A device with sleep assessment and intervention functionality

Abstract

The application discloses a device with sleep evaluation and intervention functions, which is arranged on a pillow or a headrest and comprises an information acquisition module, a sleep evaluation module and a sleep intervention module.The information acquisition module is used for acquiring physiological data, head posture data and sound data of a human body, and sending the acquired data to the sleep evaluation module.The sleep evaluation module is used for extracting multi-modal sleep features according to the acquired data, and performing sleep evaluation according to the multi-modal sleep features.The head posture data is used for analyzing the source of the physiological data and giving dynamic weights to complete an adaptive sleep staging algorithm.The sleep intervention module is used for obtaining a sleep intervention scheme according to a sleep evaluation result, and activating different working modes of a vibration motor and / or a loudspeaker according to the sleep intervention scheme to achieve the effect of intervening in the sleep of a user.The head posture data is used for analyzing the source of the physiological data and giving dynamic weights to complete an adaptive sleep staging algorithm, and the sleep staging evaluation accuracy in a real scene is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer technology, and particularly relates to a device with sleep evaluation and intervention functions. BACKGROUND

[0002] Physiological signals record the micro-currents when the human body is active, and these changes can reflect different sleep stages and states. Taking the electroencephalogram (EEG) as an example, the alpha wave (8-12Hz) is mainly observed in the wake period, and this frequency of brain waves will significantly enhance when the individual relaxes and closes his eyes; the non-rapid eye movement (NREM) sleep can be divided into N1, N2 and N3 periods, the N1 period is characterized by theta waves (4-7Hz), the N2 period will appear K complex and sleep spindle waves (12-16Hz), and the N3 period (deep sleep period or slow wave sleep) is dominated by delta waves (<4Hz); the rapid eye movement (REM) sleep is similar to the low-amplitude fast wave in the wake state, but at this time most of the body muscles are in a relaxed state. The rest of the physiological signals also have similar changes during sleep, and through the characteristics of physiological signals in different sleep periods, the sleep stage of the user can be determined.

[0003] The non-invasive electroencephalogram (EEG) acquisition device does not require any surgical operation, and uses electrodes placed on the scalp surface to record brain electrical activity, but this data acquisition method mostly adopts the form of a cap or a headband, and the user is difficult to use these acquisition devices during sleep. At the same time, the existing acquisition devices are complex to wear and inconvenient to carry, and cannot collect physiological data of the user in multiple scenes. In addition, when collecting brain electrical data in daily life using a portable device, brain region data loss or only specific brain region data collection problems occur with the change of the user's posture, and the existing brain electrical sleep evaluation method is mostly designed based on the complete brain region data, so the recognition accuracy is not high in the actual scene.

[0004] Sleep evaluation, also known as sleep staging, is a process of automatically identifying and classifying different sleep stages by analyzing electroencephalogram (EEG) signals. The existing method usually only extracts manual features in the electroencephalogram (EEG) data, such as power spectral density (PSD), time domain statistics, nonlinear dynamic characteristics, etc., and then uses convolutional neural networks, graph neural networks and Transformer models to analyze the sleep situation.

[0005] Non-pharmacological sleep interventions can be divided into psychological intervention and other auxiliary methods. The psychological intervention makes the user in a relaxed state during sleep through the guidance of a psychological consultant, so as to make it easier to fall asleep. The other auxiliary methods improve the sleep condition through light therapy, sound therapy, temperature regulation, vibration stimulation, electromagnetic field exposure and the like. The existing non-pharmacological sleep intervention method often specifies a scheme when the user is awake, and the same scheme is always executed during sleep. However, the user can be divided into different periods during sleep, and different schemes have different effects in different sleep periods. Therefore, the use of the same scheme all the time will reduce the effect of sleep intervention. SUMMARY

[0006] To at least partially solve one of the technical problems existing in the prior art, the purpose of the present application is to provide a device with sleep assessment and intervention functions.

[0007] The first technical solution adopted by the present application is:

[0008] A device with sleep assessment and intervention functions is arranged on a pillow or a headrest, comprising:

[0009] An information acquisition module is used to acquire physiological data, head posture data and sound data of a human body, and send the acquired data to a sleep assessment module,

[0010] The sleep assessment module is used to extract multi-modal sleep features according to the acquired data, and perform sleep assessment according to the multi-modal sleep features. The head posture data is used to analyze the source of the physiological data and give a dynamic weight, so as to complete an adaptive sleep staging algorithm.

[0011] The sleep intervention module is used to obtain a sleep intervention scheme according to the sleep assessment result, and activate different working modes of the vibration motor and / or the loudspeaker according to the sleep intervention scheme, so as to achieve the effect of intervening in the user's sleep.

[0012] Further, the physiological data includes at least one of electrodermal signals, electromyographic signals, heart rate signals, blood oxygen signals or body temperature signals in addition to electroencephalogram signals.

[0013] The physiological data is acquired by a multi-functional flexible sensor array implanted into a flexible fabric layer. The multi-functional flexible sensor includes an electroencephalogram electrode, and at least one of an electromyographic electrode, an electrodermal electrode, a blood oxygen sensor or a body temperature sensor.

[0014] The electroencephalogram electrode, the electromyographic electrode and the electrodermal electrode are designed as flexible structures with needle-like protrusions, so as to ensure good contact between the flexible electrodes and the scalp and reduce the electrode contact impedance.

[0015] Further, the head posture data is obtained by a flexible sensor array implanted into the flexible fabric layer, the flexible sensor array is a sensor array composed of flexible pressure sensors, and the change of the head posture is detected by measuring the distribution of pressure in the array.

[0016] Further, the working mode of the information acquisition module is:

[0017] For the electroencephalogram signal in the physiological data, the electroencephalogram signal is filtered, and then reconstructed using discrete wavelet transform to remove the noise in the electroencephalogram signal;

[0018] For the head posture data, excitation and amplification are required during acquisition, excitation is used to generate a measurable output for the sensor, and amplification is used to increase the amplitude of the measurable value to improve the signal-to-noise ratio;

[0019] For the speech data, a frame windowing method is used for preprocessing to make the speech data approximate to a stationary signal, which is convenient for subsequent feature extraction of the model;

[0020] The processed physiological data, head posture data and speech data are normalized to map the data to the same range to eliminate the dimensional difference.

[0021] Further, the expression for reconstructing the electroencephalogram signal using discrete wavelet transform is:

[0022] cA j-1 [n]=∑ k h[n-2k]cA j [k]+Σ k g[n-2k]cD j [k]

[0023] In the formula, h[n-2k] and g[n-2k] are reconstruction filters; cA j [k] is the approximation coefficient; cD j [k] is the detail coefficient obtained by layer-by-layer decomposition of the wavelet transform; c is the coefficient of each part; k is the summation index for traversing possible time points.

[0024] Further, the preprocessing by frame windowing includes:

[0025] The continuous speech data is divided into small segments of fixed length, and the formula is:

[0026]

[0027] In the formula, X[n] is speech data represented by a discrete time sequence, N is the total length of the signal, L is the frame length of each frame during frame division, S is the frame shift, i.e. the interval between adjacent two frames, and fi is the i-th frame after framing;

[0028] Each frame is windowed after framing to prevent the mutation at both ends of the frame and the problem of frequency spectrum leakage after Fourier transform.

[0029] Further, the working mode of the sleep evaluation module is:

[0030] For head posture data, the posture of the human body during sleep is calculated using a brain region adaptive analysis algorithm to obtain the brain region source of the current electroencephalogram signal, and a projection layer is used to map the electroencephalogram features and other features into the same feature space;

[0031] For physiological data, a self-attention mechanism is used to extract shallow time sequence features of the physiological data for representing the current sleep state of the human body;

[0032] For sound data, a snoring sound detection algorithm is used to extract snoring sound features to assist the judgment of the sleep evaluation model, and then a projection layer is used to map the sound features and other features into the same feature space;

[0033] Through a cross-attention mechanism, the head posture data is used to adaptively judge the signal collection situation, and dynamic weights are given to physiological data of different modalities to complete multi-modal fusion of various physiological data and sound data, and multi-modal sleep features are extracted;

[0034] According to the multi-modal sleep features, sleep staging labels are obtained, and a sleep evaluation report is generated.

[0035] Further, the working mode of the sleep intervention module is:

[0036] According to the sleep evaluation result, a dynamic matching algorithm is used to obtain a scheme with the highest similarity between the current human sleep state features and the scheme features in the preset scheme library, and the working mode of the vibration motor and / or the loudspeaker is controlled according to the obtained scheme;

[0037] The dynamic matching algorithm is a Wasserstein distance algorithm, a Hellinger distance algorithm, a JS divergence algorithm, a KL divergence algorithm, or a total variation distance algorithm.

[0038] Further, for different schemes, the control mode of the vibration motor is as follows:

[0039] In the light sleep stage (N1), an extremely low frequency vibration below 60Hz is used to provide a comfortable background feeling for the human body and help to enter a deeper sleep state faster;

[0040] In the moderate sleep stage (N2), a vibration of 60-150 Hz is used, which can maintain a certain comfort and not easily break the stable sleep pattern formed;

[0041] In the deep sleep stage (N3), a vibration of 90-300 Hz is used, which is strong enough to provide additional security and support, and does not cause awakening;

[0042] In the rapid eye movement sleep stage (REM), the brain activity increases the muscle tension and is easy to wake up, so the vibration intervention is suspended during this period;

[0043] In addition, if serious sleep apnea is detected, the vibration motor will emit a strong shock to wake up the user and prevent serious accidents.

[0044] Further, for different schemes, the control mode of the speaker is as follows:

[0045] In the light sleep stage (N1), the brain gradually transitions from a wakeful state to a shallow sleep state, and playing white noise or pink noise and other background sounds with uniformly distributed energy spectrum helps to smoothly enter deeper sleep;

[0046] In the moderate sleep stage (N2), the consciousness of the human body is further weakened, and spindle waves and K complex waves appear, and low-frequency vibration or soft natural sound is used in this stage to help stabilize sleep;

[0047] In the deep sleep stage (N3), the brain produces a large number of delta waves, and very low frequency sound is used in this period to provide additional security to the human body;

[0048] In the rapid eye movement sleep stage (REM), the brain activity increases the muscle tension and is easy to wake up, and low-frequency binaural beats (about 2 Hz to 7 Hz) are played in this period to induce the brain to appear theta brain waves, thereby optimizing sleep quality.

[0049] The beneficial effects of the present application are: the present application uses head posture data to analyze the source of physiological data and gives dynamic weight to complete the adaptive sleep staging algorithm, improves the sleep staging evaluation accuracy in real scene, and the present application also uses sound data as the auxiliary mode of the sleep evaluation module, so that the recognition result is more accurate. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following introduces the drawings of the related technical solutions in the embodiments of the present application or the prior art. It should be understood that the drawings in the following introduction are only for facilitating the clear description of part of the embodiments of the technical solutions of the present application, and for the technical personnel in the art to obtain other drawings without paying creative labor.

[0051] Figure 1 is a schematic diagram of the device with sleep evaluation and intervention function in the embodiments of the present application;

[0052] Figure 2 is a whole flow chart of the device with sleep evaluation and intervention function in the embodiments of the present application;

[0053] Figure 3 is a hardware architecture diagram of the device with sleep evaluation and intervention function in the embodiments of the present application;

[0054] Figure 4 is a working schematic diagram of the information collection module in the embodiments of the present application;

[0055] Figure 5 is a working schematic diagram of the sleep evaluation module in the embodiments of the present application;

[0056] Figure 6 is a working schematic diagram of the sleep intervention module in the embodiments of the present application. DETAILED DESCRIPTION

[0057] The embodiments of the present application are described in detail below, and the examples of the embodiments are shown in the drawings, wherein the same or similar notations represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation of the present application. For the step numbers in the following embodiments, they are only set for facilitating the description and explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of the technical personnel in the art.

[0058] In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down, front, back, left, right, etc. is based on the orientation or position relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0059] In the description of the present application, the meaning of one or more is one or more, the meaning of multiple is more than two, greater than, less than, more than, etc. Understand as not including the number, above, below, within, etc. Understand as including the number. If it is described to the first, the second is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or implicitly indicating the sequence of indicated technical features.

[0060] In the description of the present application, unless otherwise explicitly limited, the words such as setting, installing, connecting, etc. should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical scheme.

[0061] Term explanation:

[0062] PCBA: is the abbreviation of English Printed Circuit Board Assembly, printed circuit board.

[0063] As Figure 2 shown, the present embodiment provides a device with sleep evaluation and intervention function, which is arranged on a pillow or a headrest, comprising:

[0064] An information acquisition module is configured to acquire physiological data, head posture data and sound data of a human body, and send the acquired data to a sleep evaluation module,

[0065] The sleep evaluation module is configured to extract multi-modal sleep features according to the acquired data, and perform sleep evaluation according to the multi-modal sleep features; wherein the head posture data is used to analyze the source of the physiological data and give dynamic weight, so as to complete the adaptive sleep staging algorithm.

[0066] The sleep intervention module is configured to obtain a sleep intervention scheme according to the sleep evaluation result, and activate different working modes of the vibration motor and / or the loudspeaker according to the sleep intervention scheme, so as to achieve the effect of intervening the user's sleep.

[0067] The working process of the device provided by the present embodiment is as follows: the user attaches the device with sleep evaluation and intervention function to the pillow or headrest, and rests his head on the device during sleep; the information acquisition module of the device acquires the posture signal, multiple physiological signals and sound signals of the user in real time, and sends these data to the sleep evaluation module after preprocessing; the sleep evaluation module realizes the multi-modal sleep staging algorithm through the posture data, sound data and multiple physiological data, wherein the head posture data is used to analyze the source of the physiological data and give dynamic weight, so as to complete the adaptive sleep staging algorithm and obtain an accurate sleep evaluation report. The sleep intervention module adjusts the stimulation degree of sound wave and vibration in real time according to the sleep evaluation report, so as to improve the sleep quality of the user.

[0068] In an embodiment, referring to Figure 1 The main body of the device with sleep evaluation and intervention function is composed of two layers of fabric layers 3 to ensure the comfort of sleep, and the multifunctional sensor 2, the pressure sensor 4 and the printed circuit board 1 are embedded in the fabric layers 3, wherein the printed circuit board 1 further has a microphone, a speaker and a vibration motor.

[0069] 1) Flexible material

[0070] The main body material can be selected from flexible materials such as fabric and leather, and the material selection ensures comfort and aesthetics, which conforms to the daily use habits of users.

[0071] 2) Head posture detection sensor array

[0072] A flexible sensor capable of detecting changes in head posture is implanted in the flexible fabric layer, and a plurality of sensor units form an array. Specifically, a sensor array composed of flexible pressure sensors can be selected, and the changes in head posture can be detected in real time by measuring the distribution of pressure in the array.

[0073] 3) Multifunctional flexible sensor array

[0074] The multifunctional sensor includes electroencephalogram electrodes, electromyogram electrodes, electrodermal electrodes, blood oxygen sensors, body temperature sensors and other vital sign detection units. The multifunctional sensor is implanted in the flexible fabric layer and forms an array. The electroencephalogram electrodes, electromyogram electrodes and electrodermal electrodes can be designed as flexible structures with needle-like protrusions to ensure good contact between the flexible electrodes and the scalp, reduce the electrode contact impedance, and at the same time ensure comfort and safety. The blood oxygen sensor is used to measure blood oxygen, heart rate, blood pressure and respiratory rate. The temperature sensor is used to detect changes in human body temperature.

[0075] 4) Microphone

[0076] The main function is to collect the snoring sound, breathing sound and environmental noise of the user, and interact with the user's voice. The number of microphones is one or more, which are implanted in the flexible fabric layer.

[0077] 5) Speaker

[0078] During the sleep intervention stage, specific frequency band sounds are played to assist the user to enter deeper sleep and improve the user's sleep quality, or to interact with the user's voice. The number of speakers is one or more, which are implanted in the flexible fabric layer.

[0079] 6) Vibration motor

[0080] The micro-vibration is used to assist the user to enter deeper sleep in the sleep intervention stage, to improve the sleep quality of the user, or to wake up the user by vibration when the user has serious sleep apnea to prevent accidents. One or more are implanted in the flexible fabric layer.

[0081] Further as an optional implementation, referring to Figure 3 The data processing flow of the device with sleep evaluation and intervention function can mainly be divided into an input module, an output module and a communication management module. The input module is composed of an attitude detection sensor array, a multifunctional flexible sensor array and a microphone, and is mainly responsible for collecting various data of the user. The output module is composed of a loudspeaker and a vibration motor, and is an execution component when generating a sleep intervention scheme. The communication management module is composed of power management and wireless communication, and is mainly responsible for the power-on operation of the device and the communication between modules.

[0082] The various modules in the device will be explained in detail below in combination with the drawings and specific embodiments.

[0083] (1) Information collection module

[0084] Referring to Figure 4 , the information collection module uses a multifunctional sensor array, an attitude sensor array and a microphone to receive various physiological data, attitude data and sound data, respectively, to provide core data for subsequent sleep evaluation algorithms.

[0085] 1) For various physiological data, taking electroencephalogram data as an example, first use a Butterworth band-pass filter to limit the frequency range within a certain high and low frequency range to remove noise and unnecessary signals, because the frequency range of electroencephalogram signal is 0.5-70Hz, therefore a 0.5Hz low-pass filter and a 70Hz high-pass filter are adopted. The transfer function formula is as follows:

[0086]

[0087] where H(s) is the transfer function of the system, which defines the response of the system to different frequency inputs in the Laplace transform domain. G0 is the gain, which is usually a constant representing the maximum gain of the filter (usually 1, meaning no amplification or attenuation). ω0 is the center angular frequency. s is a complex frequency variable used for Laplace transform in continuous-time systems. n is the order of the filter, which determines the selectivity and steepness of the filter. Then use discrete wavelet transform to reconstruct the electroencephalogram data to remove noise in the electroencephalogram signal, the formula is as follows:

[0088] cA j-1 [n]=∑ k h[n-2k]cA j [k]+∑ k g[n-2k]cDj [k]

[0089] where h[n-2k] and g[n-2k] are the reconstruction filters, A j [k] are approximation coefficients, D j [k] are detail coefficients are obtained after the wavelet transform decomposes layer by layer.

[0090] For other physiological data, this part of data includes skin electricity, electromyography, heart rate, body temperature and other data. Skin electricity and electromyography data are physiological micro-current data which are easily disturbed by other factors, so similar filtering operation as brain electrical data is used for denoising. Heart rate and body temperature data are single numerical variables, so they are not processed. Subsequent addition of other types of physiological data also follows the preprocessing rules.

[0091] 2) For posture data, excitation and amplification are required during collection. Excitation is used to produce a measurable output for the sensor, and amplification is used to improve the amplitude of the measurable value to improve the signal-to-noise ratio. Specifically, a constant voltage source V EXC As excitation, the output voltage can be represented as:

[0092] V OUT = kV EXC ΔR / R

[0093] Then use the instrument amplifier to amplify the obtained posture signal:

[0094] V OUT = G(V IN+ -V IN- )

[0095] where V IN+ and V IN- correspond to the positive and negative output terminals of the bridge, respectively.

[0096] 3) For speech data, frame and windowing are used for preprocessing to make the speech data approximately stationary signals, which is convenient for subsequent model feature extraction. Frame segmentation means that continuous speech data is divided into small segments of fixed length, which can be expressed as:

[0097]

[0098] where X[n] is the speech data represented by a discrete time sequence, N is the total length of the signal, L is the frame length of each frame when framing, S is the frame shift, i.e. the interval between adjacent two frames, f i is the i-th frame after framing. After framing, each frame is windowed to prevent frame end mutations and Fourier transform frequency spectrum leakage problems. Here, the Hamming window is used as follows:

[0099]

[0100] New frame f after applying window function i The [n] can be expressed as an element-by-element product of the original frame and the window function:

[0101] f i [n] = f i [n] x w[n], 0 < n < L

[0102] After processing, various physiological data, posture data, and voice data are obtained. To ensure that the subsequent algorithms are more stable and efficient, the data is normalized to map the data to the same range and eliminate dimensional differences. The specific formula is as follows:

[0103]

[0104] where X is the original data, X min and X max are the minimum and maximum values of the data, respectively. Through normalization, the data can be mapped to the [0, 1] interval.

[0105] (2) Sleep evaluation module

[0106] Referring to Figure 5 , the input of the sleep evaluation module is the posture data, various physiological data, and sound data obtained in the information collection module. Then, the multi-modal sleep staging model is used to obtain the multi-modal sleep features of the user. Next, the classifier is used for classification, and finally, the visual user sleep evaluation report is output.

[0107] Specifically, in the multi-modal sleep staging model, for the posture data, the brain region adaptive analysis algorithm is used to calculate the posture of the user during sleep, accurately obtaining the brain region source of the current electroencephalogram data, and then using the projection layer to map it and other features to the same feature space. For various physiological data, the self-attention mechanism is used to extract the shallow time sequence features of these physiological data, which are used to represent the current user's sleep state. For sound data, the snoring sound detection algorithm is used to extract snoring sound features to assist the judgment of the sleep evaluation model, and then the projection layer is used to map it and other features to the same feature space. In the cross-attention layer, the posture data is used to adaptively judge the signal collection situation to give different modal physiological data dynamic weights, and the multi-modal fusion of various physiological data and sound data is completed in this layer. The multi-modal sleep features are extracted using a multi-layer Transformer stack as a time sequence feature layer, and then the sleep staging labels are obtained using CNN, LSTM, or SVM as a classifier, and finally the sleep evaluation report is visualized.

[0108] (3) Sleep intervention module

[0109] Referring to Figure 6 The sleep intervention module mainly uses a dynamic matching algorithm to obtain the most suitable sleep intervention scheme according to the sleep evaluation report output in the sleep evaluation module, and then activates different working modes of the vibration motor and the speaker according to the scheme, so as to finally achieve the effect of intervening in the user's sleep.

[0110] The scheme library is composed of preset schemes and personalized schemes. The preset schemes are composed of the frequency of the vibration motor and the audio played by the speaker preset by experts according to the sleep characteristics in different periods. The personalized schemes are composed of the frequency of the vibration motor and the audio played by the speaker customized by the user according to his own needs. The dynamic matching algorithm can use Wasserstein distance, Hellinger distance and JS divergence, etc. to filter the scheme with the highest similarity between the current user's sleep state characteristics and the scheme characteristics in the scheme library, and the algorithm can filter a more suitable scheme in real time according to the change of the sleep state. Finally, different working modes of the vibration motor and the speaker are activated according to different schemes.

[0111] Light and rhythmic vibrations can sometimes induce drowsiness, for example, it is sometimes possible to sleep better on a train or in a car. At the same time, the somatosensory music therapy system points out that vibrations in the range of 60-600Hz can stimulate the response of the human body's Pacinian corpuscles, thereby reducing symptoms such as tension and fatigue, which is conducive to relaxation and falling asleep, so this patent uses vibration to intervene in the user's sleep state. Specifically, in the light sleep period (N1), use very low frequency vibration below 60Hz to provide a comfortable background feeling for the body, help to enter deeper sleep state faster; in the moderate sleep period (N2), introduce vibration with slightly higher frequency and gentle rhythm around 60-150Hz, this degree of vibration can maintain a certain comfort and not easily break the stable sleep pattern that has been formed; in the deep sleep period (N3), use deep and slow vibration between 90-300Hz, such vibration intensity is enough to provide additional security and support, while not easily causing awakening. In the rapid eye movement sleep period (REM), the brain activity increases and the muscle tension decreases, which is easy to wake up, so the vibration intervention is suspended during this period. In addition, if it is detected that the user has a serious sleep apnea, the vibration motor will emit a strong vibration to wake up the user and prevent serious accidents.

[0112] Sound can help regulate the biological rhythm of the human body, and listening to appropriate audio during sleep can cause synchronous resonance of tissue cells, relieve the body's tension and anxiety, and thus achieve the effect of helping sleep. Specifically, in the light sleep period (N1), the brain gradually transitions from a wakeful state to a light sleep state, and playing background sounds with uniformly distributed energy spectrum such as white noise or pink noise at this time helps to smoothly enter deeper sleep; in the moderate sleep period (N2), the person's consciousness is further weakened, and spindle waves and K complex waves appear, and using low-frequency vibration or soft natural sounds such as gentle flowing water and forest bird songs in this stage helps to stabilize sleep; in the deep sleep period (N3), the brain produces a large number of delta waves, and this period uses very low frequency sound to provide additional security to the human body. In the rapid eye movement sleep period (REM), brain activity increases muscle tension and is easily awakened, and in this period, low-frequency binaural beats (about 2Hz to 7Hz) are played to induce the brain to appear theta brain waves, thereby optimizing sleep quality.

[0113] In summary, compared with the prior art, the device of the present application has the following advantages and beneficial effects:

[0114] 1) The existing non-invasive brain electrical data acquisition method mostly adopts the form of a hat or a headband, and the user is difficult to use these acquisition devices when sleeping. At the same time, the existing acquisition devices are complex to wear and inconvenient to carry, and cannot collect physiological data of users in various scenes. The present application proposes a device with brain electrical data and other physiological signal acquisition functions, which uses a specific distribution of flexible electrode arrays, and users can place it on a pillow or on a headrest when traveling, so that the acquisition of brain electrical data and other physiological signals can be completed without feeling.

[0115] 2) When the existing technical solution uses a portable device to collect brain electrical data in daily life, brain region data loss or only specific brain region data collection problems may occur with changes in user posture, and the existing brain electrical sleep evaluation method is mostly designed based on the complete brain region data, so the recognition accuracy is not high in actual scenes. The present application proposes a sleep staging algorithm based on adaptive estimation of brain regions, which can determine the source of the collected brain region data through the user's posture, thereby dynamically activating different parts of the model to complete the adaptive sleep staging algorithm of the brain region, and improve the sleep staging evaluation accuracy in real scenes.

[0116] 3) Existing sleep staging algorithms often only use electroencephalogram data, but other physiological data can also provide auxiliary information for sleep staging to improve the accuracy of identification. The present application proposes a multi-modal sleep staging model that uses electroencephalogram data as the main data supplemented by physiological data such as voice, skin electricity, electromyography, heart rate, and body temperature, etc. to jointly analyze the current sleep stage of the user, thereby improving the accuracy of the sleep staging algorithm. The sleep evaluation module of the present application can use multi-modal physiological signals based on electroencephalogram to improve the reliability of the sleep staging algorithm. At the same time, posture data is used to analyze the source of physiological data and assign dynamic weights to complete an adaptive sleep staging algorithm, thereby improving the accuracy of sleep staging evaluation in real scenarios. In addition, the present application also uses snoring sound data as an auxiliary modality for the sleep evaluation module to make the results more accurate.

[0117] 4) Existing non-pharmacological sleep intervention methods often specify a scheme when the user is awake, and the same scheme is always used during sleep. However, the user's sleep can be divided into different periods, and different schemes have different effects in different sleep periods. Therefore, using the same scheme all the time can reduce the effectiveness of sleep intervention. The present application proposes a sleep dynamic intervention scheme based on sleep staging, which dynamically selects different intervention schemes according to the sleep evaluation results to improve the sleep quality of the user. The present application can adjust the sleep intervention scheme in real time according to the sleep evaluation report through a dynamic matching algorithm, thereby achieving more effective sleep intervention effect. The user's sleep is intervened by using a vibration and audio cooperative working mode, so that the intervention effect is more effective.

[0118] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, different embodiments or examples described in the present application and the features of different embodiments or examples can be combined and combined by those skilled in the art without contradiction.

[0119] The above embodiments are only for the purpose of illustrating the technical concept and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the content of the present application and to implement it, and cannot limit the protection scope of the present application. Any equivalent changes or modifications made in accordance with the essence of the present application should be covered within the protection scope of the present application.

Claims

1. A device with sleep assessment and intervention functions, characterized in that, Placed on pillows or cushions, including: The information acquisition module is used to collect physiological data, head posture data, and sound data of the human body, and then send the collected data to the sleep assessment module. The sleep assessment module is used to extract multimodal sleep features from the collected data and to assess sleep based on these features. Specifically, head posture data is used to analyze the source of physiological data and assign dynamic weights to complete an adaptive sleep staging algorithm. The sleep intervention module is used to obtain a sleep intervention plan based on the sleep assessment results, and activate different working modes of the vibration motor and / or speaker according to the sleep intervention plan to achieve the effect of intervening in the user's sleep. The head posture data is acquired by a flexible sensor array implanted in a flexible fabric layer. The flexible sensor array is a sensor array composed of flexible pressure sensors. By measuring the pressure distribution in the array, changes in head posture can be detected. The information collection module operates as follows: For EEG signals in physiological data, the EEG signals are filtered and then reconstructed using discrete wavelet transform to remove noise from the EEG signals. For head posture data, it is necessary to excite and amplify it during acquisition. Excitation is used to generate a measurable output for the sensor, and amplification is used to increase the amplitude of the measurable value to improve the signal-to-noise ratio. For speech data, a frame-by-frame windowing method is used for preprocessing to make the speech data approximately a stationary signal, which facilitates feature extraction in subsequent models. The processed physiological data, head posture data, and voice data are normalized to map the data into the same range in order to eliminate differences in dimensions. The sleep assessment module works as follows: For head posture data, a brain region adaptive analysis algorithm is used to calculate the posture of the human body during sleep in order to obtain the brain region source of the current EEG signal. A projection layer is used to map the EEG features and other features to the same feature space. For physiological data, a self-attention mechanism is used to extract shallow temporal features of the physiological data to characterize the current sleep state of the human body. For sound data, a snoring detection algorithm is used to extract snoring features to assist the sleep assessment model in making judgments. Then, a projection layer is used to map the sound features and other features into the same feature space. By using a cross-attention mechanism, head posture data is used to adaptively determine the signal acquisition status, assign dynamic weights to physiological data of different modalities, and complete the multimodal fusion of various physiological data and sound data to extract multimodal sleep features. Sleep stage labels are obtained based on multimodal sleep characteristics, and a sleep assessment report is generated.

2. The device with sleep assessment and intervention function according to claim 1, characterized in that, The physiological data includes, in addition to, electroencephalogram (EEG) signals, at least one of the following: electrodermal (ED) signals, electromyography (EMG) signals, heart rate signals, blood oxygenation signals, or body temperature signals. The physiological data is acquired by a multifunctional flexible sensor array implanted in a flexible fabric layer. The multifunctional flexible sensor includes an EEG electrode, and at least one of an EMG electrode, an EKE electrode, a blood oxygen sensor, or a body temperature sensor. The electroencephalogram (EEG), electromyogram (EMG), and electrodermal (ED) electrodes are designed with flexible structures featuring needle-like protrusions to ensure good contact between the flexible electrodes and the scalp and reduce electrode contact resistance.

3. The device with sleep assessment and intervention function according to claim 1, characterized in that, The expression for reconstructing EEG signals using discrete wavelet transform is as follows: In the formula, and It is a reconstruction filter; It is the approximation coefficient; These are detail coefficients, all obtained by decomposing the wavelet transform layer by layer. These are the coefficients of each part; It is the summation index.

4. The device with sleep assessment and intervention function according to claim 1, characterized in that, The preprocessing method using frame-by-frame windowing includes: Continuous speech data is divided into fixed-length segments, as expressed by the formula: In the formula, It is speech data represented as a discrete time series. The total length of the signal. This refers to the frame length of each frame during framing. Frame shift is the interval between two adjacent frames. For the first frame after segmentation frame; After framing, each frame is windowed to prevent abrupt changes at both ends of the frame and spectral leakage after Fourier transform.

5. The device with sleep assessment and intervention function according to claim 1, characterized in that, The sleep intervention module works as follows: Based on the sleep assessment results, a dynamic matching algorithm is used to obtain the scheme with the highest similarity between the current human sleep state characteristics and the scheme characteristics in the preset scheme library, and the working mode of the vibration motor and / or speaker is controlled according to the obtained scheme. The dynamic matching algorithm is either the Wasserstein distance algorithm, the Hellinger distance algorithm, the JS divergence algorithm, the KL divergence algorithm, or the total variation distance algorithm.

6. The device with sleep assessment and intervention function according to claim 5, characterized in that, The control methods for the vibration motor are as follows, depending on the specific design: During light sleep, extremely low frequency vibrations below 60Hz are used to provide a comfortable background sensation for the body, helping to enter a deeper sleep state more quickly. During moderate sleep, use vibrations of 60-150Hz. This level of vibration can maintain a certain level of comfort without easily disrupting the established stable sleep pattern. During deep sleep, vibrations of 90-300Hz are used. This vibration intensity is sufficient to provide additional security and support without causing awakening. During REM sleep, increased brain activity and decreased muscle tone make it easier to wake up, so vibration intervention should be suspended during this period. In addition, if severe sleep apnea is detected, the vibration motor will emit a strong vibration to wake the user and prevent serious accidents.

7. The device with sleep assessment and intervention function according to claim 5, characterized in that, The speaker control methods are as follows, depending on the specific solution: During light sleep, the brain gradually transitions from a waking state to a light sleep state. Playing background sounds with a uniformly distributed energy spectrum during this stage helps to smoothly enter deeper sleep. During the intermediate sleep stage, a person's consciousness further weakens, and spindle waves and K complex waves begin to appear. Using low-frequency vibrations or soft, natural sounds at this stage can help stabilize sleep. During deep sleep, the brain produces a large number of delta waves. During this period, extremely low-frequency sounds are used to provide the body with an extra sense of security. During REM sleep, increased brain activity and decreased muscle tone make it easier to wake up. Playing low-frequency binaural beats during this period can induce theta brain waves, thereby optimizing sleep quality.

Images