CN110151138A - Sleep apnea segment detection method, equipment based on convolutional neural networks - Google Patents
Sleep apnea segment detection method, equipment based on convolutional neural networks Download PDFInfo
- Publication number
- CN110151138A CN110151138A CN201910460074.6A CN201910460074A CN110151138A CN 110151138 A CN110151138 A CN 110151138A CN 201910460074 A CN201910460074 A CN 201910460074A CN 110151138 A CN110151138 A CN 110151138A
- Authority
- CN
- China
- Prior art keywords
- convolutional neural
- neural networks
- segment
- electrocardiosignal
- sleep apnea
- 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.)
- Granted
Links
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 54
- 201000002859 sleep apnea Diseases 0.000 title claims abstract description 46
- 238000001514 detection method Methods 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 16
- 230000016507 interphase Effects 0.000 claims abstract description 11
- 238000006243 chemical reaction Methods 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims description 32
- 238000004519 manufacturing process Methods 0.000 claims description 17
- 238000012360 testing method Methods 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 9
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000012935 Averaging Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 230000000747 cardiac effect Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 abstract description 4
- 208000008784 apnea Diseases 0.000 abstract description 3
- 108091006146 Channels Proteins 0.000 description 13
- 238000003860 storage Methods 0.000 description 5
- 238000003745 diagnosis Methods 0.000 description 4
- 201000010099 disease Diseases 0.000 description 3
- 206010041235 Snoring Diseases 0.000 description 2
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 2
- 230000002146 bilateral effect Effects 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 229910052760 oxygen Inorganic materials 0.000 description 2
- 239000001301 oxygen Substances 0.000 description 2
- 230000000241 respiratory effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 208000024891 symptom Diseases 0.000 description 2
- 206010008132 Cerebral thrombosis Diseases 0.000 description 1
- 206010057315 Daydreaming Diseases 0.000 description 1
- 208000007590 Disorders of Excessive Somnolence Diseases 0.000 description 1
- 206010020772 Hypertension Diseases 0.000 description 1
- 201000001429 Intracranial Thrombosis Diseases 0.000 description 1
- 206010041349 Somnolence Diseases 0.000 description 1
- 210000001015 abdomen Anatomy 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 208000029078 coronary artery disease Diseases 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000003860 sleep quality Effects 0.000 description 1
- 208000011580 syndromic disease Diseases 0.000 description 1
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
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4806—Sleep evaluation
- A61B5/4818—Sleep apnoea
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Abstract
The invention discloses a kind of sleep apnea segment detection method and equipment based on convolutional neural networks, method include: the electrocardiosignal during acquiring subject's nighttime sleep；Analog-to-digital conversion is carried out to collected electrocardiosignal, obtains subject's electrocardiographicdigital digital signals；Electrocardiosignal is divided into one minute segment by minute；RR interphase signal and RR amplitude signal are extracted according to the electrocardiosignal per minute of acquisition；Classified using one-dimensional convolutional neural networks proposed by the present invention to signal.A kind of sleep apnea segment detection method based on convolutional neural networks proposed by the present invention has simple and easy, and accuracy is good, can rapidly and accurately measure the apnea segment of subject, and then realize the early detection to sleep apnea disease.
Description
Technical field
The present invention relates to medical monitoring technical field more particularly to a kind of sleep apneas based on convolutional neural networks
Segment detection method, equipment.
Background technique
There are about 9.36 hundred million people to suffer from sleep apnea (OSA) in the whole world at present, and there are about 60,000,000 people to suffer from such disease in China
Disease, sleep apnea patient are easy to produce the symptoms such as daytime sleepiness, absent minded, and long-term sleep apnea is easy
Cause the diseases such as hypertension, coronary heart disease, cerebral thrombosis.Therefore, the timely diagnosis tool of sleep apnea has very important significance.
Currently, the goldstandard of sleep apnea diagnosis is sleep analysis monitor (PSG), Polysomnography is continuously remembered
Record electrocardiogram (ECG), electroencephalogram (EEG), electroculogram (EOG), electromyogram (EMG), respiratory air flow, blood oxygen saturation, chest and abdomen
The physiological signals such as respiratory movement, the sound of snoring, by manually verifying item by item after record.Chinese 60,000,000 sleep apnea patients, really
Examine less than 1%, receive treatment less than 0.1%, Chinese diagnosis and treatment rate are all too low, wherein the main reason for someone
Be not fully recognized that the harm of sleep apnea, sleep apnea diagnosis is more troublesome, costly, and record
Signal needs doctor largely to be analyzed, and the workload of doctor is bigger.For more fast and convenient completion sleep-respiratory
The preliminary screening of pause, a large amount of researcher explore only with the signal or several letters in electrocardiogram, mouth and nose air-flow, the sound of snoring
Number complete preliminary screening to sleep apnea.1984, it is temporary that Guilleminaultd et al. observes 400 sleep-respiratories
Stop the electrocardiogram and PSG figure of continuous 24 hours of syndrome patients, discovery is when patient apnea symptom occurs and starts, the heart
Rate slows down, and when restoring normal, increased heart rate, proposing heart rate variability can be used to detect sleep apnea syndrome.
It, can also be from the width of electrocardiosignal other than detecting sleep apnea syndrome with the RR interval series extracted from electrocardiosignal
Value variation detection, people breathing when chest electrical impedance due to lung volume variation and change, secondly as heart relative to
The displacement of EEG electrode and the variation in direction, cause heart vector to change.
It is current using single derivative according to carrying out in sleep apnea detection method, most of method all only rests on benefit
Electrocardio is lost when being analyzed with the RR interphase signal extracted from electrocardiosignal, however extracting RR interphase signal from electrocardiosignal
The information of amplitude, the information for not making full use of electrocardiosignal to be included.
Summary of the invention
It is an object of the invention to, there are problem and shortage, provide for above-mentioned, one kind is technically simple easy, and measurement is accurate
With more the sleep apnea piece based on convolutional neural networks based on RR interphase and RR amplitude binary channels sequence of robustness
Section detection method, equipment.
The sleep apnea segment detection method based on convolutional neural networks that the embodiment of the invention provides a kind of, including
Following steps:
Electrocardiosignal when S1, acquisition subject's sleeping at night；
S2, analog-to-digital conversion is carried out to collected electrocardiosignal, obtains the electrocardiographicdigital digital signals of subject；
S3, the electrocardiographicdigital digital signals of acquisition are segmented by minute, obtain electrocardiosignal segment per minute；
S4, the electrocardiosignal snippet extraction cardiac RR intervals signal and RR amplitude signal per minute according to acquisition are utilized
The monitoring result of Polysomnography is labeled electrocardiosignal segment, is believed using the RR interphase signal and RR amplitude of extraction
Number composition double-channel signal make training set；
S5, one-dimensional convolutional neural networks are built, using the double-channel signal production training set of production to one-dimensional convolutional Neural
Network is trained:
S6, classified using trained one-dimensional convolutional neural networks to test signal segment.
Preferably, step S4 is specifically included:
S41, electrocardiographicdigital digital signals are denoted as Q, electrocardiosignal is segmented by minute, n-th minute electrocardiosignal segment note
Make Qn, i.e. Q={ Q1, Q2..., Qn}；
S42, electrocardiosignal segment per minute is filtered, R wave position is then extracted using Pan-Tompki algorithm
It sets and the R wave of extraction is corrected, adjacent the latter R wave subtracts previous R wave and obtains RR interval series as first passage；
The RR amplitude sequence obtained simultaneously according to R wave position is as second channel；
S43, the RR amplitude sequence of the RR interval series of first passage and second channel is filtered respectively, with removal by
The value of noise jamming and the segment for giving up signal difference；
S44, average value processing is carried out to the RR amplitude sequence of the RR interval series of first passage and second channel respectively；
S45, the RR interval series of first passage per minute and the RR amplitude sequence of second channel are filled, are filled
To uniform length k；
S46, electrocardiosignal segment per minute is labeled using the monitoring result of Polysomnography, it will be normal
Segment is labeled as 0, and pause segment is labeled as 1, forms double-channel signal system using the RR interval series and RR amplitude sequence of extraction
Make training set.
Preferably, in step S45, fill method is as follows:
The RR sequence of n-th minute first passage of note is Rn={ r1, r2..., rm, the RR of n-th minute second channel of note
Amplitude sequence is Qn={ q1, q2..., qm, it is filled into below with the data before original series, is filled into uniform length k, fills out
Filling rear result is Wherein R 'nWith Q 'nLength be K.
Preferably, above-mentioned steps S5 is specifically included:
S51, one-dimensional convolutional neural networks are built；
S52, double-channel signal production training set is carried out zero averaging and normalized, converting data to mean value is
0, the data that standard deviation is 1, converting function isWherein μ is the mean value of all training sample data, and σ is all bilaterals
Road signal makes the standard deviation of training set, and the μ and σ that double-channel signal production training set is acquired are for test set, to test set
Carry out zero averaging and normalization；
S53, one-dimensional convolutional neural networks are built, using double-channel signal production training set to one-dimensional convolutional neural networks into
Row training.
The embodiment of the invention also provides a kind of sleep apnea segment detection device based on convolutional neural networks,
It is characterized in that, including memory and processor, executable code is stored in the memory, the executable code can
It is executed by the processor to realize the sleep apnea segment detection method as above-mentioned based on convolutional neural networks.
Compared with prior art, the present invention having the following beneficial effects:
The present invention not only extracts RR interphase signal in processing cardioelectric signals, while also extracting the corresponding amplitude letter of R wave
Breath, is more fully utilized sleep apnea information included in electrocardiosignal, improves sleep apnea segment
The accuracy rate of identification；
The present invention carries out recognition detection to sleep apnea segment using the method for one-dimensional convolutional neural networks, compared to
Traditional machine learning, convolutional neural networks have stronger learning ability, with the increase of data scale, convolutional neural networks
Performance constantly increase.
Meanwhile the present invention does not need additionally to acquire breath signal, blood oxygen signal and brain electricity again after training network
The extras such as signal, it is only necessary to acquire electrocardiosignal.Simultaneously as the network that the present invention uses has powerful function
Can, with the amplification of training set, accuracy can be promoted further, therefore more feasible in terms of realization.
Detailed description of the invention
In order to illustrate more clearly of technical solution of the present invention, attached drawing needed in embodiment will be made below
Simply introduce, it should be apparent that, the accompanying drawings in the following description is only some embodiments of the present invention, general for this field
For logical technical staff, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is the method stream of the segment of one-dimensional convolutional neural networks identification sleep apnea provided in an embodiment of the present invention
Cheng Tu.
Fig. 2 is the double-channel signal waveform diagram of RR interval series and RR amplitude sequence provided in an embodiment of the present invention.
Fig. 3 is one-dimensional convolutional neural networks model schematic provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Referring to Fig. 1, the embodiment of the invention provides a kind of, the sleep apnea segment based on convolutional neural networks is examined
Survey method comprising following steps:
Electrocardiosignal when S1, acquisition subject's sleeping at night.
S2, analog-to-digital conversion is carried out to collected electrocardiosignal, obtains the electrocardiographicdigital digital signals of subject.
S3, the electrocardiographicdigital digital signals of acquisition are segmented by minute, obtain electrocardiosignal segment per minute.
S4, the electrocardiosignal snippet extraction cardiac RR intervals sequence and RR amplitude sequence per minute according to acquisition are utilized
The monitoring result of Polysomnography (PSG) is labeled electrocardiosignal segment, utilizes the RR interval series and RR of extraction
Amplitude sequence forms double-channel signal and makes training set.
Wherein, step S4 is specifically included:
S41, electrocardiographicdigital digital signals are denoted as Q, electrocardiosignal is segmented by minute, n-th minute electrocardiosignal segment note
Make Qn, i.e. Q={ Q1, Q2..., Qn}。
S42, electrocardiosignal segment per minute is filtered, R wave position is then extracted using Pan-Tompki algorithm
It sets and the R wave of extraction is corrected, adjacent the latter R wave subtracts previous R wave and obtains RR interval series as first passage；
The RR amplitude sequence obtained simultaneously according to R wave position is as second channel (as shown in Figure 2).
S43, the RR amplitude sequence of the RR interval series of first passage and second channel is filtered respectively, with removal by
The value of noise jamming and the segment for giving up signal difference.
S44, average value processing is carried out to the RR amplitude sequence of the RR interval series of first passage and second channel respectively.
S45, the RR interval series of first passage per minute and the RR amplitude sequence of second channel are filled, are filled
To uniform length k.
Wherein, specific fill method is as follows:
The RR sequence of n-th minute first passage of note is Rn={ r1, r2..., rm, the RR of n-th minute second channel of note
Amplitude sequence is Qn={ q1, q2..., qm, it is filled into below with the data before original series, is filled into uniform length k, fills out
Filling rear result is Wherein R 'nWith Q 'nLength be K.
S46, electrocardiosignal segment per minute is labeled using the monitoring result of Polysomnography, it will be normal
Segment is labeled as 0, and pause segment is labeled as 1, forms double-channel signal system using the RR interval series and RR amplitude sequence of extraction
Make training set.
That is, the embodiment of the present invention is to extract electrocardiographicdigital digital signals obtaining subject's electrocardiographicdigital digital signals
RR interphase and RR amplitude signal are combined into double-channel signal to train the one-dimensional volume built by RR interphase signal and RR amplitude signal
Product neural network.
S5, one-dimensional convolutional neural networks are built, one-dimensional convolutional neural networks is trained using the training set of production.
Wherein, step S5 is specifically included:
S51, one-dimensional convolutional neural networks are built.
Wherein, the structure of one-dimensional convolutional neural networks is as shown in Figure 3.
S52, double-channel signal production training set is carried out zero averaging and normalized, converting data to mean value is
0, the data that standard deviation is 1, converting function isWherein μ is the mean value of all training sample data, and σ is all bilaterals
Road signal makes the standard deviation of training set, and the μ and σ that double-channel signal production training set is acquired are for test set, to test set
Carry out zero averaging and normalization.
S53, one-dimensional convolutional neural networks are built, using double-channel signal production training set to one-dimensional convolutional neural networks into
Row training.
S6, classified using trained one-dimensional convolutional neural networks to test signal segment.
When it is implemented, electrocardiosignal used in production training set and test set is from Polysomnography (PSG)
The electrocardiosignal of acquisition is segmented obtained electrocardiographicdigital digital signals by minute, by doctor referring to other channel signals of PSG
Electrocardiosignal segment per minute is labeled, extracts RR interphase signal and RR amplitude signal from electrocardiosignal then to make
Training set and test set train the one-dimensional convolutional neural networks built by training set, finally with test set test network
Performance.
In conclusion a kind of sleep apnea segment detection based on convolutional neural networks provided in an embodiment of the present invention
Method is used as training network by the RR interphase and RR amplitude binary channels sequence extracted from electrocardiosignal and knows another characteristic,
Using the present invention, most of noise in electrocardiosignal can be excluded and further retain the sleep for including in electrocardiosignal
Apnea information.Only need to acquire electrocardiosignal can be rapidly completed sleep apnea after the completion of convolutional neural networks training
Identification.Method provided by the invention is simple and easy, and anti-jamming effectiveness is good, easy to operate, can rapidly and accurately measure tested
The sleep quality per minute of person, and then realize the early detection to sleep apnea.
The embodiment of the invention also provides a kind of sleep apnea segment detection device of one-dimensional convolutional neural networks,
Including memory and processor, the processor memory contains executable code, and the executable code can be by the place
Device is managed to execute to realize the method as described in above-described embodiment.
Illustratively, the executable code can be divided into one or more units, one or more of lists
Member is stored in the memory, and is executed by the processor, to complete the present invention.One or more of units can be with
It is a series of executable code instruction segments that can complete specific function, the instruction segment is for describing the executable code one
Tie up the implementation procedure in the sleep apnea segment detection device of convolutional neural networks.
The sleep apnea segment detection device of the one-dimensional convolutional neural networks may include, but are not limited to processor,
Memory.It will be understood by those skilled in the art that the schematic diagram is only the sleep apnea of one-dimensional convolutional neural networks
The example of segment detection device does not constitute the limit to the sleep apnea segment detection device of one-dimensional convolutional neural networks
It is fixed, it may include perhaps combining certain components or different components, such as described one than illustrating more or fewer components
Tie up convolutional neural networks sleep apnea segment detection device can also include input-output equipment, network access equipment,
Bus etc..
Alleged processor can be central processing unit (Central Processing Unit, CPU), can also be it
His general processor, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng the control centre of the sleep apnea segment detection device of the one-dimensional convolutional neural networks utilizes various interfaces and line
Road connects the various pieces of the sleep apnea segment detection device of entire one-dimensional convolutional neural networks.
The memory can be used for storing the executable code and/or module, and the processor is by operation or executes
Executable code in the memory and/or module are stored, and calls the data being stored in memory, described in realization
The various functions of the sleep apnea segment detection device of one-dimensional convolutional neural networks.The memory can mainly include storage
Program area and storage data area, wherein storing program area can application program needed for storage program area, at least one function
(such as sound-playing function, image player function etc.) etc.；Storage data area, which can be stored, uses created number according to mobile phone
According to (such as audio data, phone directory etc.) etc..In addition, memory may include high-speed random access memory, can also include
Nonvolatile memory, such as hard disk, memory, plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), safety
Digital (Secure Digital, SD) card, flash card (Flash Card), at least one disk memory, flush memory device or
Other volatile solid-state parts.
Wherein, if the integrated unit of the sleep apnea segment detection device of the one-dimensional convolutional neural networks is with soft
The form of part functional unit realizes and when sold or used as an independent product, can store and computer-readable deposits at one
In storage media.Based on this understanding, the present invention realizes all or part of the process in above-described embodiment method, can also pass through
Executable code is completed to instruct relevant hardware, and the executable code can be stored in a computer readable storage medium
In, the executable code is when being executed by processor, it can be achieved that the step of above-mentioned each embodiment of the method.Wherein, described to hold
Line code includes executable code code, and the executable code code can be source code form, object identification code form, can hold
Style of writing part or certain intermediate forms etc..The computer-readable medium may include: that can carry the executable code code
Any entity or device, recording medium, USB flash disk, mobile hard disk, magnetic disk, CD, computer storage, read-only memory (ROM,
Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunications letter
Number and software distribution medium etc..It should be noted that the content that the computer-readable medium includes can be managed according to the administration of justice
Local legislation and the requirement of patent practice carry out increase and decrease appropriate, such as in certain jurisdictions, according to legislation and patent
Practice, computer-readable medium does not include electric carrier signal and telecommunication signal.
It should be noted that the apparatus embodiments described above are merely exemplary, wherein described be used as separation unit
The unit of explanation may or may not be physically separated, and component shown as a unit can be or can also be with
It is not physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to actual
It needs that some or all of the modules therein is selected to achieve the purpose of the solution of this embodiment.In addition, device provided by the invention
In embodiment attached drawing, the connection relationship between module indicate between them have communication connection, specifically can be implemented as one or
A plurality of communication bus or signal wire.Those of ordinary skill in the art are without creative efforts, it can understand
And implement.
The above is a preferred embodiment of the present invention, it is noted that for those skilled in the art
For, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also considered as
Protection scope of the present invention.
Claims (5)
1. a kind of sleep apnea segment detection method based on convolutional neural networks, which comprises the following steps:
Electrocardiosignal when S1, acquisition subject's sleeping at night；
S2, analog-to-digital conversion is carried out to collected electrocardiosignal, obtains the electrocardiographicdigital digital signals of subject；
S3, the electrocardiographicdigital digital signals of acquisition are segmented by minute, obtain electrocardiosignal segment per minute；
S4, the electrocardiosignal snippet extraction cardiac RR intervals signal and RR amplitude signal per minute according to acquisition, using leading more
The monitoring result of sleep monitor is labeled electrocardiosignal segment, utilizes the RR interphase signal and RR amplitude signal group of extraction
Training set is made at double-channel signal；
S5, one-dimensional convolutional neural networks are built, using the double-channel signal production training set of production to one-dimensional convolutional neural networks
It is trained；
S6, classified using trained one-dimensional convolutional neural networks to test signal segment.
2. a kind of sleep apnea segment detection method based on convolutional neural networks according to claim 1, feature
It is, step S4 is specifically included:
S41, electrocardiographicdigital digital signals being denoted as Q, electrocardiosignal is segmented by minute, n-th minute electrocardiosignal segment is denoted as Qn,
That is Q={ Q1, Q2..., Qn}；
S42, electrocardiosignal segment per minute is filtered, R wave position pair is then extracted using Pan-Tompki algorithm
The R wave of extraction is corrected, and adjacent the latter R wave subtracts previous R wave and obtains RR interval series as first passage；Simultaneously
The RR amplitude sequence obtained according to R wave position is as second channel；
S43, the RR amplitude sequence of the RR interval series of first passage and second channel is filtered respectively, to remove by noise
The value of interference and the segment for giving up signal difference；
S44, average value processing is carried out to the RR amplitude sequence of the RR interval series of first passage and second channel respectively；
S45, the RR interval series of first passage per minute and the RR amplitude sequence of second channel are filled, are filled into system
One length k；
S46, electrocardiosignal segment per minute is labeled using the monitoring result of Polysomnography, by normal segment
It is labeled as 0, pause segment is labeled as 1, utilizes RR interval series and RR amplitude sequence the composition double-channel signal production instruction of extraction
Practice collection.
3. the sleep apnea segment detection method of one-dimensional convolutional neural networks according to claim 2, feature exist
In in step S45, fill method is as follows:
The RR sequence of n-th minute first passage of note is Rn={ r1, r2..., rm, the RR amplitude of n-th minute second channel of note
Sequence is Qn={ q1, q2..., qm, it is filled into below with the data before original series, uniform length k is filled into, after filling
As a result it is Wherein R 'nWith Q 'nLength be K.
4. a kind of sleep apnea segment detection method based on convolutional neural networks according to claim 1, feature
It is that above-mentioned steps S5 is specifically included:
S51, one-dimensional convolutional neural networks are built；
S52, double-channel signal production training set is carried out zero averaging and normalized, converting data to mean value is 0, mark
The data that quasi- difference is 1, conversion function areWherein μ is the mean value of all training sample data, and σ is all binary channels letter
The standard deviation of number production training set carries out test set the double-channel signal production training set μ that acquires and σ for test set
Zero averaging and normalization；
S53, one-dimensional convolutional neural networks are built, one-dimensional convolutional neural networks is instructed using double-channel signal production training set
Practice.
5. a kind of sleep apnea segment detection device based on convolutional neural networks, which is characterized in that including memory with
And processor, executable code is stored in the memory, the executable code can be executed by the processor with reality
The now sleep apnea segment detection method of the convolutional neural networks as described in Claims 1-4 any one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910460074.6A CN110151138B (en) | 2019-05-29 | 2019-05-29 | Sleep apnea fragment detection method and device based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910460074.6A CN110151138B (en) | 2019-05-29 | 2019-05-29 | Sleep apnea fragment detection method and device based on convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110151138A true CN110151138A (en) | 2019-08-23 |
| CN110151138B CN110151138B (en) | 2021-08-24 |
Family
ID=67629857
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910460074.6A Active CN110151138B (en) | 2019-05-29 | 2019-05-29 | Sleep apnea fragment detection method and device based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110151138B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110801221A (en) * | 2019-12-09 | 2020-02-18 | 中山大学 | Sleep apnea fragment detection method and device based on unsupervised feature learning |
| CN110811591A (en) * | 2019-10-31 | 2020-02-21 | 太原理工大学 | Heart failure grading method based on heart rate variability |
| CN111166294A (en) * | 2020-01-29 | 2020-05-19 | 北京交通大学 | Automatic sleep apnea detection method and device based on inter-heartbeat period |
| CN111297327A (en) * | 2020-02-20 | 2020-06-19 | 京东方科技集团股份有限公司 | Sleep analysis method, system, electronic equipment and storage medium |
| CN113243890A (en) * | 2021-05-10 | 2021-08-13 | 清华大学深圳国际研究生院 | Sleep apnea syndrome recognition device |
| CN113273992A (en) * | 2021-05-11 | 2021-08-20 | 清华大学深圳国际研究生院 | Signal processing method and device |
| CN113397556A (en) * | 2021-07-22 | 2021-09-17 | 福州大学至诚学院 | Method for classifying sleep disordered breathing events based on sleep data |
| CN113693608A (en) * | 2021-08-04 | 2021-11-26 | 福州大学至诚学院 | Detection method for diagnosing sleep disorder based on cloud platform |
| CN114073495A (en) * | 2021-12-03 | 2022-02-22 | 中山大学·深圳 | Sleep apnea detection method and device based on information similarity |
| TWI784513B (en) * | 2021-05-11 | 2022-11-21 | 國立勤益科技大學 | Method of monitoring apnea events based on electrocardiogram delayed reactions |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103479349A (en) * | 2013-09-25 | 2014-01-01 | 深圳市理邦精密仪器股份有限公司 | Electrocardiosignal data acquisition and processing method and system |
| US20150164410A1 (en) * | 2013-12-13 | 2015-06-18 | Vital Connect, Inc. | Sleep apnea syndrome (sas) screening using wearable devices |
| US20160287122A1 (en) * | 2015-03-30 | 2016-10-06 | Resmed Sensor Technologies Limited | Detection of periodic breathing |
| CN106361277A (en) * | 2016-08-26 | 2017-02-01 | 中山大学 | Sleep apnea syndrome assessment method based on electrocardiogram signals |
| CN109350014A (en) * | 2018-12-10 | 2019-02-19 | 苏州小蓝医疗科技有限公司 | A kind of sound of snoring recognition methods and system |
| CN109620208A (en) * | 2018-12-29 | 2019-04-16 | 南京茂森电子技术有限公司 | Sleep Apnea-hypopnea Syndrome detection system and method |
-
2019
- 2019-05-29 CN CN201910460074.6A patent/CN110151138B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103479349A (en) * | 2013-09-25 | 2014-01-01 | 深圳市理邦精密仪器股份有限公司 | Electrocardiosignal data acquisition and processing method and system |
| US20150164410A1 (en) * | 2013-12-13 | 2015-06-18 | Vital Connect, Inc. | Sleep apnea syndrome (sas) screening using wearable devices |
| US20160287122A1 (en) * | 2015-03-30 | 2016-10-06 | Resmed Sensor Technologies Limited | Detection of periodic breathing |
| CN106361277A (en) * | 2016-08-26 | 2017-02-01 | 中山大学 | Sleep apnea syndrome assessment method based on electrocardiogram signals |
| CN109350014A (en) * | 2018-12-10 | 2019-02-19 | 苏州小蓝医疗科技有限公司 | A kind of sound of snoring recognition methods and system |
| CN109620208A (en) * | 2018-12-29 | 2019-04-16 | 南京茂森电子技术有限公司 | Sleep Apnea-hypopnea Syndrome detection system and method |
Non-Patent Citations (1)
| Title |
|---|
| AHSAN H.KHANDOKER等: "Support Vector Machines for Automated Recognition of Obstructive Sleep Apnea Syndrome From ECG Recordings", 《IEEE TRANSACTIONS ON INFORMATION TECHNOLOGY IN BIOMEDICINE》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110811591A (en) * | 2019-10-31 | 2020-02-21 | 太原理工大学 | Heart failure grading method based on heart rate variability |
| CN110801221B (en) * | 2019-12-09 | 2021-04-27 | 中山大学 | Sleep apnea fragment detection equipment based on unsupervised feature learning |
| CN110801221A (en) * | 2019-12-09 | 2020-02-18 | 中山大学 | Sleep apnea fragment detection method and device based on unsupervised feature learning |
| CN111166294A (en) * | 2020-01-29 | 2020-05-19 | 北京交通大学 | Automatic sleep apnea detection method and device based on inter-heartbeat period |
| CN111297327B (en) * | 2020-02-20 | 2023-12-01 | 京东方科技集团股份有限公司 | Sleep analysis method, system, electronic equipment and storage medium |
| CN111297327A (en) * | 2020-02-20 | 2020-06-19 | 京东方科技集团股份有限公司 | Sleep analysis method, system, electronic equipment and storage medium |
| CN113243890A (en) * | 2021-05-10 | 2021-08-13 | 清华大学深圳国际研究生院 | Sleep apnea syndrome recognition device |
| CN113273992A (en) * | 2021-05-11 | 2021-08-20 | 清华大学深圳国际研究生院 | Signal processing method and device |
| TWI784513B (en) * | 2021-05-11 | 2022-11-21 | 國立勤益科技大學 | Method of monitoring apnea events based on electrocardiogram delayed reactions |
| CN113397556A (en) * | 2021-07-22 | 2021-09-17 | 福州大学至诚学院 | Method for classifying sleep disordered breathing events based on sleep data |
| CN113693608A (en) * | 2021-08-04 | 2021-11-26 | 福州大学至诚学院 | Detection method for diagnosing sleep disorder based on cloud platform |
| CN114073495A (en) * | 2021-12-03 | 2022-02-22 | 中山大学·深圳 | Sleep apnea detection method and device based on information similarity |
| CN114073495B (en) * | 2021-12-03 | 2024-03-22 | 中山大学·深圳 | Sleep apnea detection method and device based on information similarity |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110151138B (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110151138A (en) | Sleep apnea segment detection method, equipment based on convolutional neural networks | |
| US11234629B2 (en) | Method and device for self-learning dynamic electrocardiography analysis employing artificial intelligence | |
| CN110811609B (en) | Epileptic spike intelligent detection device based on self-adaptive template matching and machine learning algorithm fusion | |
| US11517212B2 (en) | Electrocardiogram information dynamic monitoring method and dynamic monitoring system | |
| Banluesombatkul et al. | Single channel ECG for obstructive sleep apnea severity detection using a deep learning approach | |
| CN106214145B (en) | Electrocardiogram classification method based on deep learning algorithm | |
| Perlman et al. | Electromyographic activity from human laryngeal, pharyngeal, and submental muscles during swallowing | |
| CN110801221B (en) | Sleep apnea fragment detection equipment based on unsupervised feature learning | |
| US11783924B2 (en) | ECG information processing method and ECG workstation | |
| US11419542B2 (en) | System and method for non-apnea sleep arousal detection | |
| CN109106350A (en) | The preliminary screening method of the Sleep Apnea-hypopnea Syndrome of low physiological stress | |
| CN103919548B (en) | Detecting device and detecting equipment for swallowing muscle disorders | |
| CN103153183A (en) | Apparatus and method for diagnosing obstructive sleep apnea | |
| CN109394188A (en) | A kind of adnormal respiration detection method, device and equipment based on heart rate variability | |
| CN109288515B (en) | Periodicity monitoring method and device based on premature beat signal in wearable electrocardiosignal | |
| CN113095302A (en) | Depth model for arrhythmia classification, method and apparatus using the same | |
| Naseer et al. | Classification of normal and abnormal ECG signals based on their PQRST intervals | |
| Sansone et al. | Influence of QT correction on temporal and amplitude features for human identification via ECG | |
| Bhattacharjee et al. | SleepTight: identifying sleep arousals using inter and intra-relation of multimodal signals | |
| CN115336973A (en) | Sleep staging system construction method based on self-attention mechanism and single-lead electrocardiosignal and sleep staging system | |
| TWI777650B (en) | A method of monitoring apnea and hypopnea events based on the classification of the descent rate of heartbeat intervals | |
| CN106377250B (en) | The tagging equipment of EEG signals data type under sleep state | |
| CN109044338A (en) | atrial fibrillation detection device and storage medium | |
| CN109394197A (en) | A kind of heart rate variability measurement method, device and equipment based on time frequency analysis | |
| CN115040074A (en) | Obstructive sleep apnea detection method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231121 Address after: No. 66, Gongchang Road, Xinhu Street, Guangming District, Shenzhen, Guangdong 518000 Patentee after: Shenzhen, Zhongshan University Patentee after: SUN YAT-SEN University Address before: 510275 No. 135 West Xingang Road, Guangdong, Guangzhou Patentee before: SUN YAT-SEN University |