CN116808391B - Sleep awakening method and system based on physiological signal decoding - Google Patents
Sleep awakening method and system based on physiological signal decoding Download PDFInfo
- Publication number
- CN116808391B CN116808391B CN202310784039.6A CN202310784039A CN116808391B CN 116808391 B CN116808391 B CN 116808391B CN 202310784039 A CN202310784039 A CN 202310784039A CN 116808391 B CN116808391 B CN 116808391B
- Authority
- CN
- China
- Prior art keywords
- sleep
- user
- layer
- physiological signal
- physiological
- 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.)
- Active
Links
- 230000007958 sleep Effects 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000008667 sleep stage Effects 0.000 claims abstract description 35
- 238000007781 pre-processing Methods 0.000 claims abstract description 24
- XPYQFIISZQCINN-QVXDJYSKSA-N 4-amino-1-[(2r,3e,4s,5r)-3-(fluoromethylidene)-4-hydroxy-5-(hydroxymethyl)oxolan-2-yl]pyrimidin-2-one;hydrate Chemical compound O.O=C1N=C(N)C=CN1[C@H]1C(=C/F)/[C@H](O)[C@@H](CO)O1 XPYQFIISZQCINN-QVXDJYSKSA-N 0.000 claims abstract description 10
- 230000000694 effects Effects 0.000 claims abstract description 9
- 230000004927 fusion Effects 0.000 claims description 37
- 238000003062 neural network model Methods 0.000 claims description 18
- 238000012549 training Methods 0.000 claims description 17
- 238000011176 pooling Methods 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical group [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 5
- 239000008280 blood Substances 0.000 claims description 5
- 210000004369 blood Anatomy 0.000 claims description 5
- 229910052760 oxygen Inorganic materials 0.000 claims description 5
- 239000001301 oxygen Substances 0.000 claims description 5
- 239000013598 vector Substances 0.000 claims description 5
- 210000004556 brain Anatomy 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 230000000638 stimulation Effects 0.000 claims description 4
- 230000002618 waking effect Effects 0.000 claims description 4
- 230000037007 arousal Effects 0.000 claims description 3
- 230000029058 respiratory gaseous exchange Effects 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 description 8
- 206010016256 fatigue Diseases 0.000 description 6
- 238000002474 experimental method Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000007500 overflow downdraw method Methods 0.000 description 3
- 230000004461 rapid eye movement Effects 0.000 description 3
- 230000000241 respiratory effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000035790 physiological processes and functions Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 206010019233 Headaches Diseases 0.000 description 1
- 206010049816 Muscle tightness Diseases 0.000 description 1
- 206010062519 Poor quality sleep Diseases 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007177 brain activity Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 231100000869 headache Toxicity 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 210000000987 immune system Anatomy 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 210000000653 nervous system Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000036387 respiratory rate Effects 0.000 description 1
- 230000003860 sleep quality Effects 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to a sleep awakening method and a sleep awakening system based on physiological signal decoding, comprising the following steps: the user enters sleep activity after setting the awakened period and the awakening mode, and physiological signals in sleep of the user are obtained; preprocessing a physiological signal based on an FMDC algorithm to obtain a residual signal; decoding the residual signal based on an XFD algorithm to obtain a sleep stage of a user; judging the sleep stage of the user, and if the user is in the awakenable sleep stage, awakening the user based on the set awakened period and awakening mode. The wake-up method provided by the invention can enable people to feel more conscious and active, reduce fatigue and uncomfortable feeling, and solve the problem that the getting-up method based on a single physiological signal is easy to misjudge.
Description
Technical Field
The invention relates to the technical field of data processing and artificial intelligence, in particular to a sleep awakening method and system based on physiological signal decoding.
Background
Human sleep is a periodic behavior that generally goes through 4 to 6 cycles from falling asleep to awake, each cycle consisting of five different sleep stages, respectively: the time required to go through these five phases, namely the fall asleep phase, the light sleep phase, the fall asleep phase, the deep sleep phase, and the rapid eye movement phase, is typically 90-120 minutes, with each individual differing slightly. In the deep sleep stage, the deep sleep stage and the rapid eye movement stage, the human body is in a dormant state, the brain activity is slow, the muscle tension is relaxed, at the moment, the physiological metabolism, the nervous system and the immune system functions of the human body are in the lowest state, and the physiological characteristics are stable. The traditional wake-up method is more difficult to wake up the user because of the lower sensitivity to external stimuli when the person is in the deep sleep, deep sleep and rapid eye movement periods. The traditional approach of getting up is usually based on time or noise stimulation to wake up the human body, which often gives uncomfortable feeling. The awakening method is easy to cause the instantaneous change of the physiological state of the human body in the deep sleep period, and causes uncomfortable symptoms such as fatigue, headache and the like. Meanwhile, the awakening method cannot fully utilize the characteristic of the sleep cycle to enable a person to get up in an optimal physiological state.
Therefore, how to develop an effective getting-up method, so that a person is woken up at a specific time of a sleep cycle to obtain better getting-up experience and physical health is an important research direction. In recent years, with the development of sleep monitoring technology, researchers have begun to explore a method of getting up based on physiological signals. However, most of the existing getting-up methods based on physiological signals are based on a single physiological signal to judge the sleep stage of the user, and the scheme is easy to misjudge, so that the user experience is very bad.
Disclosure of Invention
The invention aims to provide a sleep awakening method and system based on physiological signal decoding, which solve the problems that the traditional awakening method is easy to cause discomfort and the getting-up method based on a single physiological signal is easy to generate misjudgment.
In order to achieve the above object, the present invention provides the following solutions:
a sleep wake-up method based on physiological signal decoding, comprising:
the user enters sleep activity after setting a awakened period and a awakened mode, and physiological signals in sleep of the user are obtained;
preprocessing the physiological signal based on an FMDC algorithm to obtain a residual signal;
decoding the residual signal based on an XFD algorithm to obtain a sleep stage of the user;
judging the sleep stage of the user, and if the user is in the awakenable sleep stage, awakening the user based on the set awakened period and awakening mode.
Further, the awakened period includes a time period; the awakening mode comprises voice awakening, lamplight awakening, smell awakening, vibration awakening and electric stimulation awakening.
Further, the physiological signal during sleep of the user comprises: respiration signals, pulse signals, blood oxygen signals, electromyographic signals, and brain electrical signals.
Further, preprocessing the physiological signal based on the FMDC algorithm includes:
constructing a full-connection deep neural network model and a physiological signal data set, training and testing the full-connection deep neural network model based on the physiological signal data set, and taking the full-connection deep neural network model after training and testing as a preprocessing model;
inputting the physiological signal into the preprocessing model, decomposing the physiological signal, and outputting an IMF component of the physiological signal;
fitting an upper envelope and a lower envelope of the IMF component, and calculating an average envelope based on the upper envelope and the lower envelope;
generating an intermediate signal by differencing the physiological signal and the average envelope;
and repeatedly inputting the intermediate signals into the preprocessing model until IMF components cannot be decomposed, obtaining the residual signals, and finishing preprocessing of the physiological signals.
Further, the fully connected deep neural network model includes:
an encoder module for converting an input signal into a low-dimensional feature representation, the encoder module comprising a convolution layer, a pooling layer and a fully connected layer, the convolution layer and the pooling layer being connected by the fully connected layer, the convolution layer and the pooling layer being for extracting local features in the physiological signal, the fully connected layer being for extracting global features in the physiological signal;
a decoder module for converting the low-dimensional feature representation back to an original signal, the decoder module comprising a deconvolution layer, an upsampling layer and a residual layer, the deconvolution layer and the upsampling layer being connected by the residual layer, the residual layer being for preserving a structure of the original signal;
an IMF component and residual output layer for outputting the obtained IMF component and residual based on an output node, and taking the IMF component and residual as a supervisory signal;
and a reconstruction error calculation layer for calculating an error between the decoder output signal and the original signal.
Further, decoding the residual signal based on an XFD algorithm, and acquiring the sleep stage of the user includes:
constructing a sleep decoding model and a physiological signal data set, and training the sleep decoding model based on the physiological signal data set;
and inputting the residual signals into a trained sleep decoding model, and outputting a sleep stage classification result of the user.
Further, the sleep decoding model includes:
the data level fusion layer is used for carrying out convolution processing on the physiological signals and outputting convolution results of the physiological signals;
the feature level fusion layer is used for extracting features from the convolution result of the physiological signals, carrying out feature level fusion on the extracted features and outputting feature vectors;
and the decision-stage fusion layer is used for carrying out decision-stage fusion on the feature vectors and outputting sleep stage classification results.
The invention also provides a sleep awakening system based on the physiological signal decoding, which comprises the following steps:
the input module is used for setting a awakened period and a awakening mode before the user falls asleep;
the monitoring module is used for acquiring and storing physiological signals of the user during sleep;
the preprocessing module is used for preprocessing the physiological signal based on an FMDC algorithm to generate a residual signal;
the decoding module is used for decoding the residual signal based on an XFD algorithm to acquire the sleep stage of the user;
the wake-up module is used for generating a control instruction according to the sleep stage of the user, the wake-up period and the wake-up mode set by the user, sending the control instruction to the controlled equipment, and waking up the user by the controlled equipment by using a personalized wake-up technology;
and the display module is used for displaying prompt information in the using process.
Further, the preprocessing module is specifically configured to:
constructing a full-connection deep neural network model and a physiological signal data set, training and testing the full-connection deep neural network model based on the physiological signal data set, and taking the full-connection deep neural network model after training and testing as a preprocessing model;
inputting the physiological signal into the preprocessing model, decomposing the physiological signal, and outputting an IMF component of the physiological signal;
fitting an upper envelope and a lower envelope of the IMF component, and calculating an average envelope based on the upper envelope and the lower envelope;
generating an intermediate signal by differencing the physiological signal and the average envelope;
and repeatedly inputting the intermediate signals into the preprocessing model until IMF components cannot be decomposed, obtaining the residual signals, and finishing preprocessing of the physiological signals.
Further, the decoding module is specifically configured to:
constructing a sleep decoding model and a physiological signal data set, and training the sleep decoding model based on the physiological signal data set;
and inputting the residual signals into a trained sleep decoding model, and outputting a sleep stage classification result of the user.
The beneficial effects of the invention are as follows:
according to the invention, the sleep stage of the user is acquired by acquiring the multi-mode physiological signals of the user, and the user is awakened in the awakenable sleep stage, so that the problem that the traditional awakening method is easy to cause discomfort is effectively solved, the subjective feeling of the user is more awake and active, the fatigue feeling is greatly reduced, and the quality and efficiency of daily life are improved; the user can set the wake-up mode independently, and the wake-up time can be adjusted according to individual differences so as to achieve personalized wake-up effect, and the user interactivity and the wake-up experience are improved; comprehensively considering various physiological signals, including respiratory signals, pulse signals, blood oxygen signals, electromyographic signals, electroencephalogram signals and the like, a model for decoding various physiological signals is constructed, the classification accuracy is improved, the misjudgment problem of the traditional awakening mode is effectively avoided, and more pleasant and comfortable getting-up experience is achieved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a sleep wake-up method based on physiological signal decoding according to an embodiment of the present invention;
FIG. 2 is a workflow diagram of a FMDC algorithm-based preprocessing model in accordance with an embodiment of the present invention;
FIG. 3 is a workflow diagram of a sleep decoding model based on an XFD algorithm according to an embodiment of the invention;
fig. 4 is a schematic structural diagram of a sleep decoding model based on an XFD algorithm according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a sleep wake-up system based on decoding of physiological signals according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
The embodiment provides a sleep awakening method based on physiological signal decoding, as shown in fig. 1, which comprises the following steps:
s1, a user sets a awakened period and a awakened mode before falling asleep, wherein the awakened period is a period of time, the user can set the awakened period of not less than 1 hour, the awakened modes comprise voice awakening, lamplight awakening, smell awakening, vibration awakening and electric stimulation awakening, and the user can select one or more awakening modes.
S2, after the user wears the physiological signal detection device, the user starts to perform sleep activities.
S3, acquiring physiological signals of a user during sleep, wherein the physiological signals comprise respiratory signals, pulse signals, blood oxygen signals, electromyographic signals and brain electrical signals.
S4, in the pretreatment stage, a pretreatment model based on an FMDC algorithm is used for carrying out pretreatment on the physiological signals, the pretreated physiological signals are obtained, and individual differences of users are eliminated.
As shown in fig. 2, the specific process of preprocessing the physiological signal based on the preprocessing model of the FMDC algorithm is as follows:
s4.1, designing a full-connection depth neural network model, inputting physiological signals, and outputting IMF components and residual errors obtained through decomposition;
s4.2, preparing physiological signal data sets of more than 100 persons, and dividing the data sets into training data sets and test data sets;
s4.3, training the full-connection depth neural network model by using a training data set, wherein the aim is to enable the full-connection depth neural network model to accurately decompose the IMF component of the physiological signal;
s4.4, decomposing the test data set based on the trained full-connection depth neural network model to obtain an IMF component and a residual error;
s4.5, performing peak detection on each obtained IMF component to obtain all maximum values and minimum values;
s4.6, fitting two smooth wave crest/wave trough fitting curves, namely an upper envelope line and a lower envelope line, by using a cubic spline interpolation method;
s4.7, averaging the upper envelope curve and the lower envelope curve to obtain an average envelope curve;
s4.8, subtracting the average envelope curve from the physiological signal to obtain an intermediate signal;
s4.9, repeating the steps S4.4-S4.7 on the intermediate signal until the IMF component cannot be decomposed, wherein the last residual signal is the residual, and the obtained residual signal is the preprocessed physiological signal.
The designed full-connection depth neural network model is a self-coding neural network model and has the following structure:
an encoder module consisting of a convolution layer and a pooling layer for converting an input signal into a set of low-dimensional feature representations, extracting local patterns in the physiological signal using 5-10 layers of convolution layer and pooling layer, and extracting global features using 2-5 layers of full connection layer;
the decoder module consists of 5-10 deconvolution layers and an up-sampling layer and is used for converting the low-dimensional characteristic representation back to the original signal, meanwhile, a residual error layer is added between the deconvolution layers and the up-sampling layer, and the residual error connection is used for helping the network to better keep the structure of the original signal;
an IMF component and residual output layer for outputting the obtained IMF component and residual using a plurality of output nodes after the last layer of the decoder, using the decomposed IMF component and residual as a supervisory signal;
a reconstruction error calculation layer added after the last layer of the decoder for calculating the error between the decoder output signal and the original signal, the error being trained as a loss function and used for adjusting the parameters of the network;
by using a self-encoding neural network model, features of physiological signals can be learned without supervision, and these features can be used to decompose the signals and output IMF components and residuals; the network structure also has good expandability and generalization capability, and can adapt to various types of physiological signals.
S5, in the awakened period set by the user, decoding the preprocessed physiological signals by using a sleep decoding model based on an XFD algorithm to obtain the sleep stage of the user.
As shown in fig. 3, the specific process of decoding the preprocessed physiological signal by the sleep decoding model based on the XFD algorithm is as follows:
s5.1, acquiring a physiological signal training set of more than 100 persons;
s5.2, training a sleep decoding model based on a physiological signal training set;
s5.3, collecting physiological signals of a user for more than 10 minutes;
s5.4, inputting physiological signals of the user into a trained sleep decoding model based on the XFD algorithm to obtain a sleep stage classification result of the user.
The sleep decoding model based on the XFD algorithm includes a data level fusion layer, a feature level fusion layer and a decision level fusion layer, as shown in fig. 4, and the specific structure is as follows:
(1) Data level fusion layer: the method consists of two construction modes, namely parallel fusion and serial fusion. The parallel fusion method is to construct five independent convolutional neural networks respectively corresponding to respiratory signals, pulse signals, blood oxygen signals, electromyographic signals and brain signals, splice convolution results of the signals, and perform weighting treatment through a multi-layer perceptron to obtain a final output result. The serial fusion method is to connect different physiological signals in series and input the physiological signals as a whole into a multi-input convolutional neural network to obtain a final output result. And the output results of the parallel fusion and the serial fusion are used as the final output result of the data level fusion layer.
(2) Feature level fusion layer: and extracting features from the output of the data level fusion layer, and carrying out feature level fusion. And respectively constructing six independent feature extraction neural networks for the output results of parallel fusion and series fusion to extract different features. The feature extraction neural network is designed by using multi-layer convolution and pooling operation, the importance of each feature is weighted by adopting an additive attention model, and finally, the output result of the model is obtained by adopting multi-layer full-connection layers and an activation function. A back propagation algorithm is used in the training process to optimize to minimize classification errors. Feature selection and feature dimension reduction technology are adopted for six independent feature extraction neural networks to reduce feature dimension and redundancy, improve classification effect and reduce calculation complexity. And finally, merging the outputs of the six independent feature extraction neural networks into a comprehensive feature vector, and adding the outputs of the six independent feature extraction neural networks, wherein the total of seven outputs are taken as the final output of the feature level fusion layer.
(3) Decision-level fusion layer: and carrying out final decision level fusion according to the final output of the feature level fusion layer so as to obtain a final sleep stage classification result. Seven independent fully connected neural networks are adopted as a model for decision-level fusion. The input of the fully-connected neural network is the output of the feature level fusion layer, and the output is the corresponding sleep stage. In the training process, a cross entropy loss function is used to minimize the prediction error of the model. The optimization algorithm adopts a random gradient descent algorithm. And finally, adopting a voting or weighted average decision level fusion method to fuse the classification results of the seven independent fully-connected neural networks and obtain a final sleep stage classification result, namely the final output of the decision level fusion layer.
The design structure of the data level fusion layer, the feature level fusion layer and the decision level fusion layer of the sleep decoding model based on the XFD algorithm has the following advantages:
the accuracy is improved, and the accuracy and the precision of an algorithm can be effectively improved by fusing different levels, especially for complex sleep signal decoding tasks;
the robustness is improved, and the sensitivity of the algorithm to noise and interference can be reduced through fusion of different levels, so that the robustness of the algorithm is improved;
the interpretability is improved, and the fusion of different levels can enable the algorithm to better understand the input data and extract more meaningful features from the input data, so that the interpretability of the algorithm is improved;
by accelerating calculation and optimizing fusion of different levels, the calculated amount and memory occupation can be effectively reduced, so that the calculation efficiency and speed of the algorithm are improved.
S6, when the user is in a awakenable sleep stage, the user is awakened by using a personalized awakening technology, and the awakenable sleep stage comprises a sleep stage and a shallow sleep stage.
The personalized wake-up technology optimizes wake-up services according to feedback information of a user. The wake-up time and sleep quality information of the user are collected, and if the wake-up time is fed back by the user too early or too late, the system adjusts the wake-up time stepwise to meet the requirements of the user.
In order to further optimize the technical solution, this embodiment further provides a sleep wake-up system based on decoding of physiological signals, as shown in fig. 5, including:
the input module is used for setting a awakened period and a awakening mode before falling asleep for a user;
the monitoring module is used for acquiring and storing physiological signals of a user in a sleep period;
the pretreatment module is used for pretreating the physiological signal by using a pretreatment model based on an FMDC algorithm to obtain a pretreated physiological signal;
the decoding module is used for decoding the preprocessed physiological signals by using a sleep decoding model based on an XFD algorithm to acquire sleep stages of a user;
the wake-up module is used for sending the control instruction to the controlled equipment, and the controlled equipment wakes up the user by using the personalized wake-up technology;
and the display module is used for displaying prompt information in the using process.
To verify the effectiveness of this embodiment, the following experimental verification was performed on its effect:
experimental facilities: advanced physiological signal detection devices are used, including heart rate monitors, respiratory rate detectors, sleep posture detectors, and the like. These devices are capable of accurately detecting physiological signals of a user during sleep and transmitting these signals to a computer system for processing and analysis. Meanwhile, other auxiliary devices, such as a personalized wake-up technology, an interactive application program and the like, are also used so as to record and analyze the experimental results.
30 healthy young adults were enrolled in the experiment and randomized into two groups: experimental and control groups.
Pairing was performed to ensure that differences in age, sex, sleep habits, etc. of the two groups of subjects did not affect the experimental results.
The experimental group uses the multimode sleep awakening method based on the physiological signal decoding provided by the embodiment, and the comparison group uses the traditional alarm clock awakening mode.
Prior to the experiment, all participants were presented and trained in detail to ensure that they had knowledge of the experimental procedure and notice and signed informed consent.
Five days of the experiment were run and the subjective arousal experience and feeling of fatigue of the participants were recorded every morning.
The experimental results show that: the experimental group was superior to the control group in subjective wake experience and feeling of fatigue. After the awakening method provided by the embodiment is used by the participants of the experimental group, compared with the traditional alarm clock awakening mode, subjective feeling is more awake and active, and fatigue feeling is reduced by more than 50%. Meanwhile, the wake-up method provided by the embodiment can effectively avoid the problem of misjudgment of the traditional wake-up mode, so that the getting-up experience of the participant is more pleasant and comfortable.
In addition, the experiment also compares the wake-up effect of different physiological signals. Experimental results show that the best effect can be obtained by waking up by using various physiological signals, and the waking effect is better than that of a single physiological signal. Further illustrates the superiority of the multi-mode sleep awakening method based on the physiological signal decoding.
In general, the multimode sleep awakening method based on physiological signal decoding provided by the embodiment can improve the getting-up experience of people, reduce fatigue and uncomfortable feeling, and effectively avoid the problem of misjudgment of the traditional awakening mode. The method has practicability and universality, and has positive effects on improving the life quality of people.
The above embodiments are merely illustrative of the preferred embodiments of the present invention, and the scope of the present invention is not limited thereto, but various modifications and improvements made by those skilled in the art to which the present invention pertains are made without departing from the spirit of the present invention, and all modifications and improvements fall within the scope of the present invention as defined in the appended claims.
Claims (5)
1. A sleep wake-up method based on physiological signal decoding, comprising:
the user enters sleep activity after setting a awakened period and a awakened mode, and physiological signals in sleep of the user are obtained;
preprocessing the physiological signal based on an FMDC algorithm to obtain a residual signal, including:
constructing a full-connection deep neural network model and a physiological signal data set, training and testing the full-connection deep neural network model based on the physiological signal data set, and taking the full-connection deep neural network model after training and testing as a preprocessing model;
inputting the physiological signal into the preprocessing model, decomposing the physiological signal, and outputting an IMF component of the physiological signal;
fitting an upper envelope and a lower envelope of the IMF component, and calculating an average envelope based on the upper envelope and the lower envelope;
generating an intermediate signal by differencing the physiological signal and the average envelope;
repeatedly inputting the intermediate signals into the preprocessing model until IMF components cannot be decomposed, obtaining the residual signals, and finishing preprocessing the physiological signals;
decoding the residual signal based on an XFD algorithm to obtain a sleep stage of the user, including:
constructing a sleep decoding model and a physiological signal data set, and training the sleep decoding model based on the physiological signal data set;
inputting the residual signals into a trained sleep decoding model, and outputting a sleep stage classification result of the user;
the sleep decoding model includes:
the data level fusion layer is used for carrying out convolution processing on the physiological signals and outputting convolution results of the physiological signals;
the feature level fusion layer is used for extracting features from the convolution result of the physiological signals, carrying out feature level fusion on the extracted features and outputting feature vectors;
the decision-stage fusion layer is used for carrying out decision-stage fusion on the feature vectors and outputting sleep stage classification results;
judging the sleep stage of the user, and if the user is in the awakenable sleep stage, awakening the user based on the set awakened period and awakening mode.
2. The sleep wake-up method based on physiological signal decoding according to claim 1, wherein the awakened period comprises a period of time; the awakening mode comprises voice awakening, lamplight awakening, smell awakening, vibration awakening and electric stimulation awakening.
3. The sleep arousal method based on physiological signal decoding of claim 1 wherein the physiological signal during sleep of the user comprises: respiration signals, pulse signals, blood oxygen signals, electromyographic signals, and brain electrical signals.
4. The sleep arousal method based on physiological signal decoding of claim 1 wherein the fully connected deep neural network model comprises:
an encoder module for converting an input signal into a low-dimensional feature representation, the encoder module comprising a convolution layer, a pooling layer and a fully connected layer, the convolution layer and the pooling layer being connected by the fully connected layer, the convolution layer and the pooling layer being for extracting local features in the physiological signal, the fully connected layer being for extracting global features in the physiological signal;
a decoder module for converting the low-dimensional feature representation back to an original signal, the decoder module comprising a deconvolution layer, an upsampling layer and a residual layer, the deconvolution layer and the upsampling layer being connected by the residual layer, the residual layer being for preserving a structure of the original signal;
an IMF component and residual output layer for outputting the obtained IMF component and residual based on an output node, and taking the IMF component and residual as a supervisory signal;
and a reconstruction error calculation layer for calculating an error between the decoder output signal and the original signal.
5. A wake-up system for implementing a physiological signal decoding based sleep wake-up method as claimed in any one of claims 1-4, characterized in that it comprises:
the input module is used for setting a awakened period and a awakening mode before the user falls asleep;
the monitoring module is used for acquiring and storing physiological signals of the user during sleep;
the preprocessing module is used for preprocessing the physiological signal based on an FMDC algorithm to generate a residual signal;
the decoding module is used for decoding the residual signal based on an XFD algorithm to acquire the sleep stage of the user;
the wake-up module is used for generating a control instruction according to the sleep stage of the user, the wake-up period and the wake-up mode set by the user, sending the control instruction to the controlled equipment, and waking up the user by the controlled equipment by using a personalized wake-up technology;
and the display module is used for displaying prompt information in the using process.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310784039.6A CN116808391B (en) | 2023-06-29 | 2023-06-29 | Sleep awakening method and system based on physiological signal decoding |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310784039.6A CN116808391B (en) | 2023-06-29 | 2023-06-29 | Sleep awakening method and system based on physiological signal decoding |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116808391A CN116808391A (en) | 2023-09-29 |
CN116808391B true CN116808391B (en) | 2023-12-26 |
Family
ID=88128914
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310784039.6A Active CN116808391B (en) | 2023-06-29 | 2023-06-29 | Sleep awakening method and system based on physiological signal decoding |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116808391B (en) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011115188A (en) * | 2008-06-13 | 2011-06-16 | Heart Metrics Kk | Sleeping condition monitoring apparatus, monitoring system, and computer program |
CN105615834A (en) * | 2015-12-22 | 2016-06-01 | 深圳创达云睿智能科技有限公司 | Sleep stage classification method and device based on sleep EEG (electroencephalogram) signals |
CN107773214A (en) * | 2016-08-26 | 2018-03-09 | Tcl集团股份有限公司 | A kind of method, computer-readable medium and the system of optimal wake-up strategy |
CN108596043A (en) * | 2018-03-29 | 2018-09-28 | 中国药科大学 | The automatic method by stages of sleep of single lead EEG signals based on set empirical mode decomposition |
CN111493828A (en) * | 2020-04-23 | 2020-08-07 | 四川大学 | Sequence-to-sequence sleep disorder detection method based on full convolution network |
CN112641451A (en) * | 2020-12-18 | 2021-04-13 | 北方工业大学 | Multi-scale residual error network sleep staging method and system based on single-channel electroencephalogram signal |
WO2021248882A1 (en) * | 2020-06-08 | 2021-12-16 | 珠海格力电器股份有限公司 | Methods and apparatuses for constructing sleep state detection model and for detecting sleep state |
CN113974576A (en) * | 2021-12-23 | 2022-01-28 | 北京航空航天大学杭州创新研究院 | Sleep quality monitoring system and monitoring method based on magnetocardiogram |
CN115374812A (en) * | 2022-08-02 | 2022-11-22 | 中山大学附属第三医院 | Signal feature extraction method for multi-feature fusion extraction of electroencephalogram signals |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI725254B (en) * | 2017-11-20 | 2021-04-21 | 金寶電子工業股份有限公司 | Wearable device capable of recognizing sleep stage and recognition method thereof |
-
2023
- 2023-06-29 CN CN202310784039.6A patent/CN116808391B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011115188A (en) * | 2008-06-13 | 2011-06-16 | Heart Metrics Kk | Sleeping condition monitoring apparatus, monitoring system, and computer program |
CN105615834A (en) * | 2015-12-22 | 2016-06-01 | 深圳创达云睿智能科技有限公司 | Sleep stage classification method and device based on sleep EEG (electroencephalogram) signals |
CN107773214A (en) * | 2016-08-26 | 2018-03-09 | Tcl集团股份有限公司 | A kind of method, computer-readable medium and the system of optimal wake-up strategy |
CN108596043A (en) * | 2018-03-29 | 2018-09-28 | 中国药科大学 | The automatic method by stages of sleep of single lead EEG signals based on set empirical mode decomposition |
CN111493828A (en) * | 2020-04-23 | 2020-08-07 | 四川大学 | Sequence-to-sequence sleep disorder detection method based on full convolution network |
WO2021248882A1 (en) * | 2020-06-08 | 2021-12-16 | 珠海格力电器股份有限公司 | Methods and apparatuses for constructing sleep state detection model and for detecting sleep state |
CN112641451A (en) * | 2020-12-18 | 2021-04-13 | 北方工业大学 | Multi-scale residual error network sleep staging method and system based on single-channel electroencephalogram signal |
CN113974576A (en) * | 2021-12-23 | 2022-01-28 | 北京航空航天大学杭州创新研究院 | Sleep quality monitoring system and monitoring method based on magnetocardiogram |
CN115374812A (en) * | 2022-08-02 | 2022-11-22 | 中山大学附属第三医院 | Signal feature extraction method for multi-feature fusion extraction of electroencephalogram signals |
Also Published As
Publication number | Publication date |
---|---|
CN116808391A (en) | 2023-09-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108304917B (en) | P300 signal detection method based on LSTM network | |
Bablani et al. | Survey on brain-computer interface: An emerging computational intelligence paradigm | |
CN112656427A (en) | Electroencephalogram emotion recognition method based on dimension model | |
Modi et al. | A socially inspired framework for human state inference using expert opinion integration | |
Di Martino et al. | High-resolution physiological stress prediction models based on ensemble learning and recurrent neural networks | |
CN112509696A (en) | Health data detection method based on convolution autoencoder Gaussian mixture model | |
Banfi et al. | Efficient embedded sleep wake classification for open-source actigraphy | |
CN109166611A (en) | A kind of sleep massage method and system based on data analysis | |
Zhu et al. | ArvaNet: Deep recurrent architecture for PPG-based negative mental-state monitoring | |
Busia et al. | Eegformer: Transformer-based epilepsy detection on raw eeg traces for low-channel-count wearable continuous monitoring devices | |
Shi et al. | A smart detection method of sleep quality using EEG signal and long short-term memory model | |
Yu et al. | Modality fusion network and personalized attention in momentary stress detection in the wild | |
CN116808391B (en) | Sleep awakening method and system based on physiological signal decoding | |
Gan et al. | Human-computer interaction based interface design of intelligent health detection using PCANet and multi-sensor information fusion | |
CN114145717A (en) | Sleep state analysis method based on PPG heart rate characteristic parameters and motion quantity | |
Lian et al. | A novel time-incremental end-to-end shared neural network with attention-based feature fusion for multiclass motor imagery recognition | |
CN107577986B (en) | Respiration and heartbeat component extraction method, electronic equipment and storage medium | |
Van et al. | A Personalized Adaptive Algorithm for Sleep Quality Prediction using Physiological and Environmental Sensing Data | |
CN115770043A (en) | Dream emotion recognition method and device, electronic equipment and storage medium | |
CN113081656A (en) | Intelligent massage chair and control method thereof | |
CN113192598A (en) | Efficacy prediction method and system for taVNS therapy | |
KR102478102B1 (en) | Method of treatment for insomnia | |
Sudhamathy et al. | Hybrid convolutional neural network-long short-term memory model for automated detection of sleep stages | |
Martínez-Cruz et al. | Monwatch: A fuzzy application to monitor the user behavior using wearable trackers | |
Su et al. | Construction of regression equation for maximum oxygen uptake recognition of respiration and heart rate in exercise training |
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 |