 2009-05-05 | 2010-01-20 | 复旦大学附属儿科医院 | Digital amplitude-integrated cerebral function monitor |
| CN203354540U (en) * | 2013-06-05 | 2013-12-25 | 青岛柏恩鸿泰电子科技有限公司 | Electroencephalogram cap |
| CN203789924U (en) * | 2014-04-15 | 2014-08-27 | 武汉格林泰克科技有限公司 | Electrocerebral electrode cap |
| CN104257379A (en) * | 2014-09-23 | 2015-01-07 | 京东方科技集团股份有限公司 | Electroencephalogram processing apparatus and method and sleep monitoring worn device |
| CN104661959A (en) * | 2012-08-23 | 2015-05-27 | 莫纳什大学 | Graphene-based materials |
| CN105769121A (en) * | 2016-02-18 | 2016-07-20 | 南京清辉新能源有限公司 | Three-dimensional carbon-based pressure sensor making method |
| CN106901727A (en) * | 2017-01-13 | 2017-06-30 | 兰州大学 | A kind of depression Risk Screening device based on EEG signals |
-
2017
- 2017-09-22 CN CN201710864522.XA patent/CN107788976A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101627909A (en) * | 2009-05-05 | 2010-01-20 | 复旦大学附属儿科医院 | Digital amplitude-integrated cerebral function monitor |
| CN104661959A (en) * | 2012-08-23 | 2015-05-27 | 莫纳什大学 | Graphene-based materials |
| CN203354540U (en) * | 2013-06-05 | 2013-12-25 | 青岛柏恩鸿泰电子科技有限公司 | Electroencephalogram cap |
| CN203789924U (en) * | 2014-04-15 | 2014-08-27 | 武汉格林泰克科技有限公司 | Electrocerebral electrode cap |
| CN104257379A (en) * | 2014-09-23 | 2015-01-07 | 京东方科技集团股份有限公司 | Electroencephalogram processing apparatus and method and sleep monitoring worn device |
| CN105769121A (en) * | 2016-02-18 | 2016-07-20 | 南京清辉新能源有限公司 | Three-dimensional carbon-based pressure sensor making method |
| CN106901727A (en) * | 2017-01-13 | 2017-06-30 | 兰州大学 | A kind of depression Risk Screening device based on EEG signals |
Non-Patent Citations (4)
| Title |
|---|
| 叶大田等: "振幅整合脑电图在正常睡眠脑电分期中的应用", 《清华大学学报（自然科学报）》 * |
| 王愈: "新生儿振幅整合脑电自动识别研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 王愈: "新生儿振幅整合脑电自动识别研究", 《华东师范大学》 * |
| 王绍宾等: "基于非线性动力学的振幅整合新生儿脑电图分析", 《生物医学工程杂志》 * |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108830865A (en) * | 2018-05-08 | 2018-11-16 | 南京伟思医疗科技股份有限公司 | A kind of determination method of the stabilization up-and-down boundary for Ambulatory EEG image |
| CN108703753B (en) * | 2018-05-08 | 2020-10-27 | 南京伟思医疗科技股份有限公司 | Method for detecting sleep wake cycle of newborn based on dynamic electroencephalogram |
| CN108703753A (en) * | 2018-05-08 | 2018-10-26 | 南京伟思医疗科技股份有限公司 | A kind of neonatal sleep wake cycle detection method based on dynamic electroencephalogram |
| CN108784690A (en) * | 2018-06-19 | 2018-11-13 | 苏州修普诺斯医疗器械有限公司 | Mobile EEG signals Transmission system and method for transmitting signals |
| CN109512390A (en) * | 2018-10-26 | 2019-03-26 | 武汉大学 | Sleep stage method and wearable device based on EEG time domain various dimensions feature and M-WSVM |
| CN109497988A (en) * | 2018-12-13 | 2019-03-22 | 复旦大学 | Multi-modal electro-physiological signals monitor system for domestic sleeping monitoring field |
| US11547350B2 (en) | 2018-12-26 | 2023-01-10 | Industrial Technology Research Institute | Personalized parameter learning method, sleep-aid device and non-transitory computer readable medium |
| CN109925587A (en) * | 2019-03-18 | 2019-06-25 | 西安电子科技大学 | A kind of deep sleep improvement detection system and method based on biological low noise |
| CN110327040A (en) * | 2019-04-24 | 2019-10-15 | 武汉理工大学 | Sleep stage method and system based on cloud platform |
| CN110074778A (en) * | 2019-05-29 | 2019-08-02 | 北京脑陆科技有限公司 | A kind of extensive brain electrosleep monitoring system based on EEG equipment |
| CN110179457B (en) * | 2019-05-31 | 2021-05-07 | 电子科技大学 | Flexible wearable multi-physiological-signal detection device |
| CN110179457A (en) * | 2019-05-31 | 2019-08-30 | 电子科技大学 | A kind of more physiological signal detecting devices of flexible wearable |
| CN110575165A (en) * | 2019-09-25 | 2019-12-17 | 北京脑陆科技有限公司 | APP used for brain monitoring and intervention in cooperation with EEG equipment |
| CN111150372B (en) * | 2020-02-13 | 2021-03-16 | 云南大学 | Sleep stage staging system combining rapid representation learning and semantic learning |
| CN111150372A (en) * | 2020-02-13 | 2020-05-15 | 云南大学 | Sleep stage staging system combining rapid representation learning and semantic learning |
| CN111513675A (en) * | 2020-04-07 | 2020-08-11 | 南京伟思医疗科技股份有限公司 | Amplitude-integrated electroencephalogram-based neonatal sleep-wake cycle detection method |
| CN111513675B (en) * | 2020-04-07 | 2023-09-22 | 南京伟思医疗科技股份有限公司 | Neonate sleep wake period detection method based on amplitude integrated electroencephalogram |
| EP3900624A1 (en) * | 2020-04-20 | 2021-10-27 | Universitat Rovira I Virgili | An electrode device for amplitude-integrated electroencephalography |
| WO2021213985A1 (en) * | 2020-04-20 | 2021-10-28 | Universitat Rovira I Virgili | An electrode device for amplitude-integrated electroencephalography |
| CN111603135A (en) * | 2020-05-11 | 2020-09-01 | 江南大学 | Low-power-consumption epilepsy detection circuit based on master-slave support vector machine |
| CN111603135B (en) * | 2020-05-11 | 2021-09-28 | 江南大学 | Low-power-consumption epilepsy detection circuit based on master-slave support vector machine |
| CN114052747A (en) * | 2020-08-10 | 2022-02-18 | 西安领跑网络传媒科技股份有限公司 | Brain state adjusting method and device, computer equipment and storage medium |
| CN112115856A (en) * | 2020-09-17 | 2020-12-22 | 浙江迈联医疗科技有限公司 | Electroencephalogram quality evaluation method, storage medium and system |
| CN112156319A (en) * | 2020-09-22 | 2021-01-01 | 浙江工业大学 | Bone conduction music helps dormancy system |
| CN112545535A (en) * | 2020-12-07 | 2021-03-26 | 杭州沃维医疗科技有限公司 | Sleep-wake cycle analysis method based on amplitude integrated electroencephalogram |
| CN112545535B (en) * | 2020-12-07 | 2023-01-06 | 杭州沃维医疗科技有限公司 | Sleep-wake cycle analysis method based on amplitude integrated electroencephalogram |
| CN113208615A (en) * | 2021-06-07 | 2021-08-06 | 山东大学 | Continuous electroencephalogram monitoring and feedback system and method for cardio-pulmonary resuscitation instrument |
| CN115040754A (en) * | 2022-07-05 | 2022-09-13 | 上海全澜科技有限公司 | Sleep enhancement system based on brain activity detection |
| CN116491909A (en) * | 2023-06-27 | 2023-07-28 | 北京理工大学 | Sleep dynamics characterization method based on amplitude modulation multi-scale entropy |
| CN116491909B (en) * | 2023-06-27 | 2023-09-12 | 北京理工大学 | Sleep dynamics characterization method based on amplitude modulation multi-scale entropy |
| CN117481667A (en) * | 2023-10-24 | 2024-02-02 | 沈阳工业大学 | Electroencephalogram signal acquisition system |
| CN117349684B (en) * | 2023-12-04 | 2024-02-06 | 南京邮电大学 | Sleep electroencephalogram analysis method based on vector arrangement maximum distance |
| CN117349684A (en) * | 2023-12-04 | 2024-01-05 | 南京邮电大学 | Sleep electroencephalogram analysis method based on vector arrangement maximum distance |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107788976A (en) | Sleep monitor system based on Amplitude integrated electroencephalogram | |
| Ozkan et al. | A portable wearable tele-ECG monitoring system | |
| CN105286909B (en) | A kind of wearable heart sound and ecg characteristics information collection and monitoring system | |
| CN204306822U (en) | Wearable electrocardiosignal monitoring device | |
| WO2017075856A1 (en) | Wavelet analysis-based remote electrocardiogram monitoring and warning system and method | |
| CN106937808A (en) | A kind of data collecting system of intelligent mattress | |
| CN102462494B (en) | Novel intelligent electrocardiogram test healthcare apparatus | |
| CN106073736A (en) | A kind of extendible human parameters monitoring and control recording system | |
| CN110916647A (en) | Wearable electrocardio monitoring and arrhythmia remote real-time diagnosis device facing multiple scenes | |
| CN104605841A (en) | Wearable electrocardiosignal monitoring device and method | |
| CN106166065B (en) | Wearable electrocardio health interaction platform based on social network and implementation method thereof | |
| Krishnan | Biomedical signal analysis for connected healthcare | |
| CN109124655A (en) | State of mind analysis method, device, equipment, computer media and multifunctional chair | |
| CN107669262A (en) | Multi-lead Telediagnosis of Electrocardiogram Signals and monitor system and method based on SVM and WLT | |
| El Attaoui et al. | Wearable wireless sensors network for ECG telemonitoring using neural network for features extraction | |
| Randazzo et al. | VITAL-ECG: A portable wearable hospital | |
| CN201227272Y (en) | Dress type electro-cardio and respiration rate monitoring device based on Zigbee | |
| CN105232029A (en) | Automatic ultralow-power-consumption wearable dynamic electrocardiogram analysis equipment and implementation method thereof | |
| Yoshimoto et al. | Recent progress of biomedical processor soc for wearable healthcare application: A review | |
| CN201894645U (en) | Novel intelligent electrocardiogram test healthcare apparatus | |
| CN106344003A (en) | Multifunctional auxiliary electrocardiogram monitoring device | |
| Gong et al. | Design and implementation of wearable dynamic electrocardiograph real-time monitoring terminal | |
| CN211883766U (en) | Cardiovascular disease remote monitoring and early warning system | |
| CN114504333B (en) | Wearable vestibule monitoring system based on myoelectricity and application | |
| CN105852847A (en) | Heart and vital sign monitoring and analyzing system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180313 |