CN113974576A - Sleep quality monitoring system and monitoring method based on magnetocardiogram - Google Patents
Sleep quality monitoring system and monitoring method based on magnetocardiogram Download PDFInfo
- Publication number
- CN113974576A CN113974576A CN202111589528.3A CN202111589528A CN113974576A CN 113974576 A CN113974576 A CN 113974576A CN 202111589528 A CN202111589528 A CN 202111589528A CN 113974576 A CN113974576 A CN 113974576A
- Authority
- CN
- China
- Prior art keywords
- signal
- magnetocardiogram
- signals
- blood oxygen
- characteristic
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 75
- 230000003860 sleep quality Effects 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 50
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical group [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims abstract description 58
- 239000008280 blood Substances 0.000 claims abstract description 58
- 210000004369 blood Anatomy 0.000 claims abstract description 58
- 229910052760 oxygen Inorganic materials 0.000 claims abstract description 58
- 239000001301 oxygen Substances 0.000 claims abstract description 58
- 230000000241 respiratory effect Effects 0.000 claims abstract description 35
- 238000013441 quality evaluation Methods 0.000 claims abstract description 20
- 238000012216 screening Methods 0.000 claims abstract description 20
- 230000008878 coupling Effects 0.000 claims abstract description 19
- 238000010168 coupling process Methods 0.000 claims abstract description 19
- 238000005859 coupling reaction Methods 0.000 claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 238000010801 machine learning Methods 0.000 claims abstract description 12
- 230000002612 cardiopulmonary effect Effects 0.000 claims abstract description 11
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000004622 sleep time Effects 0.000 claims abstract description 4
- 230000005236 sound signal Effects 0.000 claims description 38
- 238000000354 decomposition reaction Methods 0.000 claims description 24
- 238000012545 processing Methods 0.000 claims description 24
- 230000006870 function Effects 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 8
- 238000012706 support-vector machine Methods 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 7
- 238000012952 Resampling Methods 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 238000001228 spectrum Methods 0.000 claims description 4
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 238000003672 processing method Methods 0.000 claims description 3
- 230000036578 sleeping time Effects 0.000 claims description 3
- 230000002802 cardiorespiratory effect Effects 0.000 claims description 2
- 238000005457 optimization Methods 0.000 claims description 2
- 230000007958 sleep Effects 0.000 description 37
- 230000029058 respiratory gaseous exchange Effects 0.000 description 20
- 230000005291 magnetic effect Effects 0.000 description 14
- 230000008569 process Effects 0.000 description 6
- 206010041235 Snoring Diseases 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 239000000523 sample Substances 0.000 description 5
- 238000003745 diagnosis Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 210000004072 lung Anatomy 0.000 description 3
- 206010020772 Hypertension Diseases 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000035790 physiological processes and functions Effects 0.000 description 2
- 238000001303 quality assessment method Methods 0.000 description 2
- 210000000614 rib Anatomy 0.000 description 2
- 208000019116 sleep disease Diseases 0.000 description 2
- 230000008667 sleep stage Effects 0.000 description 2
- 208000020685 sleep-wake disease Diseases 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 210000000779 thoracic wall Anatomy 0.000 description 2
- 210000001519 tissue Anatomy 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- 241000321096 Adenoides Species 0.000 description 1
- 206010002383 Angina Pectoris Diseases 0.000 description 1
- 206010058314 Dysplasia Diseases 0.000 description 1
- 208000017701 Endocrine disease Diseases 0.000 description 1
- 206010021143 Hypoxia Diseases 0.000 description 1
- 208000020358 Learning disease Diseases 0.000 description 1
- 208000032140 Sleepiness Diseases 0.000 description 1
- 206010041349 Somnolence Diseases 0.000 description 1
- 206010042434 Sudden death Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 210000002534 adenoid Anatomy 0.000 description 1
- 230000006793 arrhythmia Effects 0.000 description 1
- 206010003119 arrhythmia Diseases 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 206010006514 bruxism Diseases 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007211 cardiovascular event Effects 0.000 description 1
- 210000003710 cerebral cortex Anatomy 0.000 description 1
- 230000001684 chronic effect Effects 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000002124 endocrine Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000007954 hypoxia Effects 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 201000003723 learning disability Diseases 0.000 description 1
- 239000000696 magnetic material Substances 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 208000030159 metabolic disease Diseases 0.000 description 1
- 208000010125 myocardial infarction Diseases 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 208000002815 pulmonary hypertension Diseases 0.000 description 1
- 230000037321 sleepiness Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000002618 waking effect Effects 0.000 description 1
Images
Classifications
-
- 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
- A61B5/4815—Sleep quality
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration or pH-value ; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid or cerebral tissue
- A61B5/14542—Measuring characteristics of blood in vivo, e.g. gas concentration or pH-value ; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid or cerebral tissue for measuring blood gases
-
- 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
- A61B5/242—Detecting biomagnetic fields, e.g. magnetic fields produced by bioelectric currents
- A61B5/243—Detecting biomagnetic fields, e.g. magnetic fields produced by bioelectric currents specially adapted for magnetocardiographic [MCG] signals
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Optics & Photonics (AREA)
- Cardiology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Power Engineering (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention relates to a sleep quality monitoring system and method based on magnetocardiogram, the monitoring method comprises: s10, acquiring magnetocardiogram signals and blood oxygen signals of the sleep time of the target user by means of a monitoring system; s20, preprocessing and denoising the magnetocardiogram signals, screening available magnetocardiogram signal time periods and obtaining intermediate magnetocardiogram signals; s30, extracting the middle magnetocardiogram signal by adopting a QRS wave detection method and a cardiopulmonary coupling algorithm, and acquiring a characteristic signal for reflecting the sleep quality of a target user; s40, preprocessing the blood oxygen signal, extracting the characteristic value, and screening a time period capable of reflecting the respiratory event of the target user by adopting a pre-trained machine learning model; and S50, acquiring a sleep quality evaluation result of the target user based on the characteristic value and the characteristic signal of the blood oxygen signal in the screened time period. The method can monitor the respiratory events in a time sequence manner, and improves the detection accuracy and the reliability.
Description
Technical Field
The invention relates to the technical field of biomedicine, in particular to a sleep quality monitoring system and a sleep quality monitoring method based on magnetocardiogram.
Background
Sleep Disordered Breathing (SDB) is a disease which is mainly characterized by snoring in sleep at night and sleepiness in daytime, and as a patient can repeatedly stop breathing during the snoring, the patient can repeatedly arouse the cerebral cortex, the oxygen content in blood is reduced, and the important organs of the brain, the heart and the like are subjected to chronic hypoxia. When the human body is chronically anoxic for a long time, the attack of hypertension, arrhythmia, myocardial infarction and angina pectoris can be induced, and more seriously, sudden death can be caused. If the Sleep Disordered Breathing (SDB) of the children cannot be diagnosed and effectively intervened in time, a series of serious complications can be caused, such as maxillofacial dysplasia (adenoid face-face appearance), abnormal behaviors, learning disorder, backward growth and development, neurocognitive injury, endocrine and metabolic disorders, hypertension and pulmonary hypertension, and even the risk of cardiovascular events in adult stages is increased.
In the sleep quality monitoring process, in order to ensure the low-load sleep process, the respiratory signal is not directly acquired generally, and other signals capable of reflecting the sleep state are used for replacing, such as electrocardiosignals, blood oxygen signals, body position information measured by an accelerometer and the like. The prior art provides a method which adopts an accelerometer to monitor the movement of a user, judges the sleeping depth according to the micro movement of the user, and approximately records the time of falling asleep and waking of the user. Due to the problem of diagnosis precision, all the sensors cannot be compared with the gold standard PSG (sleep monitor guidance), and especially for sleep monitor of children Sleep Disordered Breathing (SDB), all portable devices cannot meet the diagnosis standard.
The PSG with the gold standard has high requirements on technicians, is complex to operate, is expensive, and cannot be popularized and used, so that a sleep quality monitoring system with low cost and simple and convenient operation needs to be provided, and the sleep state of a user can be monitored in real time, and the sleep quality of the user can be scored.
Disclosure of Invention
Technical problem to be solved
In view of the above disadvantages and shortcomings of the prior art, the present invention provides a sleep quality monitoring system and method based on magnetocardiogram.
(II) technical scheme
In order to achieve the purpose, the invention adopts the main technical scheme that:
in a first aspect, an embodiment of the present invention provides a sleep quality monitoring method based on a magnetocardiogram, including:
s10, acquiring magnetocardiogram signals and blood oxygen signals of the sleep time of the target user by means of a monitoring system;
s20, preprocessing and denoising the magnetocardiogram signals, screening available magnetocardiogram signal time periods and obtaining intermediate magnetocardiogram signals;
s30, extracting the middle magnetocardiogram signal by adopting a QRS wave detection method and a cardiopulmonary coupling algorithm, and acquiring a characteristic signal for reflecting the sleep quality of a target user;
s40, preprocessing the blood oxygen signal, extracting a characteristic value, and screening a time period capable of reflecting the respiratory event of the target user based on the characteristic value by adopting a pre-trained machine learning model;
and S50, acquiring a sleep quality evaluation result of the target user based on the characteristic value of the blood oxygen signal in the screened time period and the characteristic signal in the screened time period.
Optionally, the method further comprises:
s60, acquiring sound signals and/or image signals of the sleeping time of the target user by means of a monitoring system;
processing the sound signals by adopting another machine learning model trained in advance, screening effective sound signals, and pairing time periods corresponding to the effective sound signals with the time periods of the characteristic signals/the screened time periods;
if the time is overlapped, analyzing the characteristic signal, the blood oxygen signal and the sound signal of the overlapped time period to obtain a sleep quality evaluation result of the target user;
or analyzing the characteristic signals, the blood oxygen signals, the sound signals and the image signals of the overlapped time periods to obtain the sleep quality evaluation result of the target user.
Optionally, the S20 includes:
and optimizing the magnetocardiogram signals by adopting the quality factors of the magnetocardiogram signals, and performing noise reduction processing on the optimized magnetocardiogram signals by using a soft thresholding processing method based on empirical mode decomposition to obtain intermediate magnetocardiogram signals.
Optionally, the S20 includes:
s21, segmenting the magnetocardiogram signals based on the sliding window of 1 minute, adopting 'harr' wavelet basis to carry out 7-layer wavelet packet decomposition on the magnetocardiogram signals in each window, and carrying out 7-layer wavelet packet decomposition on the magnetocardiogram signals with the sampling rate of 256Hz
Decomposing in 6 sub-bands, calculating
Layer 7 in minutes
The proportion of wavelet in the energy of the layer
，
Wherein,
is a magnetocardiogram signal
Within minute is
The energy of the individual subspaces;
s22, extracting magnetocardiogram signals respectively
Wavelet energy ratio of 6 sub-bands in minutes
，
Entropy of energy
And kurtosis of magnetocardiogram signals
A total of 13 eigenvalues are used as quality factors for evaluating the quality of magnetocardiogram signals,
is as follows
The average of the sampled points in minutes,
is as follows
Standard deviation of sampling points in minutes;
s23, inputting the extracted 13 quality factors into a trained support vector machine to sequentially judge hearts per minuteMagnetic signal quality, screening time periods for sleep quality assessment to obtain quality label of magnetocardiogram signal
And quality optimized magnetocardiogram signal
；
S24, decomposing the quality-optimized magnetocardiogram signals into a plurality of intrinsic mode functions IMFs in a self-adaptive manner based on signal local characteristics by using an EMD method, using multiples of a Donoho threshold as a threshold, soft-thresholding the obtained IMFs, superposing and reconstructing the soft-thresholded IMFs, removing noise of the magnetocardiogram signals, and obtaining intermediate magnetocardiogram signals
。
Optionally, the S30 includes:
s31, performing square operation on each item of the middle magnetocardiogram signal, increasing the intensity value and increasing the high-frequency component to obtain a first signal
For the first signal
Carrying out three-time moving average filtering to respectively extract QRS characteristics
QRS threshold
And baseline noise level in the signal
(ii) a Calculating a QRS threshold that includes an offset,
；
s32, obtaining the positions of all points with QRS characteristics larger than the QRS threshold value according to the threshold value method, marking the positions of the points with 0 and 1,
the duration of the QRS characteristic at the QRS wave position being larger than the threshold value is larger than
Removing T wave, P wave and noise as criterion to obtain position function representing QRS waveT（n）；
S33, extracting RR interval sequence of intermediate signal according to extracted QRS wave
And respiratory signals
；
S34, using cubic spline interpolation method to sequence RR intervals
And respiratory signals
Resampling to 4Hz, then segmenting the resampled respiratory signal using a 3min sliding window with an overlap rate of 25%, calculating the cardiopulmonary coupling strength in each window
；
Extracting the cardiorespiratory coupling strength in each window
The normalized low-frequency power LFC-N and the ratio LVHC of the low-frequency power to the high-frequency power of LFC-N are used as the characteristic signal of the magnetocardiogram signal.
Optionally, the S40 includes:
s41, preprocessing the blood oxygen signal and extracting a characteristic value;
specifically, removing the artifact of the blood oxygen signal, resampling the signal to 25Hz, and keeping two decimal places;
segmenting the signal by using a 3min sliding window with the overlapping rate of 25%, extracting a characteristic value of the blood oxygen signal, and extracting the ODI3 of the blood oxygen signal and the frequency spectrum skewness of the interested frequency band as the characteristic value capable of reflecting the respiratory event;
and S42, inputting the extracted characteristic values into a machine learning model trained in advance, and screening time periods capable of reflecting the respiratory events.
Optionally, S50 includes:
quality label based on magnetocardiogram signals
Judging whether the characteristic signals in the time period accord with a preset available standard or not, and if so, acquiring a sleep quality evaluation result of the target user according to the characteristic value of the blood oxygen signals in the screened time period and the characteristic signals in the screened time period;
otherwise, acquiring a sleep quality evaluation result of the target user according to the characteristic value of the blood oxygen signal of the screened time period.
Optionally, the S60 includes:
respectively denoising the sound signal and the image signal by using a Butterworth filter and dividing the signals by using a 3min sliding window with an overlapping rate of 25%;
for the image signal of each segmented window, extracting dynamic features by using a convolutional neural network;
and for the sound signals of each divided window, judging whether the period of time exceeds the sound signals of the threshold value by using a fully-connected neural network.
Optionally, the S23 includes:
s23-1, initialTransformingi=1First, the quality of the 1 st minute signal is evaluated;
s23-2, the first stepiInputting 13 feature values of minutes into the trained support vector machine model, and outputting the result asqu i =1Orqu i =0Signal quality label
If outputqu i =1Then represents the firstiThe minute signal is available, execute S23-3; if outputqu i =0Then represents the firstiThe signal of minute is not available, the signal quality of the next minute is continuously judged until the available signal of the first section is found, and the order is giveni=i+1Repeating S23-2;
s23-3, ifiIf minute signal is available, determining the first available signal, assigning the value of the first available signal to each unavailable signal, and optimizing magnetocardiogram signal
，
Wherein
Continuing to judge the next section of signal, order
；
S23-4, the first step
Inputting 13 characteristic values of minutes into the trained support vector machine model to obtain
Value of (2), signal quality label
(ii) a If it is
Then represents the first
The minute signal can be used, the optimized magnetocardiogram signal is the same as the original signal,
(ii) a If it is
Then represents the firstiMinute signal is not available, willi-1Assigning the minute signal to the current signal, and optimizing the magnetocardiogram signal
Alternatively, the S24 includes:
inputting the quality-optimized magnetocardiogram signal
Performing EMD decomposition and initializing iteration times
And residual terms
And c represents the number of iterative decompositions,
residual term determination before starting of the c-th decomposition
All poles, all maximum points of
Upper envelope curve of
All minimum points forming the lower envelope
The mean line of the signal is obtained according to the upper and lower envelope lines
(ii) a Subtracting the envelope mean line from the residual term
(ii) a If it is
The zero point and the extreme point have equal number or differ by at most 1, and the average value of the envelope curve formed by the extreme points at any time is zero, then
As an IMF value to
And outputting and updating the number of iterations
Up to
Satisfying the criterion of iteration termination, outputting residual terms
Completing EMD decomposition;
after EMD decomposition
Wherein
The intrinsic mode component IMF is represented,
representing a residual term or a trend term;
using a multiple of the Donoho threshold as the threshold:
where C is a constant and N is the data length; different IMFs have different energies, different scale factors C are selected for the different IMFs, and 0.6-1.2 are sequentially selected;
each intrinsic mode is subjected to soft thresholding, and the w mode after the soft thresholding
Overlapping and reconstructing the IMF subjected to soft thresholding to obtain a denoised magnetocardiogram signal
。
In a second aspect, an embodiment of the present invention further provides a sleep quality monitoring system based on magnetocardiogram, which includes:
a magnetocardiogram monitoring structure for acquiring magnetocardiogram signals;
a blood oxygen collecting component for collecting blood oxygen signals;
the signal processing device is connected with the magnetocardiogram monitoring structure and the blood oxygen acquisition assembly;
the magnetocardiogram monitoring structure comprises: the magnetocardiogram acquisition waistcoat is internally provided with a magnetocardiogram sensor and is worn on the upper body of a target user; the signal processing device executes the sleep quality monitoring method based on the magnetocardiogram according to any one of the first aspect;
or,
the sleep quality monitoring system comprises:
a magnetocardiogram monitoring structure for acquiring magnetocardiogram signals;
a blood oxygen collecting component for collecting blood oxygen signals;
a video acquisition component for acquiring image signals;
a sound collection assembly for collecting sound signals;
the signal processing device is connected with the magnetocardiogram monitoring structure and the blood oxygen acquisition assembly; the video acquisition assembly and the sound acquisition assembly are both connected with the signal processing device;
the signal processing device executes the sleep quality monitoring method based on the magnetocardiogram according to any one of the first aspect.
(III) advantageous effects
According to the method, the portable magnetocardiogram sensor is used for obtaining the magnetocardiogram signals of the target user sleeping all night, then the magnetocardiogram signals are optimized according to the quality factor of the magnetocardiogram signals, the magnetocardiogram signals with poor quality are marked, after the noise of the magnetocardiogram signals is removed, the long-time magnetocardiogram signal QRS wave detection method is adopted for carrying out QRS wave detection on the denoised magnetocardiogram signals, the QRS wave detection accuracy is improved, the respiratory events can be monitored in a time sequence, and the detection accuracy and the reliability are improved.
Further, in the method of the present invention, the trained machine learning model can be used to detect the sleep quality and respiratory events of each period of time in combination with the blood oxygen signal and the image signal, and then the sleep quality is scored and the sleep disorder is assisted to be evaluated.
In an automated signal processing process: the cardiopulmonary coupling is combined with the blood oxygen signal, the image signal and the sound signal, a trained machine learning model is used for detecting the sleep quality and the respiratory event in each period of time, and the sleep quality is scored and the sleep disorder is assisted to be evaluated on the basis of the detection, so that a sleep quality report is generated.
The signal acquisition is more targeted: the target user can select the required sensor according to the requirement, and the daily sleep quality monitoring only needs the magnetocardiogram sensor; if the user suspects that the user has the Sleep Disordered Breathing (SDB) or recommends further Sleep Disordered Breathing (SDB) examination after daily sleep quality monitoring, the user can add an oximeter to accurately detect whether the user has the Sleep Disordered Breathing (SDB) and the severity of the Sleep Disordered Breathing (SDB); if the user wants to record the detailed information of the user during the sleep period and carry out fine sleep monitoring, the image acquisition system can be continuously added to obtain a detailed sleep monitoring report.
Drawings
Fig. 1 is a schematic structural diagram of a sleep quality monitoring system based on a magnetocardiogram according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a sleep quality monitoring method of a sleep quality monitoring system based on a magnetocardiogram according to an embodiment;
fig. 3 is a schematic diagram of a sleep quality monitoring method of a sleep quality monitoring system based on a magnetocardiogram according to another embodiment.
Detailed Description
For the purpose of better explaining the present invention and to facilitate understanding, the present invention will be described in detail by way of specific embodiments with reference to the accompanying drawings.
The magnetocardiogram signal contains more comprehensive physiological information than the electrocardio signal, and can reflect the physiological state of the heart. Compared with electrocardiogram, the magnetic signal is more stable than the electric signal, because the human tissue belongs to non-magnetic substance, the magnetic conductivity is approximately the same as that in vacuum, so the magnetic signal is not interfered by the medium of lung, chest wall, rib, etc., and the obtained result is more reliable. In addition, for the current with the same annular current and the same magnitude and the opposite direction, the electric effects are mutually counteracted and cannot be displayed on the electrocardiogram, but the obvious magnetic effect is generated and displayed on the electrocardiogram. Therefore, the study on sleep quality by using a magnetic field precision measurement method to measure magnetocardiogram instead of traditional electrocardio is a future development trend. In addition, the accuracy of monitoring the body movement in the sleep by using the accelerometer is greatly controversial, and the body movement and the snore in the sleep process can be better monitored by adopting image acquisition equipment comprising a camera and a microphone.
That is, the magnetocardiogram signal contains more comprehensive physiological information than the electrocardiographic signal, and can reflect the physiological state of the heart. Compared with electrocardiogram, the magnetic signal is more stable than the electric signal, because the human tissue belongs to non-magnetic substance, the magnetic conductivity is approximately the same as that in vacuum, so the magnetic signal is not interfered by the medium of lung, chest wall, rib, etc., and the obtained result is more reliable. In addition, for the current with the same annular current and the same magnitude and the opposite direction, the electric effects are mutually counteracted and cannot be displayed on the electrocardiogram, but the obvious magnetic effect is generated and displayed on the electrocardiogram. Combining magnetocardiogram signals with a CPC algorithm will improve the accuracy of the diagnosis of Sleep Disordered Breathing (SDB) in children.
Example one
As shown in fig. 1, the present embodiment provides a schematic structural diagram of a sleep quality monitoring system based on a magnetocardiogram, and the sleep quality monitoring system based on a magnetocardiogram of the present embodiment may include: a magnetocardiogram monitoring structure for acquiring magnetocardiogram signals; a blood oxygen collecting component for collecting blood oxygen signals; a signal processing device.
The signal processing device 6 (which may be a computer, for example) is connected to the magnetocardiogram monitoring structure and the blood oxygen collection assembly, and may perform the magnetocardiogram-based sleep quality monitoring method described below with reference to fig. 2 and 3.
The magnetocardiogram monitoring structure of the present embodiment includes: the magnetocardiogram acquisition waistcoat 1 is used for being worn on the upper body of a target user and internally provided with a magnetocardiogram sensor 2;
the magnetocardiogram acquisition waistcoat can be made of non-magnetic materials, has certain elasticity, is suitable for target users with different body types, can wrap a body during wearing, is not easy to displace, and is used as a magnetocardiogram acquisition channel for measuring magnetic signals generated by the heart during sleeping; the magnetocardiogram sensor fixing base is arranged on the magnetocardiogram acquisition waistcoat, and a user inserts the probe into the array fixing base 3 at different positions according to different heart positions and can select the number of the inserted sensors. That is to say, the magnetocardiogram sensor of this embodiment can be flexibly arranged based on the sensor array base, and is laid out above the heart of the subject, and a plurality of magnetocardiogram sensors can be installed, and the number of corresponding acquisition channels is selected according to the number of sensors.
The magnetocardiogram monitoring structure further comprises an analog output device and an analog-to-digital conversion acquisition card for acquiring magnetocardiogram signals and background magnetic noise, the magnetocardiogram signals and the background magnetic noise are amplified and filtered by the analog output device and then output to the analog-to-digital conversion acquisition card, and the converted digital signals are recorded in the signal processing device in real time. The analog output device and the analog-to-digital conversion acquisition card are sequentially connected with the signal processing device.
In other embodiments, the magnetocardiogram-based sleep quality monitoring system may further include: the video acquisition assembly is used for acquiring image signals, and the sound acquisition assembly is used for acquiring sound signals; at the moment, the video acquisition assembly and the sound acquisition assembly are both connected with the signal processing device; that is, in practical applications, a user may configure a sound collection component such as a microphone or a video collection component such as a CCD image sensor according to his or her needs. That is, the blood oxygen collecting component, the sound collecting component and the video collecting component are all components which can be selected by the user to add or not.
A user can select required components according to the requirement, and the daily sleep quality monitoring only needs a magnetocardiogram monitoring structure; if the user suspects that the user has the Sleep Disordered Breathing (SDB) or suggests to further carry out the Sleep Disordered Breathing (SDB) examination after daily sleep quality monitoring, the user can add a blood oxygen collecting component to accurately detect whether the user has the Sleep Disordered Breathing (SDB) and the severity of the Sleep Disordered Breathing (SDB); if the user wants to record the detailed information of the sleep period of the user and carry out fine sleep monitoring, the image and sound acquisition assembly for acquiring the image signal and the sound signal can be continuously added to obtain a detailed sleep monitoring report.
If only daily sleep quality monitoring is needed and only a magnetocardiogram sensor is needed, if a user suspects that the user has Sleep Disordered Breathing (SDB) or suggests to further perform Sleep Disordered Breathing (SDB) examination after the daily sleep quality monitoring, the user can add an oximeter 4 to accurately detect whether the user has the Sleep Disordered Breathing (SDB) and the severity of the Sleep Disordered Breathing (SDB); if the user wishes to record the detailed information of the sleep period of the user and perform the fine sleep monitoring, the image acquisition component 5 comprising a camera and a microphone can be continuously added to obtain a detailed sleep monitoring report.
Example two
As shown in fig. 2 and fig. 3, an embodiment of the present invention provides a sleep quality monitoring method based on magnetocardiogram, the execution subject of the method may be a signal processing device in a monitoring system, and specifically, the monitoring may include the following steps:
s10, acquiring magnetocardiogram signals and blood oxygen signals of the sleep time of the target user by means of a monitoring system;
s20, preprocessing and denoising the magnetocardiogram signals, screening available magnetocardiogram signal time periods and obtaining intermediate magnetocardiogram signals.
For example, the magnetocardiogram signal quality factor can be adopted to optimize the magnetocardiogram signal, the available magnetocardiogram signal time is screened, and the optimized magnetocardiogram signal is subjected to noise reduction processing based on the soft thresholding processing method of empirical mode decomposition to obtain the intermediate magnetocardiogram signal.
And S30, extracting the middle magnetocardiogram signal by adopting a QRS wave detection method and a cardiopulmonary coupling algorithm, and acquiring a characteristic signal for reflecting the sleep quality of the target user.
And S40, preprocessing the blood oxygen signal, extracting characteristic values, and screening by adopting a pre-trained machine learning model based on the extracted characteristic values to screen out a time period capable of reflecting the breathing events of the target user.
For example, the blood oxygen signal of each time segment may be preprocessed and feature value extracted; for example, removing artifacts of the blood oxygen signal, resampling the signal to 25Hz, reserving two decimal places, segmenting the signal by using a 3min sliding window with an overlapping rate of 25%, extracting a characteristic value of the blood oxygen signal, extracting ODI3 of the blood oxygen signal and a frequency spectrum skewness of an interested frequency band as characteristic values capable of reflecting respiratory events, inputting the extracted characteristic values into a pre-trained machine learning model, and screening time periods capable of reflecting respiratory events.
And S50, obtaining the sleep quality assessment result of the target user based on the characteristic value of the blood oxygen signal in the screened time period and the characteristic signal in the screened time period.
In the specific implementation process, the quality label of the magnetocardiogram signal can be obtained according to the following steps
Judging whether the characteristic signal in the time period is in accordance with (namely, in accordance with the available standard, namely, the preset available standard), if so, acquiring the sleep quality evaluation result of the target user according to the characteristic value of the blood oxygen signal in the screened time period and the characteristic signal; otherwise, acquiring a sleep quality evaluation result of the target user according to the characteristic value of the blood oxygen signal in the screened time period.
In an alternative implementation, when the user selects the sound collection assembly and the image collection assembly, the method may further include the following step S60 not shown in the figure:
s60, acquiring sound signals and/or image signals of the sleeping time of the target user by means of a monitoring system;
processing the sound signals by adopting another machine learning model trained in advance, screening effective sound signals, and pairing time periods corresponding to the effective sound signals with the time periods of the characteristic signals/the screened time periods;
if the time is overlapped, analyzing the characteristic signal, the blood oxygen signal and the sound signal of the overlapped time period to obtain a sleep quality evaluation result of the target user;
or analyzing the characteristic signals, the blood oxygen signals, the sound signals and the image signals of the overlapped time periods to obtain the sleep quality evaluation result of the target user.
In this embodiment, the butterworth filters may be used to denoise the image signals separately and segment the signals with a 3min sliding window with an overlap rate of 25%; and then, extracting dynamic characteristics or dynamic change information such as turning over, body movement and the like by using a convolutional neural network for the image signal of each divided window, further effectively screening the image signal, and correspondingly, performing subsequent sleep quality evaluation results of the target user according to the effectively screened image signal.
The method comprises the steps of denoising sound signals by using a Butterworth filter, dividing the sound signals by using a 3min sliding window with the overlapping rate of 25%, judging whether the sound signals exceed a threshold value in the period of time such as snoring, dreaming and teeth grinding by using a fully-connected neural network for the sound signals of each divided window, and taking the sound signals exceeding the threshold value as effective sound signals for screening, thereby carrying out the subsequent sleep quality evaluation result of a target user according to the effectively screened sound signals or carrying out the subsequent sleep quality evaluation result of the target user by combining other blood oxygen signals and magnetocardiogram signals.
The method of the embodiment carries out sleep monitoring in all directions by means of the magnetocardiogram signal, the blood oxygen signal, the sound signal and the image signal, better analyzes the sleep quality of the target user, and obtains a sleep quality evaluation result with higher accuracy.
EXAMPLE III
The embodiment of the invention provides a sleep quality monitoring method based on a magnetocardiogram, which carries out detailed description on the processing of each signal.
The preprocessing and noise reduction processing on the magnetocardiogram signal in the step S20, and the screening of the available magnetocardiogram signal time period and the obtaining of the intermediate magnetocardiogram signal may include the following sub-steps:
the substeps S21,Dividing the magnetocardiogram signal based on 1-minute sliding window, and adopting the method in each windowThe 'harr' wavelet base carries out 7-layer wavelet packet decomposition on the magnetocardiogram signal, and the magnetocardiogram signal with the sampling rate of 256Hz
Decomposing in 6 sub-bands (such as 0-1Hz, 1-5Hz, 5-15Hz, 15-50Hz, 15-100Hz and 100-120 Hz), and calculating the second sub-band
Layer 7 in minutes
The proportion of wavelet in the energy of the layer
Wherein
is a magnetocardiogram signal
Within minute is
The energy of the individual subspaces;
is the largest integer that is less than the number of sample points and that can be divided exactly by 256,
。
the substeps S22,Respectively extracting magnetocardiogram signals at
Wavelet energy ratio of 6 sub-bands in minutes
，
Entropy of energy
And kurtosis of magnetocardiogram signals
A total of 13 eigenvalues are used as quality factors for evaluating the quality of magnetocardiogram signals,
is as follows
The average of the sampled points in minutes,
is as follows
Standard deviation of sampling points in minutes;
the substeps S23,Inputting the extracted 13 quality factors into a trained support vector machine to sequentially judge the quality of magnetocardiogram signals per minute, screening available magnetocardiogram signals, and obtaining a quality label of the magnetocardiogram signals
And quality optimized magnetocardiogram signal
。
In a specific implementation process, the sub-step S23 may include the following processes:
s23-1, initializationi=1First, the quality of the 1 st minute signal is evaluated;
s23-2, the first stepiMinute 13 eigenvalues input already trained supportThe model of the measuring machine outputs the result ofqu i =1Orqu i =0Signal quality label
If outputqu i =1Then represents the firstiThe minute signal is available, execute S23-3; if outputqu i =0Then represents the firstiThe signal of minute is not available, the signal quality of the next minute is continuously judged until the available signal of the first section is found, and the order is giveni=i+1Repeating S23-2;
s23-3, ifiIf the minute signal is available, a first segment of available signal can be determined, the value of the segment of available signal is assigned to each segment of unavailable signal, so that the detection precision of QSR wave is not reduced due to sudden change of the optimized signal caused by baseline drift, the length of the signal is unchanged, the number of respiratory events per minute can be conveniently judged by subsequently combining the blood oxygen signal, and the optimized magnetocardiogram signal
，
Wherein
Continuing to judge the next section of signal, orderi =i+1；
S23-4, the first stepiInputting 13 characteristic values of minutes into the trained support vector machine model to obtain
Value of (2), signal quality label
(ii) a If it is
Then represents the firstiThe minute signal can be used, the optimized magnetocardiogram signal is the same as the original signal,
(ii) a If it is
Then represents the firstiMinute signal is not available, willi-1 minute signal assignment to current signal, optimized magnetocardiogram signal
The substeps S24,Adaptively decomposing the quality-optimized magnetocardiogram signal into a plurality of intrinsic mode functions IMFs (intrinsic mode functions) by using an EMD (empirical mode decomposition) method based on signal local characteristics, soft thresholding the IMFs by using multiples of a Donoho threshold as the threshold to obtain the IMFs, and superposing and reconstructing the soft thresholded IMFs to obtain the denoised magnetocardiogram signal
。
For better understanding, this substep is explained in detail:
inputting the quality-optimized magnetocardiogram signal
Performing EMD decomposition and initializing iteration times
And residual terms
And c represents the number of iterative decompositions,
residual term determination before starting of the c-th decomposition
All poles, all maximum points of
Upper envelope curve of
All minimum points forming the lower envelope
The mean line of the signal is obtained according to the upper and lower envelope lines
(ii) a Subtracting the envelope mean line from the residual term
(ii) a If it is
The zero point and the extreme point have equal number or differ by at most 1, and the average value of the envelope curve formed by the extreme points at any time is zero, then
As an IMF value to
And outputting and updating the number of iterations
Up to
Satisfying the criterion of iteration termination, outputting residual terms
Completing EMD decomposition;
after EMD decomposition
Wherein
The intrinsic mode component IMF is represented,
representing a residual term or a trend term;
using a multiple of the Donoho threshold as the threshold:
where C is a constant and N is the data length; different IMFs have different energies, different scale factors C are selected for the different IMFs, and 0.6-1.2 are sequentially selected;
each intrinsic mode is subjected to soft thresholding, and the w mode after the soft thresholding
Overlapping and reconstructing the IMF subjected to soft thresholding to obtain a denoised magnetocardiogram signal
。
In another possible implementation manner, the extracting the magnetocardiogram signal by using the QRS wave detection method and the cardiopulmonary coupling algorithm in step S30 to obtain the characteristic signal reflecting the respiratory event of the target user and the corresponding time period of the characteristic signal may include the following sub-steps:
the substeps S31,Performing square operation on each item of the intermediate magnetocardiogram signal (i.e. denoised magnetocardiogram signal) to increase the intensity and increase the high-frequency component to obtain a first signal
，
；
For the first signal
Carrying out three-time moving average filtering to respectively extract QRS characteristics
QRS threshold
And baseline noise level in the signal
(ii) a Calculating a QRS threshold that includes an offset,
；
for example, cubic filtering is illustrated as follows:
for the first signal
Performing moving average filtering, and optimizing the width of the filter
25, filtered signal
。
For QRS wavesDetecting variations in heart rate, locally estimating heart rate by short-time Fourier transform, and estimating filter width by these locally
。
At time t
Average value of (2)
Wherein t represents any integer time,
. Amplitude at frequency b at time t
b is any integer frequency between 1 and 50. Maximum frequency at time t
Wherein
To prevent large frequency jumps between successive time points. Heart rate estimate at time t
And then calculating the heart rate estimated value at each sampling point by a proximity interpolation method
Calculating the width of the filter
。
For the first signal
Performing moving average filtering with a filter width of
Calculating QRS threshold
。
For the first signal
Moving average filtering to locally estimate baseline noise level in a signal
。
Wherein the width of the filter
1281, offset factor
. Calculating a QRS threshold that includes an offset,
。
the substeps S32,Obtaining the positions of all points with QRS characteristics larger than the QRS threshold value according to a threshold value method, marking the positions of the points with 0 and 1,
；
secondly, the calculation is centered on each sample point,
is the sum of the labels of the bandwidths:
then, whether a positive integer exists at each sample point n is judged one by one
So that the following conditions are all true;
if the QRS wave exists, the QRS wave is marked here, and the position function of the QRS wave is obtainedT（n）；
That is, the duration of time that the QRS feature is greater than the threshold at the QRS wave is greater than
Removing T wave, P wave and noise as criterion to obtain position function representing QRS waveT（n）。
The substeps S33,Extraction of RR interval sequence of intermediate signal from extracted QRS wave
And respiratory signals
；
Specifically, the abscissa corresponding to each QRS wave peak is the position of the R wave, and is recorded as
，
Is as follows
Sampling points corresponding to the R waves
Amplitude of a sequence of RR intervals
RR interval sequence
. Calculating respiratory signals by area mapping
The area of the QRS wave is
Respiratory signal
。
The substeps S34,Interpolating the RR interval sequence by cubic spline
And respiratory signals
Resampling to 4Hz to obtain new RR interval sequence
And respiratory signals
(ii) a The resampled respiratory signal was then segmented using a 3min sliding window with an overlap rate of 25%, and the cardiopulmonary coupling strength in each window was calculated
。
The heart-lung coupling strength reflects the synchronous degree of the heart rate and the respiratory cycle, and the respiratory signal is multiplied by the RR interval sequence coherence and the cross power spectrum square to obtain the heart-lung coupling strength of the heart-lung coupling strength
. The method comprises the following specific steps:
first, RR interval sequence is calculated
And respiratory signals
Cross correlation cross power spectral density of
Wherein
，
。
Then calculating RR interval sequence
Sequence of coherent RR intervals with a respiratory signal
And respiratory signals
Cross correlation cross power spectral density of
Wherein
. Cardio-pulmonary coupling strength of electrocardio signal and respiration signal
。
The degree of synchronization between the heart rate and the respiratory cycle varies significantly with the sleep stage, so that the heart-lung coupling strength in each window can be determined according to the heart-lung coupling strength
The current sleep state is determined. Cardiopulmonary coupling strength when sleep is stable
The high frequency power (0.1-0.5 Hz for adults and 0.2-0.45 Hz for children) is large, and the coupling strength of the heart and lung is high when respiratory events occur
The low frequency power (0.04-0.1 Hz for adults and 0.04-0.2 Hz for children) is larger. Extracting each window
The normalized low-frequency power LFC-N (ratio of low-frequency power to total power) and the ratio LVHC of low-frequency power to high-frequency power of the magnetocardiogram signal are used as eigenvalues/signatures of the magnetocardiogram signal.
By processing the magnetocardiogram signals, the blood oxygen signals, the images and the sound signals, sleep stage information, respiratory event information and body movement and snoring information can be obtained and input into the trained fully-connected neural network for sleep quality monitoring and auxiliary diagnosis of Sleep Disordered Breathing (SDB) of children and adults.
It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the terms first, second, third and the like are for convenience only and do not denote any order. These words are to be understood as part of the name of the component.
Furthermore, it should be noted that in the description of the present specification, the description of the term "one embodiment", "some embodiments", "examples", "specific examples" or "some examples", etc., means that a specific feature, structure, material or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, the claims should be construed to include preferred embodiments and all changes and modifications that fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include such modifications and variations.
Claims (10)
1. A sleep quality monitoring method based on magnetocardiogram is characterized by comprising the following steps:
s10, acquiring magnetocardiogram signals and blood oxygen signals of the sleep time of the target user by means of a monitoring system;
s20, preprocessing and denoising the magnetocardiogram signals, screening available magnetocardiogram signal time periods and obtaining intermediate magnetocardiogram signals;
s30, extracting the middle magnetocardiogram signal by adopting a QRS wave detection method and a cardiopulmonary coupling algorithm, and acquiring a characteristic signal for reflecting the sleep quality of a target user;
s40, preprocessing the blood oxygen signal, extracting a characteristic value, and screening a time period capable of reflecting the respiratory event of the target user based on the characteristic value by adopting a pre-trained machine learning model;
and S50, acquiring a sleep quality evaluation result of the target user based on the characteristic value of the blood oxygen signal in the screened time period and the characteristic signal in the screened time period.
2. The monitoring method of claim 1, further comprising:
s60, acquiring sound signals and/or image signals of the sleeping time of the target user by means of a monitoring system;
processing the sound signals by adopting another machine learning model trained in advance, screening effective sound signals, and pairing time periods corresponding to the effective sound signals with the time periods of the characteristic signals/the screened time periods;
if the time is overlapped, analyzing the characteristic signal, the blood oxygen signal and the sound signal of the overlapped time period to obtain a sleep quality evaluation result of the target user;
or analyzing the characteristic signals, the blood oxygen signals, the sound signals and the image signals of the overlapped time periods to obtain the sleep quality evaluation result of the target user.
3. The monitoring method according to claim 1, wherein the S20 includes:
and optimizing the magnetocardiogram signals by adopting the quality factors of the magnetocardiogram signals, and performing noise reduction processing on the optimized magnetocardiogram signals by using a soft thresholding processing method based on empirical mode decomposition to obtain intermediate magnetocardiogram signals.
4. The monitoring method according to claim 1, wherein the S20 includes:
s21, segmenting the magnetocardiogram signals based on the sliding window of 1 minute, adopting 'harr' wavelet basis to carry out 7-layer wavelet packet decomposition on the magnetocardiogram signals in each window, and carrying out 7-layer wavelet packet decomposition on the magnetocardiogram signals with the sampling rate of 256Hz
Decomposing in 6 sub-bands, calculating
Layer 7 in minutes
The proportion of wavelet in the energy of the layer
，
Wherein,
is a magnetocardiogram signal
Within minute is
The energy of the individual subspaces;
s22, extracting magnetocardiogram signals respectively
Wavelet energy ratio of 6 sub-bands in minutes
，
Entropy of energy
And kurtosis of magnetocardiogram signals
A total of 13 eigenvalues are used as quality factors for evaluating the quality of magnetocardiogram signals,
is as follows
The average of the sampled points in minutes,
is as follows
Standard deviation of sampling points in minutes;
s23, inputting the extracted 13 quality factors into a trained support vector machine to sequentially judge the quality of the magnetocardiogram signals per minute, and screening time periods for sleep quality evaluation to obtain quality labels of the magnetocardiogram signals
And qualityOptimized magnetocardiogram signal
；
S24, decomposing the quality-optimized magnetocardiogram signals into a plurality of intrinsic mode functions IMFs in a self-adaptive manner based on signal local characteristics by using an EMD method, using multiples of a Donoho threshold as a threshold, soft-thresholding the obtained IMFs, superposing and reconstructing the soft-thresholded IMFs, removing noise of the magnetocardiogram signals, and obtaining intermediate magnetocardiogram signals
。
5. The monitoring method according to claim 1, wherein the S30 includes:
s31, performing square operation on each item of the middle magnetocardiogram signal, increasing the intensity value and increasing the high-frequency component to obtain a first signal
For the first signal
Carrying out three-time moving average filtering to respectively extract QRS characteristics
QRS threshold
And baseline noise level in the signal
(ii) a Calculating a QRS threshold that includes an offset,
；
s32, obtaining the positions of all points with QRS characteristics larger than the QRS threshold value according to the threshold value method, marking the positions of the points with 0 and 1,
the duration of the QRS characteristic at the QRS wave position being larger than the threshold value is larger than
Removing T wave, P wave and noise as criterion to obtain position function representing QRS waveT（n）；
S33, extracting RR interval sequence of intermediate signal according to extracted QRS wave
And respiratory signals
；
S34, using cubic spline interpolation method to sequence RR intervals
And respiratory signals
Resampling to 4Hz, then segmenting the resampled respiratory signal using a 3min sliding window with an overlap rate of 25%, calculating the cardiopulmonary coupling strength in each window
；
Extracting the cardiorespiratory coupling strength in each window
The normalized low-frequency power LFC-N and the ratio LVHC of the low-frequency power to the high-frequency power of LFC-N are used as the characteristic signal of the magnetocardiogram signal.
6. The monitoring method according to claim 1, wherein the S40 includes:
s41, preprocessing the blood oxygen signal and extracting a characteristic value;
specifically, removing the artifact of the blood oxygen signal, resampling the signal to 25Hz, and keeping two decimal places;
segmenting the signal by using a 3min sliding window with the overlapping rate of 25%, extracting a characteristic value of the blood oxygen signal, and extracting the ODI3 of the blood oxygen signal and the frequency spectrum skewness of the interested frequency band as the characteristic value capable of reflecting the respiratory event;
and S42, inputting the extracted characteristic values into a machine learning model trained in advance, and screening time periods capable of reflecting the respiratory events.
7. The monitoring method according to claim 4, wherein S50 includes:
quality label based on magnetocardiogram signals
Judging whether the characteristic signals in the time period accord with a preset available standard or not, and if so, acquiring a sleep quality evaluation result of the target user according to the characteristic value of the blood oxygen signals in the screened time period and the characteristic signals in the screened time period;
otherwise, acquiring a sleep quality evaluation result of the target user according to the characteristic value of the blood oxygen signal of the screened time period.
8. The monitoring method according to claim 2, wherein the S60 includes:
respectively denoising the sound signal and the image signal by using a Butterworth filter and dividing the signals by using a 3min sliding window with an overlapping rate of 25%;
for the image signal of each segmented window, extracting dynamic features by using a convolutional neural network;
and for the sound signals of each divided window, judging whether the period of time exceeds the sound signals of the threshold value by using a fully-connected neural network.
9. The monitoring method according to claim 4, wherein the S23 includes:
s23-1, initialization
First, the quality of the 1 st minute signal is evaluated;
s23-2, the first step
Inputting 13 feature values of minutes into the trained support vector machine model, and outputting the result as
Signal quality label
If output
Then represents the first
The minute signal is available, execute S23-3; if output
Then represents the first
Minute signal is not availableContinuing to judge the signal quality of the next minute until finding the available signal of the first segment, and enabling
Repeating S23-2;
s23-3, if
If minute signal is available, determining the first available signal, assigning the value of the first available signal to each unavailable signal, and optimizing magnetocardiogram signal
，
Wherein
Continuing to judge the next section of signal, order
；
S23-4, the first step
Inputting 13 characteristic values of minutes into the trained support vector machine model to obtain
Value of (2), signal quality label
(ii) a If it is
Then represents the first
Minute signal available, optimized magnetocardiogram signal andthe original signals are the same as each other,
(ii) a If it is
Then represents the first
Minute signal is not available, will
Assigning the minute signal to the current signal, and optimizing the magnetocardiogram signal
Alternatively, the S24 includes:
inputting the quality-optimized magnetocardiogram signal
Performing EMD decomposition and initializing iteration times
And residual terms
And c represents the number of iterative decompositions,
residual term determination before starting of the c-th decomposition
All poles, all maximum points of
Upper envelope curve of
All minimum points forming the lower envelope
The mean line of the signal is obtained according to the upper and lower envelope lines
(ii) a Subtracting the envelope mean line from the residual term
(ii) a If it is
The zero point and the extreme point have equal number or differ by at most 1, and the average value of the envelope curve formed by the extreme points at any time is zero, then
As an IMF value to
And outputting and updating the number of iterations
Up to
Satisfying the criterion of iteration termination, outputting residual terms
Completing EMD decomposition;
after EMD decomposition
Wherein
The intrinsic mode component IMF is represented,
representing a residual term or a trend term;
using a multiple of the Donoho threshold as the threshold:
where C is a constant and N is the data length; different IMFs have different energies, different scale factors C are selected for the different IMFs, and 0.6-1.2 are sequentially selected;
each intrinsic mode is subjected to soft thresholding, and the w mode after the soft thresholding
Overlapping and reconstructing the IMF subjected to soft thresholding to obtain a denoised magnetocardiogram signal
。
10. A magnetocardiogram-based sleep quality monitoring system, comprising:
a magnetocardiogram monitoring structure for acquiring magnetocardiogram signals;
a blood oxygen collecting component for collecting blood oxygen signals;
a video acquisition component for acquiring image signals;
a sound collection assembly for collecting sound signals;
the signal processing device is connected with the magnetocardiogram monitoring structure and the blood oxygen acquisition assembly; the video acquisition assembly and the sound acquisition assembly are both connected with the signal processing device;
the magnetocardiogram monitoring structure comprises: the magnetocardiogram acquisition waistcoat is internally provided with a magnetocardiogram sensor and is worn on the upper body of a target user; the signal processing device executes the magnetocardiogram-based sleep quality monitoring method as set forth in any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111589528.3A CN113974576B (en) | 2021-12-23 | 2021-12-23 | Sleep quality monitoring system and monitoring method based on magnetocardiogram |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111589528.3A CN113974576B (en) | 2021-12-23 | 2021-12-23 | Sleep quality monitoring system and monitoring method based on magnetocardiogram |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113974576A true CN113974576A (en) | 2022-01-28 |
| CN113974576B CN113974576B (en) | 2022-04-22 |
Family
ID=79734159
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111589528.3A Active CN113974576B (en) | 2021-12-23 | 2021-12-23 | Sleep quality monitoring system and monitoring method based on magnetocardiogram |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113974576B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115251939A (en) * | 2022-06-21 | 2022-11-01 | 西南交通大学 | Method, device and equipment for detecting arousal and readable storage medium |
| CN115392327A (en) * | 2022-10-27 | 2022-11-25 | 博睿康科技（常州）股份有限公司 | Signal baseline updating method, signal detection method and measuring device |
| CN116808391A (en) * | 2023-06-29 | 2023-09-29 | 北京理工大学 | Sleep awakening method and system based on physiological signal decoding |
Citations (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1552285A (en) * | 2003-05-26 | 2004-12-08 | 中国科学院物理研究所 | Magnetocardiogram and its analyzing method |
| US20050256418A1 (en) * | 2004-05-17 | 2005-11-17 | Beth Israel Deaconess Medical Center | Assessment of sleep quality and sleep disordered breathing based on cardiopulmonary coupling |
| US20050267362A1 (en) * | 2004-05-17 | 2005-12-01 | Beth Israel Deaconess Medical Center | Assessment of sleep quality and sleep disordered breathing based on cardiopulmonary coupling |
| CN104523262A (en) * | 2014-11-18 | 2015-04-22 | 南京丰生永康软件科技有限责任公司 | Sleep quality detection method based on electrocardiosignals |
| CN107361760A (en) * | 2017-07-12 | 2017-11-21 | 中国科学院上海微系统与信息技术研究所 | The magnetocardiograph of magnetocardiograph diagnostic system and the application system |
| CN107590508A (en) * | 2017-08-25 | 2018-01-16 | 漫迪医疗仪器（上海）有限公司 | A kind of mcg-signalses adaptive processing method and system |
| CN107753012A (en) * | 2016-08-19 | 2018-03-06 | 中国科学院上海微系统与信息技术研究所 | A kind of mcg-signalses method for evaluating quality, system and server |
| CN107894965A (en) * | 2017-11-30 | 2018-04-10 | 陕西师范大学 | A kind of coupled processing method for being used for two groups of signal with different type |
| WO2018130036A1 (en) * | 2017-01-16 | 2018-07-19 | 清华大学 | Multi-sensor cardiopulmonary coupling sleep quality detection system and detection method thereof |
| CN109359506A (en) * | 2018-08-24 | 2019-02-19 | 浙江工业大学 | A kind of mcg-signals noise-reduction method based on wavelet transformation |
| CN109431470A (en) * | 2018-12-20 | 2019-03-08 | 西安交通大学医学院第二附属医院 | Sleep breath monitoring method and device |
| CN109770892A (en) * | 2019-02-01 | 2019-05-21 | 中国科学院电子学研究所 | A kind of sleep stage method based on electrocardiosignal |
| CN110325108A (en) * | 2017-08-18 | 2019-10-11 | 克里沃医疗技术有限公司 | Signal processing in medical magnetic meter |
| CN209966365U (en) * | 2018-11-09 | 2020-01-21 | 中科数字健康科学研究院（南京）有限公司 | Portable sleep monitoring system |
| CN110742585A (en) * | 2019-10-10 | 2020-02-04 | 北京邮电大学 | Sleep staging method based on BCG (BCG-broadcast) signals |
| CN110824944A (en) * | 2019-11-22 | 2020-02-21 | 珠海格力电器股份有限公司 | Sleep behavior information prediction method and system based on intelligent household equipment |
| US20200107775A1 (en) * | 2018-10-02 | 2020-04-09 | Spacelabs Healthcare L.L.C. | Methods and Systems for Monitoring Sleep Apnea |
| CN111685774A (en) * | 2020-05-28 | 2020-09-22 | 西安理工大学 | OSAHS diagnosis method based on probability integration regression model |
| CN111938627A (en) * | 2020-08-10 | 2020-11-17 | 浙江工业大学 | Method for filtering magnetocardiogram signal baseline drift |
| WO2020245595A1 (en) * | 2019-06-07 | 2020-12-10 | Prevayl Limited | Method of controlling access to activity data from a garment |
| CN112294264A (en) * | 2020-10-31 | 2021-02-02 | 无锡中物云信息科技有限公司 | Sleep staging method based on BCG and blood oxygen saturation rate |
| CN112890777A (en) * | 2021-01-22 | 2021-06-04 | 深圳市苏仁智能科技有限公司 | Sleep state staging method and device based on cardiopulmonary coupling and computer readable storage medium |
| CN112930139A (en) * | 2018-09-17 | 2021-06-08 | 脉源有限公司 | System and method for analyzing sleep disordered breathing events |
| US20210369165A1 (en) * | 2020-05-28 | 2021-12-02 | Hi Llc | Systems and methods for recording biomagnetic fields of the human heart |
-
2021
- 2021-12-23 CN CN202111589528.3A patent/CN113974576B/en active Active
Patent Citations (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1552285A (en) * | 2003-05-26 | 2004-12-08 | 中国科学院物理研究所 | Magnetocardiogram and its analyzing method |
| US20050256418A1 (en) * | 2004-05-17 | 2005-11-17 | Beth Israel Deaconess Medical Center | Assessment of sleep quality and sleep disordered breathing based on cardiopulmonary coupling |
| US20050267362A1 (en) * | 2004-05-17 | 2005-12-01 | Beth Israel Deaconess Medical Center | Assessment of sleep quality and sleep disordered breathing based on cardiopulmonary coupling |
| CN104523262A (en) * | 2014-11-18 | 2015-04-22 | 南京丰生永康软件科技有限责任公司 | Sleep quality detection method based on electrocardiosignals |
| CN107753012A (en) * | 2016-08-19 | 2018-03-06 | 中国科学院上海微系统与信息技术研究所 | A kind of mcg-signalses method for evaluating quality, system and server |
| WO2018130036A1 (en) * | 2017-01-16 | 2018-07-19 | 清华大学 | Multi-sensor cardiopulmonary coupling sleep quality detection system and detection method thereof |
| CN107361760A (en) * | 2017-07-12 | 2017-11-21 | 中国科学院上海微系统与信息技术研究所 | The magnetocardiograph of magnetocardiograph diagnostic system and the application system |
| CN110325108A (en) * | 2017-08-18 | 2019-10-11 | 克里沃医疗技术有限公司 | Signal processing in medical magnetic meter |
| CN107590508A (en) * | 2017-08-25 | 2018-01-16 | 漫迪医疗仪器（上海）有限公司 | A kind of mcg-signalses adaptive processing method and system |
| CN107894965A (en) * | 2017-11-30 | 2018-04-10 | 陕西师范大学 | A kind of coupled processing method for being used for two groups of signal with different type |
| CN109359506A (en) * | 2018-08-24 | 2019-02-19 | 浙江工业大学 | A kind of mcg-signals noise-reduction method based on wavelet transformation |
| CN112930139A (en) * | 2018-09-17 | 2021-06-08 | 脉源有限公司 | System and method for analyzing sleep disordered breathing events |
| US20200107775A1 (en) * | 2018-10-02 | 2020-04-09 | Spacelabs Healthcare L.L.C. | Methods and Systems for Monitoring Sleep Apnea |
| CN209966365U (en) * | 2018-11-09 | 2020-01-21 | 中科数字健康科学研究院（南京）有限公司 | Portable sleep monitoring system |
| CN109431470A (en) * | 2018-12-20 | 2019-03-08 | 西安交通大学医学院第二附属医院 | Sleep breath monitoring method and device |
| CN109770892A (en) * | 2019-02-01 | 2019-05-21 | 中国科学院电子学研究所 | A kind of sleep stage method based on electrocardiosignal |
| WO2020245595A1 (en) * | 2019-06-07 | 2020-12-10 | Prevayl Limited | Method of controlling access to activity data from a garment |
| CN110742585A (en) * | 2019-10-10 | 2020-02-04 | 北京邮电大学 | Sleep staging method based on BCG (BCG-broadcast) signals |
| CN110824944A (en) * | 2019-11-22 | 2020-02-21 | 珠海格力电器股份有限公司 | Sleep behavior information prediction method and system based on intelligent household equipment |
| CN111685774A (en) * | 2020-05-28 | 2020-09-22 | 西安理工大学 | OSAHS diagnosis method based on probability integration regression model |
| US20210369165A1 (en) * | 2020-05-28 | 2021-12-02 | Hi Llc | Systems and methods for recording biomagnetic fields of the human heart |
| CN111938627A (en) * | 2020-08-10 | 2020-11-17 | 浙江工业大学 | Method for filtering magnetocardiogram signal baseline drift |
| CN112294264A (en) * | 2020-10-31 | 2021-02-02 | 无锡中物云信息科技有限公司 | Sleep staging method based on BCG and blood oxygen saturation rate |
| CN112890777A (en) * | 2021-01-22 | 2021-06-04 | 深圳市苏仁智能科技有限公司 | Sleep state staging method and device based on cardiopulmonary coupling and computer readable storage medium |
Non-Patent Citations (5)
| Title |
|---|
| 刘冬冬等: "基于经验模式分解的心肺耦合技术在儿童睡眠分析中的应用", 《中国医疗设备》 * |
| 刘冬冬等: "基于经验模式分解的心肺耦合技术在睡眠分析中的应用", 《中国医疗设备》 * |
| 樊蕾: "心肺耦合分析技术的临床应用与研究进展", 《世界睡眠医学杂志》 * |
| 符浩等: "基于心电信号的心肺耦合技术在睡眠相关疾病中的应用", 《上海交通大学学报（医学版）》 * |
| 莫晓云等: "心肺耦合技术在睡眠呼吸障碍疾病诊断中的应用", 《现代医学与健康研究》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115251939A (en) * | 2022-06-21 | 2022-11-01 | 西南交通大学 | Method, device and equipment for detecting arousal and readable storage medium |
| CN115392327A (en) * | 2022-10-27 | 2022-11-25 | 博睿康科技（常州）股份有限公司 | Signal baseline updating method, signal detection method and measuring device |
| CN116808391A (en) * | 2023-06-29 | 2023-09-29 | 北京理工大学 | Sleep awakening method and system based on physiological signal decoding |
| CN116808391B (en) * | 2023-06-29 | 2023-12-26 | 北京理工大学 | Sleep awakening method and system based on physiological signal decoding |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113974576B (en) | 2022-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113974576B (en) | Sleep quality monitoring system and monitoring method based on magnetocardiogram | |
| US7862515B2 (en) | Apparatus for detecting sleep apnea using electrocardiogram signals | |
| US20240130627A1 (en) | Continuous non-invasive monitoring of a pregnant human subject | |
| CN110269600B (en) | Non-contact video heart rate detection method based on multivariate empirical mode decomposition and combined blind source separation | |
| US7672717B1 (en) | Method and system for the denoising of large-amplitude artifacts in electrograms using time-frequency transforms | |
| US7343198B2 (en) | System, software, and method for detection of sleep-disordered breathing using an electrocardiogram | |
| US7559903B2 (en) | Breathing sound analysis for detection of sleep apnea/popnea events | |
| US12102450B2 (en) | System and method for assessing physiological signal quality | |
| CN111387959A (en) | Non-contact physiological parameter detection method based on IPPG | |
| KR20220037669A (en) | Non-invasive type ecg monitoring device and method | |
| WO2019079829A9 (en) | Method of preprocessing and screening auscultatory sound signals | |
| Cho et al. | Detection of arousals in patients with respiratory sleep disorders using a single channel EEG | |
| CN116784860B (en) | Electrocardiosignal characteristic extraction system based on morphological heart beat template clustering | |
| CN117786470A (en) | Electrocardiosignal and respiratory signal separation and reconstruction method based on radar signal | |
| Ari et al. | On a robust algorithm for heart sound segmentation | |
| CN115024716B (en) | Heart attack graph signal reconstruction method based on heart rate label generation | |
| Sbrollini et al. | Evaluation of the low-frequency components in surface electromyography | |
| CN116616709A (en) | Working memory load assessment method based on multichannel video in non-stationary state | |
| Liu et al. | A modified approach of peak extraction from BCG for heart rate estimation | |
| CN113349753A (en) | Arrhythmia detection method based on portable dynamic electrocardiogram monitor | |
| CN112826451A (en) | Anesthesia depth and sleep depth assessment method and device | |
| CN115886742B (en) | Sleep dynamics analysis method based on variable heart-lung coupling | |
| Akhbardeh et al. | Ballistocardiogram classification using a novel transform so-called AliMap and biorthogonal wavelets | |
| US20220361798A1 (en) | Multi sensor and method | |
| Arvinti et al. | Frequency Correlated Filtering for Ambulatory Electrocardiograms |
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 |