CN104107037A - Physiological information acquiring and processing system - Google Patents

Abstract

The invention provides a physiological information acquiring and processing system. The system comprises a sensor system, a transmission line, a signal processing circuit and a computer host, wherein the sensor system is used for detecting pressure measurement signals in sleep of a patient without contact; the signal processing circuit is used for separating the measured signals into various independent physiological signals; the computer host is used for extracting useful information and integrating the information into physiological case state change information so as to analyze sleep indexes and evaluate the sleep level. Through the physiological information acquiring and processing system, automatic detection is performed when the patient falls asleep, data are automatically transmitted when the patient gets off the bed, and no disturbance is caused to life of the monitored person. An alarm can be given in time for abnormality in sleep. Physiological information in sleep is detected in real time for scientific analysis, and the analysis result is displayed.

Description

Physiologic information acquisition and processing system

Technical field

The present invention relates to data acquisition process field, particularly physiologic information acquisition and processing in sleep.

Background technology

Research to vital sign in sleep is directly connected to the research to disease, therefore becomes the topic of being relatively concerned about in medical science.Further investigation reveals that and breathe in sleep and all standing of heart beating has not been an example, it is greatly threatening, and people's is healthy.Reasonably diagnosis and treatment as early as possible, the quality of life that can improve patient is prevented the generation of various complications, obviously improves patient's survival rate.Therefore, to the monitoring of sign in sleep procedure be the first step of prevention and diagnose and treat diseases.

Traditional Sleep architecture measuring technique (Rechtschaffen & Kales slightly writes R & K) nineteen sixty-eight is proposed by each experience of the comprehensive summary of California, USA university Brain Research Institute.The Data Source that it is analyzed is made up of 2 road brain electricity, 2 tunnel eye movement electricity and 1 road mentalis electricity, and these data are analyzed through comprehensive, obtains the Sleep architecture of 6 phases that were divided into.From above summary, with R & K-method detection Sleep architecture, at least to paste 10 pieces with top electrode at examinee's head, can bring certain Physiological Psychology load to examinee, some is higher to the requirement for environmental conditions of falling asleep especially, and comparatively this load of sensitive subjects is sizable.Therefore R & K-method is only suitable for those persons that easily do not fall asleep especially (as suffering from more serious sleep apnea syndrome person).Certainly R & K-method is more not suitable for the required Sleep architecture monitoring of professional job safety.

In fact the many physiological signals on the person, as cardiac cycle, respiratory wave parameter, the moving parameter of body, skin resistance, body temperature etc., all can present corresponding variation with the cyclically-varying of each phase of sleep.Therefore many scholars of countries in the world all be devoted to various physiological signals beyond brain electricity and Sleep architecture change between the research of coupled relation.But existing technical scheme all fails signal to excavate and information fusion engages, thus cannot be exactly at the potential useful information of sleep physiology signal extraction, various physiology case state change informations in complete reliable sleep can not be provided.

Therefore,, for existing the problems referred to above in correlation technique, effective solution is not yet proposed at present.

Summary of the invention

For solving the existing problem of above-mentioned prior art, the present invention proposes a kind of physiologic information acquisition and processing system, utilize the win situation of change of signal of non-contacting approach, and under the contrast of R & K-method, sum up feature and rule that these physiological signals have in Sleep architecture cyclically-varying, obtain non-contacting Sleep architecture measuring technique, work out various physiological signals beyond brain electricity and Sleep architecture change between coupled relation.

The present invention adopts following technical scheme: a kind of physiologic information acquisition and processing system, comprising:

Sensing system, for detecting the pressure measurement signal of patient's sleep under noncontact condition;

Transmission line, for transferring to signal processing circuit by physiological signal;

Signal processing circuit, for analyzing the vibration output of detected physiological signal, and is separated into each independently physiological signal by this signal;

Main frame, the physiological signal obtaining for receiving sleep, and:

Extract useful information, and by these information, under predefine constraints, be integrated into physiology case state change information; And

Based on detected physiological signal, according to index analysis and the sleep level evaluation of sleeping of predefine standard.

Preferably, described sensing system comprises air mattress and air pressure probe, and the faint pressure of this air pressure probe sensing mattress changes.

Preferably, described physiological signal comprises pulse wave signal, respiratory wave signal, and snoring signal, body moves signal.

Preferably, described sensing system also comprises comb filter, processes for utilizing bandpass filtering modules block to implement FFT to the data of sensor output, extracts the each physiological signal component under sleep state.

Preferably, described main frame takes redundant computation and credibility to calculate, and to evaluate the reliability of different approaches institute extracting parameter, chooses the result that credibility is greater than predetermined threshold, for sleep architecture stage.

Preferably, the parameter of described sleep architecture stage utilization comprises: the moving information of cardiac cycle, breathing cycle, body and/from bed information;

Described main frame, based on mass data, is found the Sleep architecture dependency rule based on obtaining parameter, sets up sleep stage knowledge base; Application uncertain reasoning theory, merges multiparameter information, carries out the analysis ratiocination of Sleep architecture.

Preferably, described sensing system utilizes Active signal controlling, eliminates the interference of intrinsic vibration cycle to physiological signal in sleep around.

Preferably, when described system detects while occurring that heartbeat is abnormal in sleep or when respiratory arrest, system automatic alarm.

Than prior art, the having the following advantages of technical scheme of the present invention:

1) " passive type " detects, and automatically detects and starts, data automatic transmission during from bed while falling asleep.Without the collection and the transmission that initiatively get involved daily life information, live noiseless to monitored people.

2) abnormality alarming, Sleeping Center is jumped, and breathes, and enters timely alarm when abnormal from bed, and sends note.

3) long distance monitoring, jumps based on real-time detection Sleeping Center, breathes, and snoring, the physiologic information such as stands up, and carries out science dissection process, displayed map table analysis result.Lasting detection, can provide, week/month/year statement-of-health.By mobile phone, the real-time monitorings such as computer.

Brief description of the drawings

Fig. 1 is according to the module map of the physiologic information acquisition and processing system of the embodiment of the present invention.

Fig. 2 is the bio-signal acquisition schematic flow sheet according to the embodiment of the present invention.

Fig. 3 passes schematic diagram according to the sleep level of the embodiment of the present invention.

Detailed description of the invention

Below provide the detailed description to one or more embodiment of the present invention together with illustrating the accompanying drawing of the principle of the invention.Describe the present invention in conjunction with such embodiment, but the invention is not restricted to any embodiment.Scope of the present invention is only defined by the claims, and the present invention contain manyly substitute, amendment and equivalent.Set forth in the following description many details to provide thorough understanding of the present invention.These details are provided for exemplary purposes, and also can realize the present invention according to claims without some or all details in these details.

In sleep, the excavation of the signal of physiological and pathological state and information fusion technology particular content are: at the potential useful information of sleep physiology signal extraction, and by these from heterogeneity, dissimilar, the potential information of different approaches, under certain constraints, through mutually supplementing and checking mutually, integrate out than various physiology case state change informations in more complete more reliable sleep originally.After coupled relation beyond brain electricity between physiological signal and Sleep architecture variation finds, with contactless, noiseless state, the technical method that obtains these physiological signals just becomes key of the present invention.For example, by the ballistocardiogram obtaining from air mattress, the analysis of the moving waveform of respiratory movement and body, extracts cardiac cycle variability, breathing cycle variability, the respiratory wave amplitude variation opposite sex, and the moving information of body, and the information of mutual relation between them, judge Sleep architecture; Then cardiac cycle when the asphyxia, respiratory waveform, the moving feature changing of ballistocardiogram and body, picks out and whether has asphyxia, and suspend be obstructive or central, and whether there is firmly the information such as breathing pattern microarousal.

Generally, we's surface technology scheme be physiologic information to gather from sensor special by analog filtering processing, be separated into that end pulse, breathing, body are moving, snoring signal, through high speed Fourier transform, comb shape Filtering Processing, and dormant judgement is carried out in calculation.One of key point is how from the signal mixing, correctly to extract physiological signal out.Utilize merely software (FFT calculation) system of filtering to be difficult to realize this function, and with suitable hardware simulation filtration, improve S/N ratio.The present invention increases on this basis software and judges, further improves precision.The diagnostic method of one of them software application is, by by parallel pulse, breathe, the moving signal of body is separated, within the unit interval, there is back number for same scope amount and carry out Pareto diagram, solve its envelope, in certain probable range, extract data time-sequencing once again out, establish sleep state cyclically-varying.These numerical value are processed statistical calculation method, and physiological data Changing Pattern has application extremely widely.

Fig. 1 is according to the structure chart of the physiologic information acquisition and processing system of the embodiment of the present invention.As shown in Figure 1, system provided by the invention is made up of air mattress, air pressure probe, transmission line, signal processing circuit (5-10Hz wave trap, band filter etc.) and main frame.Main frame comprises the 32 bit processor MPU that calculate for data, for storing the read only memory ROM of calculation procedure data, for the LCD display device to user's display process result, for reading and writing the internal memory SD card of the signal data before and after computing, for signal being sent to the wireless communication module of other device of remote port, COM1 etc.

Air mattress forms a sensing system to faint pressure sensitive together with air pressure probe, detects the measurement of various physiological signals in sleep under noncontact condition.When heartbeat, pulse are beaten and breathe while causing health to have small vibration, the responsive mattress of fine motion has corresponding faint variation output, is processed by signal processing circuit.

Utilize algorithm in this paper that measured signal segregation is become to pulse wave signal, respiratory wave signal, snoring signal, body moves signal, and Sleeping Center is dirty while beating speed or irregular rhythm or automatic alarm when breathing occurs stopping.Meanwhile, based on the pulse wave signal obtaining, respiratory wave signal, the moving signal of the body index analysis of sleeping, carries out the evaluation of sleep level according to R & K standard.

Fig. 3 passes curve chart according to the sleep level of the embodiment of the present invention.Be followed successively by from top to bottom: according to the sleep level of sleep signal; According to the sleep level of heart arteries and veins signal; The sleep level that the length of one's sleep, relation function was established; Consider the sleep level that the moving impact of body is established.

Such structural design has the feature of safety, easy care.Design mattress has the performance of many outputs, and its advantage is: improve detection sensitivity; And implicit certain redundant signals, be beneficial to post processing level and carry out reliable parameter extraction and analysis.

The present invention further excavates and information fusion the signal of physiological and pathological state in sleep.Its particular content is: at the potential useful information of sleep physiology signal extraction, and by these from heterogeneity, dissimilar, the potential information of different approaches, under certain constraints, through mutually supplementing and checking mutually, integrate out than various physiology case state change informations in more complete more reliable sleep originally.

1) extraction of life parameters

The initial data obtaining from sensing system is isolated heart shock wave, pulse wave and respiratory wave, and then extracts the life parameterses such as cardiac cycle and breathing rate, and this is the basic and crucial of whole accurate natural sleep detection technique.Primary signal feature is subject to the impact of individual morphology, sleeping position large, has the advantages that pattern makes a variation, metamorphosis is large.Parameter information wherein thereby there is the features such as incomplete, fuzzy, in post processing in order to overcome above factor, take redundant computation, and design credibility and calculate, in order to evaluate the reliability of different approaches institute extracting parameter, extract result and choose credibility the greater, and for next step sleep architecture stage.

2) sleep architecture stage

The parameter of sleep architecture stage utilization comprises: the moving information of cardiac cycle, breathing cycle, body and/from bed information.The problem solving is how to pass through these gain of parameter Hypnograms.Artificial intelligence's thought and theory used in the realization of this sleep stage method.First investigating on the basis of mass data, finding and summed up a large amount of Sleep architecture (expert analyzes gained with R & K-method) dependency rules based on obtaining parameter, setting up sleep stage knowledge base; Uncertain reasoning theory in using artificial intelligence, according to knowledge base, scientifically merges multiparameter information, carries out the analysis ratiocination of Sleep architecture.

The present invention proposes cardiac cycle and sleep level relation mathematic model is as follows:

Middle cycle frequency range heart arteries and veins variation h _u(t), pass S with frequency range sleep level _m(t) time difference equation:

${\hat{S}}_m\left(t\right)+a_1{\hat{S}}_m(t-1)+...+a_n{\hat{S}}_m(t-n)=b_0{h}_u\left(t\right)+b_1{h}_u(t-1)+...+b_m{h}_u(t-m)$ $\left(1\right)$

A _n: the denominator coefficients (n=1-5) of difierence equation

B _m; The denominator coefficients (m=1-5) of difierence equation

The continuous sleep level S of middle cycle frequency range _mand middle cycle frequency range heart arteries and veins variation h (t) _u(t)

${\hat{S}}_m\left(t\right)=0.996{\hat{S}}_m(t-1)+0.255h_u\left(t\right)-0.259h_u(t-1)---\left(2\right)$

According to heart arteries and veins, the function His length of one's sleep (k) of continuous sleep level is inferred in variation

$\mathrm{his}\left(k\right)=N\·\frac{\mathrm{his}\left(k\right)}{S_h}---\left(3\right)$

Wherein S _hfor the amplitude of fluctuation of sleeping continuously

Ask the minima of regular distribution function

$\mathrm{Min}\left(\underset{k=0}{\overset{N}{\mathrm{\Σ}}}{\{\mathrm{his}\left(k\right)-\underset{i=1}{\overset{6}{\mathrm{\Σ}}}{w}_i\frac{1}{{\mathrm{\σ}}_i\sqrt{2\mathrm{\π}}}\exp (-\frac{{(k-m_i)}^2}{2{\mathrm{\σ}}_i^2})\}}^2\right)---\left(4\right)$

Night each sleep level coefficient w _i

w ₆＝0.102，w ₅＝0.161，w ₄＝0.132，

w ₃＝0.472，w ₂＝0.089，w ₁＝0.056， (5)

REM sleep, when awakening, heart arteries and veins amplitude of variation is very large, how to distinguish REM sleep very difficult with awakening, for accurate judgement, need to judge that body moves information.The moving signal intensity of body is far better than heart arteries and veins signal intensity.According to heart arteries and veins oscillogram, determine that 1 minute heart arteries and veins changes.If in this 1 minute moving signal of body, the ratio of establishing this 1 minute body fatigue resistance.Body moving with awakening relation in, REM sleep (Rapid eye movement sleep) relation is as follows:

The moving relation of Sleep stages and body adopts mathematical model below.

$M^{\′}\left(t\right)=\frac{M\left(t\right)-M_{\max }}{M_{\max }-M_{\min }}$

T: time;

M (t): the moving size of body;

M ' is (t): the moving size of standardization body

M _max: in the moving size of body, the meansigma methods of upper 5 numerical value;

M _min: in the moving size of body, the meansigma methods of the next 5 numerical value

(t) determine REM sleep below and awakening judge index Iwr with M (t) according to M ':

Iwr (t)=M ' (t)/Mo, a meansigma methods of M (t) when wherein Mo enters bed for tester and when clear-headed before getting up.

A large amount of clinical datas prove:

Iwr (t) > 0.87 determines awakening

(REM sleep is determined in t)≤0.87 to Iwr

Continuously sleep level is to standard 6 grade transformation of sleeping

$y_i=e{\left(\frac{{(k-m_i)}^2}{2{\mathrm{\σ}}_i^2}\right)}^2,i=\mathrm{1,2},...6$

M _i: the meansigma methods (i=1-6) of each sleep level

σ _i: the deviation (i=1-6) of each sleep level

K: sleep level continuously

Y _i, i=1,2 ... 6: the probit that each sleep level is corresponding, between 0-1.

A large amount of clinical datas are determined

m ₆＝0.62±0.16，σ ₆＝0.15±0.08

m ₅＝0.65±0.10，σ ₅＝0.18±0.05

m ₄＝0.51±0.13,σ ₄＝0.13±0.10

m ₃＝0.44±0.08,σ ₃＝0.20±0.03

m ₂＝0.35±0.09,σ ₂＝0.17±0.06

m ₁＝0.33±0.14,σ ₁＝0.14±0.07

According to mathematical model above, determine sleep level result of determination.

The application of multiparameter information fusion thought has overcome the uncertain limitation of the insufficient conclusion of information in single parameter source, has solved well the sleep analysis under sensor as aforementioned prerequisite, a large amount of verification experimental verifications feasibility and the reliability of this method.

Fig. 2 is the bio-signal acquisition schematic flow sheet according to the embodiment of the present invention.Describe the system of the embodiment of the present invention in detail to the decision method of each physiological status below in conjunction with Fig. 2, and the detailed step of the storage of system acquisition information and transmission data.

1 sleep is judged

1. extract key element out: Pulse Rate, Respiration Rate, body moving (body moves size exclusion pulse key element non-active ingredients+significantly moving time of origin of body), snoring (snoring time of origin)

2. data sampling interval/calculation object data number: 39ms/sample, approximately 10 Miao Jian Even continued

3. pulse is extracted mode out: extract pulse composition by simulation filtration system loop and implement microcomputer FFT calculation.

Among the Pulse Rate of 48 times～600 times, carry out cutting with the intervals such as 0 time～59 times, 60 times～119 times, 120 times～179 times, 180 times～239 times, 240 times～299 times, 300 times～359 times, 360 times～419 times, 420 times～479 times, 480 times～539 times, 540 times～599 times

Extract respectively the maximum Pulse Rate among each Pulse Rate interval out.Great majority become in the Pulse Rate composition of multiple relation, get minimum Pulse Rate as pulse signal.The Pulse Rate that (utilize wave component to a high-profile, improve specific precision) extracted as unit taking one minute ', calculate by following algorithm:

Pulse Rate A=(Pulse Rate ' [1]+Pulse Rate ' [2]+Pulse Rate ' [3]) ÷ 3

Pulse Rate B=(Pulse Rate ' [4]+Pulse Rate ' [5]+Pulse Rate ' [6]) ÷ 3

Pulse variation=(Pulse Rate A-Pulse Rate B)/2

Result of calculation is kept in file.

4. breathe extraction mode: extract respiratory component by simulation filtration system loop and implement microcomputer FFT calculation.

Using the Pulse Rate among the Respiration Rate of 6 times～90 times as judgement, from the corresponding maximum breathing number of its medium and small numerical value as breathing key element.

Data are preserved, and take to get its meansigma methods again and preserved according to being divided into 6 deciles with 10 number of seconds.

Directly resolve [algorithm alone] by waveform and carry out repeatedly breathing cycle parsing, number is returned in the breathing of extracting one minute with this.

5. the extraction mode of snoring: utilize speech band territory (10Hz～100Hz), judge as breathing the moment using certain volume, differentiate and have or not snoring with this.

For unit is used as minute, extracted the snoring time that occurs at 60 seconds.

6. human body has or not extraction mode: measure human body weight by pressure transducer, judge that with this human body has or not.

7. the moving extraction mode of body:

Signal in upper note each Pulse Rate interval that 2. a pulse is extracted out is tried to achieve body as the summation of judgment value with the ratio of the summation of all values and is moved.

The moving signal of (summation-conduct of all values judges the summation of the value [comprising high harmonic] of signal)=body

Implement waveform and directly resolve [algorithm alone], in the data sampling data with [having taking 30 seconds as unit of account according to detecting required precision] between 10 seconds, extract the moving time of body that occurs out.

8. sleep starts to judge: human body is in bed perception and the moving probability perception of body occurs significantly and breathe fixing perception or press SLEEP ANALYZE SW button to start to judge that sleep starts.

9. the judgement of getting up: when continuing from 20 points of above perception of bed or can not extracting pulse/breathing out for a long time or press WAKEUP SW and press the button as the judgement of getting up.

10. sleep is judged: get up judge after, the judgement of sleeping of following formula.

Calculate RSI (indicating the index of REM sleep state), SDI (indicating the index of Depth of sleep) by the data that obtain.

RSI calculates method: using calculate between pulse variable signal between first 10 points of object data [pulse variable signal between 1 point], latter 10 points pulse variable signal as

Rolling average result

SDI calculates method: 0.5log2 × (summation of all values of the moving signal ÷ of body)

Sleep state is judged: following table is benchmark, and compare the each dormant average time of obtaining according to each age group, obtains respectively following several state

Awakening/REM sleep/NonREM sleep 1/NonREM sleep 2/NonREM sleep 3/NonREM sleep 4.

The REM judgement of sleeping

The RSI value calculating is arranged by descending, according to the age information of login, be judged to be REM sleep with the sleep portion approaching till the most moving in the poor scope of individual average REM length of one's sleep [standard deviation].

Awakening/NonREM1 condition judgement

Except judging the data of REM sleep, SDI data are arranged with descending, according to the age information of login, to approach average wakefulness

Part till the most moving in the poor scope of time individual [standard deviation] is judged to be wakefulness.

Be judged to be NonREM1 sleep with part till approaching the most moving in the poor scope of individual average N onREM1 length of one's sleep [standard deviation].

NonREM2/NonREM3/NonREM4 condition judgement

SDI is arranged by descending, is calculated with following method:

SDI descending [ADR1]-SDI descending [ADR2]=SSDI[ADR1],

SDI descending [ADR2]-SDI descending [ADR3]=SSDI[ADR2],

SDI descending [ADRn-1]-SDI descending [ADRn]=SSDI[ADRn-1].

Till approaching the maximum SSDI area part in the average NonREM2 poor scope of the individual length of one's sleep [standard deviation], be judged to be NonREM2 sleep.Till approaching the maximum SSDI area part in the average NonREM3 poor scope of the individual length of one's sleep [standard deviation], be judged to be NonREM3 sleep.And the NonREM3 later area part of sleeping is judged to be to NonREM4 sleep.

SSDI＝Slope of SDI

The data that sequence is cut apart each Sleep stages are on time preserved hereof with ordered series of numbers form.

2 apneas are judged

Directly resolve [algorithm alone] according to waveform and judge and breathe no more, continue 10 seconds above occasions, be judged to be apnea state one time.

If breathe while stopping health is not caused to harmful effect for discontented 10 seconds, this situation happens occasionally conventionally, is called non-disease respiratory arrest.

This state is exported as mechanical contacts.

3 according to environment, the automatic adjustment of user

While considering feature, sensor setting position, the initial setting of user body weight, bed, heart beating, breathes, and stands up the suitableeest set-up function of the signal extractions such as snoring.

4SD memory storage

Save data form: the csv file of cutting apart with comma form between data.

Filename:

Heartbeat signal, H0000000.csv filename starts, and every 2 points of intervals generate 1 file, while exceeding H0020160.csv, return to H0000000.csv, rewritable paper.

Breath signal, B0000000.csv filename start, and every 2 points of intervals generate 1 file, while exceeding B0020160.csv, return to B0000000.csv, rewritable paper.

Physiologic information state, S0000000.csv filename start, and every 2 points of intervals generate 1 file, while exceeding S0020160.csv, return to S0000000.csv, rewritable paper.

Obtain in the same time above-mentioned three files, letter below numeral is identical.

Sleep state, SLEEP000.csv filename start, and every 2 points of intervals generate 1 file, exceed SLEEP999.csv, return to SLEEP000.csv, rewritable paper.

Save data content (all data are with ASCII character form):

Heartbeat signal data format: the moving 39ms of pulse body interval gathers initial data [60 seconds intervals]

Time series data 1[3 byte 16 systems] the funny number 1 of+comma [1 byte 0x2c]+data 2+ comma+data 3+, 535+ comma+data 1,536+CR[1 byte 0x0d]

Breath signal form: 39ms interval gathers initial data [dividing between 60 seconds]

Time series data 1[3 byte 16 is entered]+comma [1 byte 0x2c]+data 2+ comma+data 3+ comma

Data 1,535+ comma+data 1,536+CR[1 byte 0x0d]

Status file form:

1. date temporal information

2,0,1,3, (year [8 byte 10 systems]) 0,5, (month [4 byte 10 systems]) 3,0, (day [4 byte 10 systems]) 1,2, (time [4 byte 10 systems]) 1,2, (point [4 byte 10 systems])

2. Pulse Rate change information

0,0,5, (inferior [6 byte 10 systems])

3. Pulse Rate A information

0,9,1, (inferior [6 byte 10 systems])

4. Pulse Rate B information

1,0,2, (inferior [6 byte 10 systems])

5. the moving size information [SDI value] of body

3,0,1, (rate [6 byte 10 systems])

6. Respiration Rate information

0,1,8, (inferior [8 byte 10 systems])

7. apnea temporal information

1,2, (second [4 byte 10 systems])

8. the temporal information of snoring

2,5, (second [4 byte 10 systems])

9. the moving temporal information of pressure sensor body

1,5, (second [4 byte 10 systems])

10. human body with/without, start/awakening of sleep, warning output with/without mark (FLG)

H, U, 0or1, S, L, 0or1, A, L, 0or1 (17 byte ASCII)

machine state FLG

E, R, 0or1,0or1,0or1,0or1,0or1,0or1,0or1,0or1, (20 byte)

Abnormal FLG[has extremely=1, abnormal without=0] order with as described below, as, E, R,, radio equipment has or not extremely, and maintain communications has or not extremely, pulse signal has or not extremely, breath signal has or not extremely, and snoring signal has or not extremely, and SD card has or not extremely, wave analysis module has or not extremely, and sleep sensor hardware has or not extremely.

Program version information

V, 1,0,0, (8 byte ASCII)

Each Sleep stages document format data:

Date temporal information

2,0,1,3, (year [8 byte 10 systems]) 0,5, (month [4 byte 10 systems]) 3,0, (day [4 byte 10 systems]) 1,2,

(time [4 byte 10 systems]) 1,2, (point [4 byte 10 systems])

The identification FLG of each state is respectively, wakefulness: A

REM sleep state: B

NonREM1 sleep state: C

NonREM2 sleep state: D

NonREM3 sleep state: E

The get up form of state: Z of NonREM4 sleep state: F, in time series with comma

State+time is divided and made a distinction.Each change of state frequency is limited with maximum 200, and file size is fixed as 2K byte.

(example) Z, 0,2,5, A, 0,0,5, D, 0,1,0, E, 0,0,3, F, 0,3,0, E, 0,0,2, D, 0,0,5, B, 0,4,2,

5 signal communications

1. communication for conservation port/radio communication function [Zigbee]

When power supply On: communicate according to 5 communication designs by ancillary equipment, can extract out various settings and save data.

In sleep analysis: termly [suppose now 60 seconds for unit] by described in 6-4 item with the fixed form of csv file from holding end to deliver letters.

Aspect data collecting card, the present invention applies the quick conversion chip of 12, meanwhile, in order to adapt to the sampling of upper frequency real-time under Windows operating system, installs one 8253 intervalometers on sampling card additional.Its parameter index:

Analog input channel; Single-ended 16 tunnels (this is for four berths, should use 64 tunnels for 16 berths)

A/D changes figure place: 12

Input voltage range :-5V～+ 5V

Conversion time: 7 microseconds

Output code system: offset binary code

Systematic error: 0.1% (comprising: passage, sampling keep, A/D transformed error)

Computational methods: (nominal value-measured value)/ull-scale value * l00%

Input impedance: > 10Mohms

The setting of 8253 intervalometers:

Connect timing signal l.5MHz, the highest count frequency is 1.5MHz.

Under 6 kinds of working methods, counting finishes backward main frame and sends out interruption.(PC bus takies IRQ2, and AT-bus takies IRQ9).

Below running environment and the configuration of system of the present invention.

Appearance design:

Waterproofing protection grade: nothing

This body profile cun method: 150mm (W) × 110mm (H) × 40mm (D)

Sensor portion profile cun method: 500mm (W) × 250mm (H) × 20mm (D)

This body weight: about Kg

Sensor portion: about Kg

Electrical design:

Power input voltage scope: DC5V ± 5% following (from the input of AC adapter)

Following (while conventionally the action)/A/VDC of consumption electric current: A/VDC following (when maximum actuation)

Inrush current: maximum A/VDC following (ms is following)

Communication mode 1:RS232[non-same period of communication] (baud rate 9,600bps/ data bit 8bit/ parity is without/position of rest 1bit)

Communication that communication mode 2:Zigbee or Wi-fi is wireless [2.4GHz is with territory]

Environment Design

1. serviceability temperature :+5～+ 35 DEG C

2. storage temperature: 0-+50 DEG C

3. use/preserve humidity: as long as 30%～90% Antidewing]

4. condition is set

4.-1 noncorrosive gases occasion

4.-2 little dust, the occasion that aeration is good

4. the very little occasion of-3 vibration

4.-4 occasions without forceful electric power Jie Jiqiangci circle

Reliability design

In sum, the physiologic information acquisition and processing system that the present invention proposes, has following characteristics than prior art:

1) without physiologic information collection in constraint, contactless sleep.On the bed of tested object, arrange at least more than one without constraint, contactless sensor, detect multiple physiology signals of moving, the breathing of the described tested object of the h.d. body relevant with sleep quality in bed of above-mentioned tested object, heartbeat and so on, then use comb filter to carry out date processing, extraction is accompanied by the moving signal of dormant pulse signal and body, is provided with to eliminate the structural system that imports external signal into by bed and prevent external disturbance on the pillar of bed.The physiology signal that air type pressure transducer is measured, by bandpass filtering modules block, is separated into cardiac cycle part, breathes part, snoring part, and body moves part, waits physiological signal separating treatment.

2) comb filtering physiologic information separates.The data of the present invention to sensor output, implement FFT and process, and according to the basic wave of pulse, harmonic component and other component, moving to pulse component and body component are separated.But, following the body disorder of internal organs standing up etc. to contain near low frequency component heart rate number more, therefore, the heart rate number of sometimes asking for according to the peaks spectrum of FFT contains error.In the present invention, propose to have a kind of method, it separates pulse and body moving accurately in order to effectively utilize the harmonic component of pulse component, has used comb filter.

3) utilize Active signal controlling, eliminate cycle signal around superly, the intrinsic vibration cycles such as super expressway are jumped Sleeping Center, the interference of breath signal.

4) heartbeat signal based on measuring, body moves signal, each Sleep stages average originating rate and standard deviation calculation function.

5) heartbeat signal based on measuring, body moves signal, infers the algorithm of Sleep stages.

Should be understood that, above-mentioned detailed description of the invention of the present invention is only for exemplary illustration or explain principle of the present invention, and is not construed as limiting the invention.Therefore any amendment of, making, be equal to replacement, improvement etc., within protection scope of the present invention all should be included in without departing from the spirit and scope of the present invention in the situation that.In addition, claims of the present invention are intended to contain whole variations and the modification in the equivalents that falls into claims scope and border or this scope and border.

Claims (9)

1. a physiologic information acquisition and processing system, is characterized in that, comprising:

Sensing system, for detecting the pressure measurement signal of patient's sleep under noncontact condition;

Transmission line, for transferring to signal processing circuit by physiological signal;

Signal processing circuit, for analyzing the vibration output of detected physiological signal, and is separated into each independently physiological signal by this signal;

Main frame, the physiological signal obtaining for receiving sleep, and:

Extract useful information, and these information are integrated under predefine constraints to physiology case state change information; And

Based on detected physiological signal, according to index analysis and the sleep level evaluation of sleeping of predefine standard.

2. system according to claim 1, is characterized in that, described sensing system comprises air mattress and air pressure probe, and the faint pressure of this air pressure probe sensing mattress changes.

3. system according to claim 2, is characterized in that, described physiological signal comprises pulse wave signal, respiratory wave signal, and snoring signal, body moves signal.

4. system according to claim 3, is characterized in that, described system also comprises comb filter, processes for utilizing bandpass filtering modules block to implement FFT to the data of sensor output, extracts the each physiological signal component under sleep state.

5. system according to claim 4, is characterized in that, described main frame takes redundant computation and credibility to calculate, and to evaluate the reliability of different approaches institute extracting parameter, chooses the result that credibility is greater than predetermined threshold, for sleep architecture stage.

6. system according to claim 5, is characterized in that, the parameter of described sleep architecture stage utilization comprises: the moving information of cardiac cycle, breathing cycle, body and/from bed information;

Described main frame, based on mass data, is found the Sleep architecture dependency rule based on obtaining parameter, sets up sleep stage knowledge base; Application uncertain reasoning theory, merges multiparameter information, carries out the analysis ratiocination of Sleep architecture.

7. system according to claim 6, is characterized in that, described sensing system utilizes Active signal controlling, eliminates the interference of intrinsic vibration cycle to physiological signal in sleep around.

8. system according to claim 7, is characterized in that, when described system detects while occurring that heartbeat is abnormal in sleep or when respiratory arrest, and system automatic alarm.

9. system according to claim 6, is characterized in that, cardiac cycle and sleep level relation mathematic model are as follows:

Middle cycle frequency range heart arteries and veins variation h _u(t), middle cycle frequency range sleep level is passed S _m(t) time difference equation:

${\hat{S}}_m\left(t\right)+a_1{\hat{S}}_m(t-1)+...+a_n{\hat{S}}_m(t-n)=b_0{h}_u\left(t\right)+b_1{h}_u(t-1)+...+b_m{h}_u(t-m)$

Wherein a _nfor the denominator coefficients (n is the natural number between 1 to 5) of difierence equation

B _mfor the denominator coefficients (m is the natural number between 1 to 5) of difierence equation

Calculate ${\hat{S}}_m\left(t\right)=0.996{\hat{S}}_m(t-1)+0.255h_u\left(t\right)-0.259h_u(t-1);$

According to heart arteries and veins, the function His length of one's sleep (k) of continuous sleep level is inferred in variation:

$\mathrm{his}\left(k\right)=N\·\frac{\mathrm{his}\left(k\right)}{S_h};$

Wherein S _hfor the amplitude of fluctuation of sleeping continuously;

Then, ask the minima of regular distribution function:

$\mathrm{Min}\left(\underset{k=0}{\overset{N}{\mathrm{\Σ}}}{\{\mathrm{his}\left(k\right)-\underset{i=1}{\overset{6}{\mathrm{\Σ}}}{w}_i\frac{1}{{\mathrm{\σ}}_i\sqrt{2\mathrm{\π}}}\exp (-\frac{{(k-m_i)}^2}{2{\mathrm{\σ}}_i^2})\}}^2\right)$

Night each sleep level coefficient w _i

w ₆＝0.102，w ₅＝0.161，w ₄＝0.132，

w ₃＝0.472，w ₂＝0.089，w ₁＝0.056，

Calculate continuous sleep level to the sleep probit of 6 grade transformation of standard:

$y_i=e{\left(\frac{{(k-m_i)}^2}{2{\mathrm{\σ}}_i^2}\right)}^2$

Wherein i=1,2 ... 6, m _irepresent the meansigma methods of each sleep level;

σ _irepresent the deviation of each sleep level;

K represents continuous sleep level;

Y _irepresent the probit that each sleep level is corresponding;

And the moving relation of described Sleep stages and body adopts following mathematical model to represent:

$M^{\′}\left(t\right)=\frac{M\left(t\right)-M_{\max }}{M_{\max }-M_{\min }}$

Wherein t express time;

M (t) represents the moving size of body;

M ' (t) represents that standardization body is moving big or small

M _maxrepresent in the moving size of body the meansigma methods of upper 5 numerical value;

M _minrepresent in the moving size of body the meansigma methods of the next 5 numerical value;

A meansigma methods Mo of M (t) while entering bed according to tester and when clear-headed before getting up, M ' (t) determines REM sleep below and awakening judge index Iwr:

Iwr(t)＝M’(t)/Mo

In the time of Iwr (t) > 0.87, be defined as awakening,

When Iwr (, is defined as REM sleep when t the)≤0.87.