WO2021248882A1 - 睡眠状态检测模型构建、睡眠状态检测方法及装置 - Google Patents

睡眠状态检测模型构建、睡眠状态检测方法及装置 Download PDF

Info

Publication number
WO2021248882A1
WO2021248882A1 PCT/CN2020/140494 CN2020140494W WO2021248882A1 WO 2021248882 A1 WO2021248882 A1 WO 2021248882A1 CN 2020140494 W CN2020140494 W CN 2020140494W WO 2021248882 A1 WO2021248882 A1 WO 2021248882A1
Authority
WO
WIPO (PCT)
Prior art keywords
sleep state
signals
breathing
user
sleep
Prior art date
Application number
PCT/CN2020/140494
Other languages
English (en)
French (fr)
Inventor
董明珠
赵杰磊
唐杰
徐洪伟
李昱
Original Assignee
珠海格力电器股份有限公司
珠海联云科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 珠海格力电器股份有限公司, 珠海联云科技有限公司 filed Critical 珠海格力电器股份有限公司
Publication of WO2021248882A1 publication Critical patent/WO2021248882A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/08—Detecting, measuring or recording devices for evaluating the respiratory organs
    • A61B5/0826—Detecting or evaluating apnoea events
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/48—Other medical applications
    • A61B5/4806—Sleep evaluation
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/48—Other medical applications
    • A61B5/4806—Sleep evaluation
    • A61B5/4809—Sleep detection, i.e. determining whether a subject is asleep or not
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/48—Other medical applications
    • A61B5/4806—Sleep evaluation
    • A61B5/4818—Sleep apnoea
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction

Definitions

  • the embodiments of the present disclosure relate to the field of smart bedrooms, and in particular to a method and device for constructing a sleep state detection model and sleep state detection.
  • the method for judging human snoring behavior is to determine the snoring behavior based on the characteristics of the signal extracted by the acoustic sensor.
  • the frequency of the snoring signal overlaps the frequency of the respiratory signal and is similar to the amplitude of the human body motion signal, which makes it difficult for related technologies to distinguish between each other. This kind of signal will cause inaccurate judgment of human sleep state.
  • embodiments of the present disclosure provide a sleep state detection model construction, sleep state detection method and device.
  • embodiments of the present disclosure provide a method for constructing a sleep state detection model, including:
  • Input a plurality of the multi-dimensional feature vectors into the initial model, and perform deep learning training until the similarity between the output result of the initial model and the label is greater than or equal to a set threshold, then it is determined that the initial model training is completed, Use the trained initial model as the sleep state detection model.
  • the method further includes:
  • a corresponding first respiration signal is obtained from each of the first electrical signals, and multiple first respiration signals corresponding to multiple first electrical signals are obtained.
  • the method further includes:
  • the method further includes:
  • Filtering is performed on each of the first electrical signals according to the frequency range of the first electrical signals to obtain multiple first breathing signals corresponding to when the user is breathing.
  • the method further includes:
  • Multidimensional feature vector Perform vectorization processing on the multiple parameter information of each of the first respiratory signals to obtain the multi-dimensional feature vector corresponding to each of the first respiratory signals, and then obtain the multiple corresponding to the multiple first respiratory signals. Multidimensional feature vector.
  • embodiments of the present disclosure provide a method for detecting a user's sleep state, including:
  • a plurality of the multi-dimensional feature vectors are input into the sleep state detection model constructed in any one of the first aspect, so that the sleep state detection model outputs the sleep state of the user.
  • the method further includes:
  • the sleep state is sent to the terminal device, so that the terminal device can display the sleep state.
  • embodiments of the present disclosure provide a device for constructing a sleep state detection model, including:
  • An acquiring module configured to acquire multiple first breathing signals when the user is in a sleep state, and tags corresponding to the sleep state
  • the data processing module is configured to determine the multi-dimensional feature vector corresponding to each of the first respiratory signals according to the multiple parameter information in each of the first respiratory signals, and obtain the multiple corresponding to the multiple first respiratory signals. Said multi-dimensional feature vectors;
  • the training module is configured to input a plurality of the multi-dimensional feature vectors into the initial model and perform deep learning training until the similarity between the output result of the initial model and the label is greater than or equal to a set threshold, then it is determined The initial model training is completed, and the trained initial model is used as the sleep state detection model.
  • embodiments of the present disclosure provide a device for detecting a user's sleep state, including:
  • the acquisition module is set to acquire sleep electrical signals when the user is in a sleep state
  • a data processing module configured to extract a plurality of first breathing signals from the sleep electrical signal
  • the data processing module is further configured to determine a plurality of parameter information from each of the first respiratory signals, and generate a multi-dimensional feature vector corresponding to each of the first respiratory signals based on the plurality of parameter information, And then obtain a plurality of said multi-dimensional feature vectors corresponding to a plurality of said first breathing signals;
  • the determining module is configured to input a plurality of the multi-dimensional feature vectors into the sleep state detection model, so that the sleep state detection model outputs the sleep state of the user.
  • an embodiment of the present disclosure provides a sleep detection device, including: a processor and a memory, the processor is configured to execute sleep state detection model construction and user sleep state detection programs stored in the memory to achieve The method for constructing a sleep state detection model according to any one of the above-mentioned first aspect and the method for detecting a user's sleep state according to any one of the above-mentioned second aspect.
  • embodiments of the present disclosure provide a storage medium, including: the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the first The sleep state detection model construction method according to any one of the aspects and the user sleep state detection method according to any one of the second aspects.
  • the solution for constructing a sleep state detection model is to obtain multiple first breathing signals of a user in a sleep state and tags corresponding to the sleep state; Pieces of parameter information, determine the multi-dimensional feature vector corresponding to each of the first breathing signals, and obtain multiple of the multi-dimensional feature vectors corresponding to the multiple of the first breathing signals; input the multiple of the multi-dimensional feature vectors to the initial model
  • deep learning training is performed until the similarity between the output result of the initial model and the label is greater than or equal to the set threshold, then it is determined that the initial model training is completed, and the trained initial model is used as the sleep state detection model, With this method, it is possible to accurately determine the sleep state of the human body by detecting the physiological signal of the human body, which provides convenience for medical auxiliary diagnosis.
  • FIG. 1 is a schematic flowchart of a method for constructing a sleep state detection model provided by an embodiment of the disclosure
  • FIG. 2 is a schematic flowchart of another method for constructing a sleep state detection model provided by an embodiment of the disclosure
  • FIG. 3 is a schematic flowchart of a method for detecting a user's sleep state according to an embodiment of the disclosure
  • FIG. 4 is a schematic structural diagram of a device for constructing a sleep state detection model provided by an embodiment of the disclosure
  • FIG. 5 is a schematic structural diagram of a device for detecting a user's sleep state provided by an embodiment of the disclosure
  • FIG. 6 is a schematic structural diagram of a sleep detection device provided by an embodiment of the disclosure.
  • FIG. 1 is a schematic flowchart of a method for constructing a sleep state detection model provided by an embodiment of the disclosure. As shown in FIG. 1, the method specifically includes:
  • the model will calculate the output user sleep state results based on the input data.
  • the results may be inconsistent with the pre-recorded state labels.
  • Perform deep learning training until the similarity between the user's sleep state result output by the model and the state label is greater than or equal to the set threshold (for example, 90%), then the model training is considered complete, and the trained model at this time is used as sleep state detection Model.
  • This solution provides a method for constructing a sleep state detection model by acquiring multiple first breathing signals when the user is in a sleep state and tags corresponding to the sleep state; according to multiple parameters in each of the first breathing signals Information, determine the multi-dimensional feature vector corresponding to each of the first breathing signals, and obtain multiple of the multi-dimensional feature vectors corresponding to the multiple of the first breathing signals; input the multiple of the multi-dimensional feature vectors into the initial model, Perform deep learning training until the similarity between the output result of the initial model and the label is greater than or equal to the set threshold, then it is determined that the initial model training is completed, and the trained initial model is used as the sleep state detection model. , Can realize and predict the sleep state of the human body by collecting the physiological signal of the human body.
  • FIG. 2 is a schematic flowchart of another method for constructing a sleep state detection model provided by an embodiment of the present disclosure. As shown in FIG. 2, the method specifically includes:
  • S21 Acquire sleep electrical signals when the user is in a sleep state through the piezoelectric sensor.
  • the sleep detection device is placed under the mattress.
  • the sleep detection device is in standby mode during the daily non-detection stage.
  • the piezoelectric sensor installed on the sleep detection device senses a sudden change in the signal, and the sleep detection device turns on the normal working mode, and sleep detection
  • the piezoelectric sensor installed on the device starts to collect piezoelectric signals.
  • the piezoelectric signals collected at this time include signals for all periods of time when the user is awake and sleeping.
  • S22 Intercept a plurality of first electrical signals with a preset duration from the sleep signal.
  • the electrical sleep signals are intercepted multiple times to obtain multiple first electrical signals.
  • the intermediate interval time of the first electrical signal is intercepted twice for the previous time.
  • each measurement value needs to be weighted.
  • x is the signal measurement value
  • n is the width of the sliding weighting process
  • ⁇ is the corresponding weight
  • the sum is 1.
  • the preset duration can be 30s, 40 or 45s. This solution is adjusted according to the actual use process, and there is no specific limitation here.
  • an IIR multi-stage filter preset an IIR multi-stage filter, and filter each first electrical signal of the multiple intercepted first electrical signals, and separate the corresponding first electrical signal from each first electrical signal.
  • the first breath signal preset an IIR multi-stage filter, and filter each first electrical signal of the multiple intercepted first electrical signals, and separate the corresponding first electrical signal from each first electrical signal.
  • S24 Process a plurality of the first breathing signals according to the time-domain characteristics of the signals generated when the user breathes, to remove the Gaussian error of each of the first breathing signals, to obtain a plurality of the first breaths after processing Signal.
  • Short-time Fourier transform is performed on the processed multiple first breathing signals, and each first breathing signal is decomposed into a frequency spectrum, and the frequency domain characteristics corresponding to each first breathing signal are obtained.
  • the corresponding multiple respiratory frequencies are determined according to the frequency domain characteristics corresponding to each first respiratory signal obtained above.
  • the multiple parameter information of the first respiratory signal is obtained according to the respiratory frequency, and the parameter information includes at least parameters such as high frequency, low frequency, amplitude, energy, mean value, variance, root mean square error or median value of the respiratory frequency.
  • the multiple parameter information of each first breathing signal is formed into a multi-dimensional feature vector, and then multiple multi-dimensional feature vectors corresponding to the multiple first breathing signals are obtained.
  • the accuracy of the output result of the sleep state detection model is used as the fitness, and the parameter information is optimized to remove one or more unnecessary parameters to speed up the sleep detection model.
  • the sleep state detection model construction solution obtains the sleep electrical signal of the user in sleep state through a piezoelectric sensor, processes the sleep electrical signal, separates the respiration signal, and then performs data processing on the respiration signal to obtain Multiple signal parameter information forms multiple multi-dimensional feature vectors, and input multiple multi-dimensional feature vectors into the sleep state detection model to predict the user’s sleep state.
  • This method can be achieved by converting human physiological signals into electrical signals.
  • the processing of electrical signals can effectively distinguish various physiological signals, achieve the result of accurately judging the human sleep state, and provide convenience for medical auxiliary diagnosis and improve user experience.
  • FIG. 3 is a schematic flowchart of a method for detecting a user's sleep state provided by an embodiment of the present disclosure. As shown in FIG. 3, the method specifically includes:
  • the sleep detection device is placed under the mattress.
  • the sleep detection device is in standby mode during the daily non-detection stage.
  • the piezoelectric sensor installed on the sleep detection device senses a sudden change in the signal, and the sleep detection device turns on the normal working mode, and sleep detection
  • the piezoelectric sensor installed on the device starts to collect piezoelectric signals.
  • the collected piezoelectric signals include signals during all periods of the user's awake state and sleep state.
  • the collected electrical signals are based on the waking breathing frequency and heart rate. With other parameters, the electrical signal in the user's waking state is removed, and the sleep electrical signal in the user's sleep state is retained.
  • the electrical sleep signals are intercepted multiple times to obtain multiple first electrical signals, and an IIR multi-order filter is preset according to the known breathing frequency of ordinary people, and the intercepted multiple first electrical signals are preset Each first electrical signal of the electrical signal is filtered, and the corresponding first respiration signal is separated from each first electrical signal.
  • S33 Determine a plurality of parameter information from each of the first respiratory signals, and generate a multi-dimensional feature vector corresponding to each of the first respiratory signals based on the plurality of the parameter information, so as to obtain a plurality of the first respiratory signals.
  • the piezoelectric sensor since the piezoelectric sensor has high sensitivity, a slight touch will generate a large jump signal. Therefore, it is necessary to compare the collected first breathing signal according to the time-domain characteristics of the signal generated when the user breathes.
  • the data is processed to remove the Gaussian error of each of the first breathing signals, and the missing values in the signal processing process are filled by polynomial fitting and interpolation operations to ensure signal continuity, and then multiple processed first breathing signals are obtained , Performing short-time Fourier transform on the processed multiple first breathing signals, decomposing each first breathing signal into a frequency spectrum, and then obtaining the frequency domain characteristics corresponding to each first breathing signal, in some embodiments According to the frequency domain characteristics corresponding to each first breathing signal obtained above, corresponding multiple breathing frequencies are determined, and multiple parameter information of the first breathing signal is obtained according to the breathing frequencies.
  • multiple parameter information of each first breathing signal is formed into a multi-dimensional feature vector, and then multiple multi-dimensional feature vectors corresponding to the multiple first breathing signals are obtained.
  • the multiple multi-dimensional feature vectors obtained above are input into the trained sleep state detection model, and the detected sleep state of the user is output.
  • the sleep state includes breathing rate (times/min), snoring period and apnea period.
  • the output result is fed back to the app of the mobile terminal held by the user to facilitate the user to check the sleep state.
  • the method for detecting the sleep state of a user obtains the sleep electrical signal when the user is in the sleep state; extracts a plurality of first breathing signals from the sleep electrical signal; and determines from each of the first breathing signals A plurality of parameter information, and based on a plurality of the parameter information, generate a multi-dimensional feature vector corresponding to each of the first breathing signals, and then obtain a plurality of the multi-dimensional feature vectors corresponding to a plurality of the first breathing signals; A plurality of the multi-dimensional feature vectors are input into the sleep state detection model, so that the sleep state detection model outputs the sleep state of the user.
  • This method can detect human physiological signals to accurately determine the human sleep state and is useful for doctors. Provides convenience for diagnosing sleep-related diseases.
  • FIG. 4 is a schematic structural diagram of a sleep state detection model construction device provided by an embodiment of the disclosure, which specifically includes:
  • the acquiring module 401 is configured to acquire multiple first breathing signals when the user is in a sleep state, and tags corresponding to the sleep state;
  • the data processing module 402 is configured to determine the multi-dimensional feature vector corresponding to each of the first respiratory signals according to the multiple parameter information in each of the first respiratory signals, and obtain multiple corresponding to the first respiratory signals. A plurality of said multi-dimensional feature vectors;
  • the training module 403 is configured to input a plurality of the multi-dimensional feature vectors into the initial model and perform deep learning training until the similarity between the output result of the initial model and the label is greater than or equal to a set threshold, then it is determined The initial model training is completed, and the trained initial model is used as the sleep state detection model.
  • the acquiring module is specifically configured to acquire the sleep electrical signal of the user in the sleep state; intercept a plurality of first electrical signals with a preset duration from the sleep signal; acquire the corresponding first electrical signal from each of the first electrical signals A respiration signal is used to obtain multiple first respiration signals corresponding to multiple first electrical signals.
  • the acquisition module is further configured to acquire sleep electrical signals of the user in a sleep state through the piezoelectric sensor.
  • the acquisition module is further configured to perform filtering processing on each of the first electrical signals according to the frequency range of the first electrical signals, to obtain multiple corresponding first electrical signals when the user is breathing A breathing signal.
  • the data processing module is specifically configured to process a plurality of the first breathing signals according to the time domain characteristics of the signals generated when the user breathes, remove the Gaussian error of each first breathing signal, and obtain the processed multiple Each of the first breathing signals; perform short-time Fourier transform on each of the processed first breathing signals to determine the corresponding frequency domain characteristics; determine the corresponding user’s characteristics based on each of the frequency domain characteristics Respiration frequency; determine the corresponding multiple parameter information of the first respiration signal from each of the respiration frequencies; perform vectorization processing on the multiple parameter information of each of the first respiration signals to obtain each of the The multi-dimensional feature vectors corresponding to the first breathing signal are further obtained to obtain the multiple multi-dimensional feature vectors corresponding to the multiple first breathing signals.
  • the device for constructing a sleep state detection model may be the device for constructing a sleep state detection model as shown in FIG. 4, which can perform all the steps of the method for constructing a sleep state detection model in FIG. 1-2, and then realize
  • FIG. 4 The device for constructing a sleep state detection model as shown in FIG. 4, which can perform all the steps of the method for constructing a sleep state detection model in FIG. 1-2, and then realize
  • FIG. 4 For the technical effects of the method of constructing the sleep state detection model shown in Figure 1-2, please refer to the related description in Figure 1-2 for details. It is a concise description and will not be repeated here.
  • FIG. 5 is a schematic structural diagram of a device for detecting a user's sleep state provided by an embodiment of the disclosure, which specifically includes:
  • the obtaining module 501 is configured to obtain sleep electrical signals when the user is in a sleep state
  • the data processing module 502 is configured to extract multiple first breathing signals from the sleep electrical signals
  • the data processing module 502 is further configured to determine a plurality of parameter information from each of the first respiratory signals, and generate a multi-dimensional feature vector corresponding to each of the first respiratory signals based on the plurality of the parameter information , And then obtain a plurality of said multi-dimensional feature vectors corresponding to a plurality of said first breathing signals;
  • the determining module 503 is configured to input a plurality of the multi-dimensional feature vectors into the sleep state detection model, so that the sleep state detection model outputs the sleep state of the user.
  • the user sleep state detection device provided in this embodiment may be the user sleep state detection device shown in FIG. 5, which can perform all the steps of the method for constructing a sleep state detection model in FIG. 3, thereby realizing the user sleep shown in FIG. 3
  • the state detection method please refer to the related description in FIG. 3 for details, which is a concise description and will not be repeated here.
  • FIG. 6 is a schematic structural diagram of a sleep detection device provided by an embodiment of the disclosure.
  • the electronic device 600 shown in FIG. 6 includes: at least one processor 601, a memory 602, at least one network interface 604, and other user interfaces 603.
  • the various components in the sleep detection device 600 are coupled together through the bus system 605.
  • the bus system 605 is configured to implement connection and communication between these components.
  • the bus system 605 also includes a power bus, a control bus, and a status signal bus.
  • various buses are marked as the bus system 605 in FIG. 6.
  • the user interface 603 may include a display, a keyboard, or a pointing device (for example, a mouse, a trackball (trackball), a touch panel, or a touch screen, etc.).
  • a pointing device for example, a mouse, a trackball (trackball), a touch panel, or a touch screen, etc.
  • the memory 602 in the embodiment of the present disclosure may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory.
  • the non-volatile memory can be read-only memory (Read-Only Memory, ROM), programmable read-only memory (Programmable ROM, PROM), erasable programmable read-only memory (Erasable PROM, EPROM), and electrically available Erase programmable read-only memory (Electrically EPROM, EEPROM) or flash memory.
  • the volatile memory may be a random access memory (Random Access Memory, RAM), which is used as an external cache.
  • RAM static random access memory
  • DRAM dynamic random access memory
  • DRAM synchronous dynamic random access memory
  • DDRSDRAM Double Data Rate Synchronous Dynamic Random Access Memory
  • Enhanced SDRAM, ESDRAM Synchronous Link Dynamic Random Access Memory
  • Synch link DRAM SLDRAM
  • DRRAM Direct Rambus RAM
  • the memory 602 described herein is intended to include, but is not limited to, these and any other suitable types of memory.
  • the memory 602 stores the following elements, executable units or data structures, or their subsets, or their extended sets: operating system 6021 and application programs 6022.
  • the operating system 6021 includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are configured to implement various basic services and process hardware-based tasks.
  • the application program 6022 includes various application programs, such as a media player (Media Player), a browser (Browser), etc., and is set to implement various application services.
  • the program for implementing the method of the embodiments of the present disclosure may be included in the application program 6022.
  • the processor 601 by calling a program or instruction stored in the memory 602, specifically, a program or instruction stored in the application program 6022, the processor 601 is set to execute the method steps provided in each method embodiment, for example, include:
  • the sleep electrical signal of the user in a sleep state is obtained; a plurality of first electrical signals with a preset duration are intercepted from the sleep signal; and the corresponding first electrical signal is obtained from each of the first electrical signals.
  • a first breathing signal to obtain a plurality of the first breathing signals corresponding to a plurality of the first electrical signals.
  • a piezoelectric sensor is used to obtain sleep electrical signals when the user is in a sleep state.
  • filtering processing is performed on each of the first electrical signals according to the frequency range of the first electrical signals to obtain multiple first breathing signals corresponding to the user's breathing.
  • a plurality of the first breathing signals are processed according to the time-domain characteristics of the signals generated when the user is breathing, and the Gaussian error of each first breathing signal is removed to obtain the processed multiple Each of the first breathing signals; perform short-time Fourier transform on each of the processed first breathing signals to determine the corresponding frequency domain characteristics; determine the corresponding user’s characteristics based on each of the frequency domain characteristics Respiration frequency; determine the corresponding multiple parameter information of the first respiration signal from each of the respiration frequencies; perform vectorization processing on the multiple parameter information of each of the first respiration signals to obtain each of the The multi-dimensional feature vectors corresponding to the first breathing signal are further obtained to obtain the multiple multi-dimensional feature vectors corresponding to the multiple first breathing signals.
  • a sleep electrical signal when the user is in a sleep state Acquire a sleep electrical signal when the user is in a sleep state; extract a plurality of first breathing signals from the sleep electrical signal; determine a plurality of parameter information from each of the first breathing signals, and based on a plurality of the parameter information , Generating a multi-dimensional feature vector corresponding to each of the first breathing signals, and then obtaining a plurality of the multi-dimensional feature vectors corresponding to a plurality of the first breathing signals; inputting the plurality of the multi-dimensional feature vectors into the sleep state detection model , So that the sleep state detection model outputs the sleep state of the user.
  • the sleep state is sent to the terminal device, so that the terminal device displays the sleep state.
  • the methods disclosed in the foregoing embodiments of the present disclosure may be applied to the processor 601 or implemented by the processor 601.
  • the processor 601 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 601 or instructions in the form of software.
  • the aforementioned processor 601 may be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (ASIC), a ready-made programmable gate array (Field Programmable Gate Array, FPGA) or other Programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
  • DSP Digital Signal Processor
  • ASIC application specific integrated circuit
  • FPGA Field Programmable Gate Array
  • the methods, steps, and logical block diagrams disclosed in the embodiments of the present disclosure can be implemented or executed.
  • the general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like.
  • the steps of the method disclosed in the embodiments of the present disclosure may be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software units in the decoding processor.
  • the software unit may be located in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, registers.
  • the storage medium is located in the memory 602, and the processor 601 reads the information in the memory 602, and completes the steps of the foregoing method in combination with its hardware.
  • the embodiments described herein can be implemented by hardware, software, firmware, middleware, microcode, or a combination thereof.
  • the processing unit can be implemented in one or more application specific integrated circuits (ASICs), digital signal processors (Digital Signal Processing, DSP), digital signal processing devices (DSPDevice, DSPD), programmable logic Devices (Programmable Logic Device, PLD), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA), general-purpose processors, controllers, microcontrollers, microprocessors, and others configured to perform the functions described in this disclosure Electronic unit or its combination.
  • ASICs application specific integrated circuits
  • DSP digital signal processors
  • DSPDevice digital signal processing devices
  • DSPD programmable logic Devices
  • PLD Field-Programmable Gate Array
  • FPGA Field-Programmable Gate Array
  • the technology described herein can be implemented by a unit that performs the functions described herein.
  • the software codes can be stored in the memory and executed by the processor.
  • the memory can be implemented in the processor or external to the processor.
  • the sleep detection device provided by this embodiment may be the sleep detection device as shown in FIG. 6, which can perform all the steps of the method of constructing the sleep state detection model in FIG. 1-2 and the user's sleep state detection method in FIG. 3, and then To achieve the technical effects of the sleep state detection model construction method shown in Figure 1-2 and the user sleep state detection method in Figure 3, please refer to the related descriptions in Figure 1-2 and Figure 3 for details. For concise description, we will not repeat them here.
  • the embodiment of the present disclosure also provides a storage medium (computer-readable storage medium).
  • the storage medium here stores one or more programs.
  • the storage medium may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state hard disk; the memory may also include the above-mentioned types of memory. combination.
  • One or more programs in the storage medium can be executed by one or more processors to realize the above-mentioned automatic patterning method executed on the side of the sleep detection device.
  • the processor is configured to execute the sleep state detection model construction and user sleep state detection programs stored in the memory to implement the following steps of the sleep state detection model construction method and the user sleep state detection method executed on the sleep detection device side:
  • the sleep electrical signal of the user in a sleep state is obtained; a plurality of first electrical signals with a preset duration are intercepted from the sleep signal; and the corresponding first electrical signal is obtained from each of the first electrical signals.
  • a first breathing signal to obtain a plurality of the first breathing signals corresponding to a plurality of the first electrical signals.
  • the sleep electrical signal of the user in a sleep state is obtained through a piezoelectric sensor.
  • filter processing is performed on each of the first electrical signals according to the frequency range of the first electrical signals to obtain multiple first breathing signals corresponding to when the user is breathing.
  • a plurality of the first breathing signals are processed according to the time-domain characteristics of the signals generated when the user is breathing, and the Gaussian error of each first breathing signal is removed to obtain the processed multiple Each of the first breathing signals; perform short-time Fourier transform on each of the processed first breathing signals to determine the corresponding frequency domain characteristics; determine the corresponding user’s characteristics based on each of the frequency domain characteristics Respiration frequency; determine the corresponding multiple parameter information of the first respiration signal from each of the respiration frequencies; perform vectorization processing on the multiple parameter information of each of the first respiration signals to obtain each of the The multi-dimensional feature vectors corresponding to the first breathing signal are further obtained to obtain the multiple multi-dimensional feature vectors corresponding to the multiple first breathing signals.
  • a sleep electrical signal when the user is in a sleep state Acquire a sleep electrical signal when the user is in a sleep state; extract a plurality of first breathing signals from the sleep electrical signal; determine a plurality of parameter information from each of the first breathing signals, and based on a plurality of the parameter information , Generating a multi-dimensional feature vector corresponding to each of the first breathing signals, and then obtaining a plurality of the multi-dimensional feature vectors corresponding to a plurality of the first breathing signals; inputting the plurality of the multi-dimensional feature vectors into the sleep state detection model , So that the sleep state detection model outputs the sleep state of the user.
  • the sleep state is sent to the terminal device, so that the terminal device displays the sleep state.
  • the steps of the method or algorithm described in combination with the embodiments disclosed in this document can be implemented by hardware, a software module executed by a processor, or a combination of the two.
  • the software module can be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or all areas in the technical field. Any other known storage media.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Animal Behavior & Ethology (AREA)
  • Veterinary Medicine (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Surgery (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • Physiology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Pulmonology (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Theoretical Computer Science (AREA)
  • Psychiatry (AREA)
  • Signal Processing (AREA)
  • Mathematical Physics (AREA)
  • Anesthesiology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Fuzzy Systems (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

本公开实施例涉及一种睡眠状态检测模型构建、睡眠状态检测方法及装置，所述方法包括：获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型，由此，可以实现准确判断人体睡眠状态。

Description

睡眠状态检测模型构建、睡眠状态检测方法及装置
本公开要求于2020年06月08日提交中国专利局、申请号为202010517041.3、发明名称为“睡眠状态检测模型构建、睡眠状态检测方法及装置”的中国专利申请的优先权，其全部内容通过引用结合在本公开中。
技术领域
本公开实施例涉及智慧卧室领域，尤其涉及一种睡眠状态检测模型构建、睡眠状态检测方法及装置。
背景技术
随和科学技术的不断进步，物联网和智能家居发展迅猛，与此同时，人们越来越注重智能化时代给人们的生活和为人们的健康监测带来的便利，因此，关注人体睡眠状态的智慧卧室的发展受到极大关注，同时，睡眠数据的分析对于医生诊断与睡眠相关的疾病的作用也至关重要。
相关技术中，对于人体鼾动行为的判断方法是，基于声学传感器提取信号特征判断鼾动行为，但是，鼾动信号与呼吸信号频率重叠，与人体体动信号幅度近似，导致相关技术难以区分各种信号，会造成人体睡眠状态判断不准确。
发明内容
鉴于此，为解决上述无法准确判断人体睡眠状态的技术问题，本公开实施例提供一种睡眠状态检测模型构建、睡眠状态检测方法及装置。
第一方面，本公开实施例提供一种睡眠状态检测模型构建的方法，包括：
获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；
根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；
将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
在一个可能的实施方式中，所述方法还包括：
获取用户处于睡眠状态下的睡眠电信号；
从所述睡眠信号中截取多个预设时长的第一电信号；
从每个所述第一电信号中获取对应的第一呼吸信号，得到多个所述第一电信号对应的多个所述第一呼吸信号。
在一个可能的实施方式中，所述方法还包括：
通过压电传感器获取用户处于睡眠状态下的睡眠电信号。
在一个可能的实施方式中，所述方法还包括：
根据所述第一电信号的频度范围对每个所述第一电信号进行滤波处理，得到用户呼吸时对应的多个第一呼吸信号。
在一个可能的实施方式中，所述方法还包括：
根据用户呼吸时所产生信号的时域特征，对多个所述第一呼吸信号进行处理，去除每个所述第一呼吸信号的高斯误差，得到处理后的多个所述第一呼吸信号；
将处理后的每个所述第一呼吸信号执行短时傅里叶变换，确定对应的频域特征；
基于每个所述频域特征确定对应的所述用户的呼吸频率；
从每个所述呼吸频率中确定对应的所述第一呼吸信号的多个参数信 息；
对每个所述第一呼吸信号的多个参数信息进行向量化处理，得到每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量。
第二方面，本公开实施例提供一种用户睡眠状态检测方法，包括：
获取用户处于睡眠状态下的睡眠电信号；
从所述睡眠电信号中提取多个第一呼吸信号；
从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；
将多个所述多维特征向量输入至第一方面中任一项所构建的睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
在一个可能的实施方式中，所述方法还包括：
将所述睡眠状态发送至终端设备，以使所述终端设备恩显示所述睡眠状态。
第三方面，本公开实施例提供一种睡眠状态检测模型构建的装置，包括：
获取模块，被设置为获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；
数据处理模块，被设置为根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；
训练模块，被设置为将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
第四方面，本公开实施例提供一种用户睡眠状态检测装置，包括：
获取模块，被设置为获取用户处于睡眠状态下的睡眠电信号；
数据处理模块，被设置为从所述睡眠电信号中提取多个第一呼吸信号；
所述数据处理模块，还被设置为从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；
确定模块，被设置为将多个所述多维特征向量输入至睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
第五方面，本公开实施例提供一种睡眠检测设备，包括：处理器和存储器，所述处理器被设置为执行所述存储器中存储的睡眠状态检测模型构建和用户睡眠状态检测程序，以实现上述第一方面中任一项所述的睡眠状态检测模型构建方法和上述第二方面中任一项所述的用户睡眠状态检测方法。
第六方面，本公开实施例提供一种存储介质，包括：所述存储介质存储有一个或者多个程序，所述一个或者多个程序可被一个或者多个处理器执行，以实现上述第一方面中任一项所述的睡眠状态检测模型构建方法和上述第二方面中任一项所述的用户睡眠状态检测方法。
本公开实施例提供的睡眠状态检测模型构建的方案，通过获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型，由此方法可以实现通过检测人体生理信号进而准确判断人体睡眠状态， 为医疗辅助诊断提供了方便。
附图说明
图1为本公开实施例提供的一种睡眠状态检测模型构建的方法的流程示意图；
图2为本公开实施例提供的另一种睡眠状态检测模型构建的方法的流程示意图；
图3为本公开实施例提供的一种用户睡眠状态检测方法的流程示意图；
图4为本公开实施例提供的一种睡眠状态检测模型构建装置的结构示意图；
图5为本公开实施例提供的一种用户睡眠状态检测装置的结构示意图；
图6为本公开实施例提供的一种睡眠检测设备的结构示意图。
具体实施方式
为使本公开实施例的目的、技术方案和优点更加清楚，下面将结合本公开实施例中的附图，对本公开实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本公开一部分实施例，而不是全部的实施例。基于本公开中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本公开保护的范围。
为便于对本公开实施例的理解，下面将结合附图以具体实施例做进一步的解释说明，实施例并不构成对本公开实施例的限定。
图1为本公开实施例提供的一种睡眠状态检测模型构建的方法的流程示意图，如图1所示，该方法具体包括：
S11、获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠 状态对应的标签。
从压电传感器采集到的用户处于睡眠状态下的压电信号中，根据预设时长和呼吸频率范围，多次提取出多个第一呼吸信号，并将多个第一呼吸信号代表的用户的多个睡眠状态记录到睡眠状态标签中，其中，状态标签至少包括鼾动情况和呼吸暂停情况。
S12、根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量。
将每个第一呼吸信号进行数据处理，得到每个第一呼吸信号对应的多个参数信息，将每个第一呼吸信号对应的多个参数信息组成一个多维特征向量，进而得到多个第一呼吸信号对应的多个多维特征向量。
S13、将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
将多个第一呼吸信号对应的多个多维特征向量输入至初始模型中，模型会根据输入的数据计算得到输出的用户睡眠状态结果，该结果可能与预先记录的状态标签不一致，所以需要对模型进行深度学习训练，直到模型输出的用户睡眠状态结果与状态标签的相似度大于或等于设定的阈值(例如，90％)，则认为模型训练完成，将此时训练好的模型作为睡眠状态检测模型。
本方案提供一种睡眠状态检测模型构建方法，通过获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所 述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型，由此，可以实现通过采集人体生理信号分析并预测人体睡眠状态。
图2为本公开实施例提供的另一种睡眠状态检测模型构建的方法的流程示意图，如图2所示，该方法具体包括：
S21、通过压电传感器获取用户处于睡眠状态下的睡眠电信号。
睡眠检测设备置于床垫下方，日常非检测阶段，睡眠检测设备处于待机模式，当用户上床后，睡眠检测设备上安装的压电传感器感知到信号突变，睡眠检测设备开启正常工作模式，睡眠检测设备上安装的压电传感器开始采集压电信号，此时采集的压电信号包括用户处于清醒状态和睡眠状态所有时段的信号。
在一些实施方式中，需要从采集到的电信号里根据清醒时的呼吸频率和心率等参数，去除用户清醒状态下的电信号，保留用户处于睡眠状态下的睡眠电信号。
S22、从所述睡眠信号中截取多个预设时长的第一电信号。
从上述保留下来的睡眠电信号中，根据预设的时长，多次截取睡眠电信号，得到多个第一电信号，其中，每两次截取第一电信号的中间间隔时间，用于上一次截取到的第一电信号的数据处理，为了确保信号的采样有效性，去除单次测量误差，需要对每次测量值进行加权处理，有如下公式1：
Figure PCTCN2020140494-appb-000001
其中，x为信号测量值，n为滑动加权处理的宽度，ω为对应权重,其和为1。
预设时长可以是30s、40或45s，本方案根据实际使用过程进行调整，在此不做具体限定。
S23、根据所述第一电信号的频度范围对每个所述第一电信号进行滤 波处理，得到用户呼吸时对应的多个第一呼吸信号。
根据已知的一般人呼吸频率，预设一个IIR多阶滤波器，将截取到的多个第一电信号的每一个第一电信号都进行滤波处理，从每个第一电信号中分离出对应的第一呼吸信号。
S24、根据用户呼吸时所产生信号的时域特征，对多个所述第一呼吸信号进行处理，去除每个所述第一呼吸信号的高斯误差，得到处理后的多个所述第一呼吸信号。
由于压电传感器灵敏度较高，轻微的触碰即会产生大的跳变信号，因此需要根据用户呼吸时所产生信号的时域特征对采集到的第一呼吸信号数据进行处理，滤除跳变值，即去除每个所述第一呼吸信号的高斯误差，并通过多项式拟合与插值操作填补信号处理过程中的丢失值，保证信号连贯性，进而得到多个处理后的第一呼吸信号。
S25、将处理后的每个所述第一呼吸信号执行短时傅里叶变换，确定对应的频域特征。
S26、基于每个所述频域特征确定对应的所述用户的呼吸频率。
将处理后的多个第一呼吸信号进行短时傅里叶变换，将每个第一呼吸信号分解成频率谱，进而得到每个第一呼吸信号对应的频域特征。
在一些实施方式中，根据上述得到的每个第一呼吸信号对应的频域特征，确定对应的多个呼吸频率。
S27、从每个所述呼吸频率中确定对应的所述第一呼吸信号的多个参数信息。
根据呼吸频率得到第一呼吸信号的多个参数信息，参数信息至少包括，呼吸频率的高频、低频、幅值、能量、均值、方差、均方根误差或中位值等参数。
S28、对每个所述第一呼吸信号的多个参数信息进行向量化处理，得到每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼 吸信号对应的多个所述多维特征向量。
将每一个第一呼吸信号的多个参数信息组成一个多维特征向量，进而得到多个第一呼吸信号对应的多个多维特征向量。
在一些实施方式中，根据并行粒子群优化算法以睡眠状态检测模型的输出结果的准确率作为适应度，对所述参数信息进行优化处理，去除一个或多个非必要参数，以加快睡眠检测模型的训练收敛速度和提高睡眠检测模型的预测结果准确度。
本公开实施例提供的睡眠状态检测模型构建的方案，通过压电传感器获取用户处于睡眠状态下的睡眠电信号，对睡眠电信号进行处理，分离出呼吸信号，进而对呼吸信号进行数据处理，得到多个信号参数信息，组成多个多维特征向量，将多个多维特征向量输入至睡眠状态检测模型中，预测用户的睡眠状态，由此方法，可以实现通过将人体生理信号转化为电信号，进而对电信号进行处理，能够有效区分各种生理信号，达到准确判断人体睡眠状态的结果，为医疗辅助诊断提供了方便提高用户体验。
图3为本公开实施例提供的一种用户睡眠状态检测方法的流程示意图，如图3所示，该方法具体包括：
S31、获取用户处于睡眠状态下的睡眠电信号。
睡眠检测设备置于床垫下方，日常非检测阶段，睡眠检测设备处于待机模式，当用户上床后，睡眠检测设备上安装的压电传感器感知到信号突变，睡眠检测设备开启正常工作模式，睡眠检测设备上安装的压电传感器开始采集压电信号，此时采集的压电信号包括用户处于清醒状态和睡眠状态所有时段的信号，进而，从采集到的电信号里根据清醒时的呼吸频率和心率等参数，去除用户清醒状态下的电信号，保留用户处于睡眠状态下的睡眠电信号，本方案可以实现与人体非直接触式的压电传感器获取人体生理信号，提高用户体验。
S32、从所述睡眠电信号中提取多个第一呼吸信号。
从上述保留下来的睡眠电信号中，多次截取睡眠电信号，得到多个第一电信号，根据已知的一般人呼吸频率，预设一个IIR多阶滤波器，将截取到的多个第一电信号的每一个第一电信号都进行滤波处理，从每个第一电信号中分离出对应的第一呼吸信号。
S33、从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量。
本公开实施例中，首先，由于压电传感器灵敏度较高，轻微的触碰即会产生大的跳变信号，因此需要根据用户呼吸时所产生信号的时域特征对采集到的第一呼吸信号数据进行处理，去除每个所述第一呼吸信号的高斯误差，并通过多项式拟合与插值操作填补信号处理过程中的丢失值，保证信号连贯性，进而得到多个处理后的第一呼吸信号，将处理后的多个第一呼吸信号进行短时傅里叶变换，将每个第一呼吸信号分解成频率谱，进而得到每个第一呼吸信号对应的频域特征，在一些实施方式中，根据上述得到的每个第一呼吸信号对应的频域特征，确定对应的多个呼吸频率，根据呼吸频率得到第一呼吸信号的多个参数信息。
在一些实施方式中，将每个第一呼吸信号的多个参数信息组成一个多维特征向量，进而得到多个第一呼吸信号对应的多个多维特征向量。
S34、将多个所述多维特征向量输入至如权利要求1-5任一项所构建的睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
S35、将所述睡眠状态发送至终端设备，以使所述终端设备恩显示所述睡眠状态。
将上述得到的多个多维特征向量输入到训练好的睡眠状态检测模型中，输出检测出的用户的睡眠状态。
其中，睡眠状态包括呼吸频率(次/分)、鼾动时段和呼吸暂停时段。
在一些实施方式中，将输出结果反馈至用户持有的移动终端的app上，方便用户查看睡眠状态。
本公开实施例提供的用户睡眠状态检测方法，通过获取用户处于睡眠状态下的睡眠电信号；从所述睡眠电信号中提取多个第一呼吸信号；从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态，由此方法可以实现通过检测人体生理信号进而准确判断人体睡眠状态并且对于医生诊断与睡眠相关的疾病提供了方便。
图4为本公开实施例提供的一种睡眠状态检测模型构建装置的结构示意图，具体包括：
获取模块401，被设置为获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；
数据处理模块402，被设置为根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；
训练模块403，被设置为将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
获取模块，具体被设置为获取用户处于睡眠状态下的睡眠电信号；从所述睡眠信号中截取多个预设时长的第一电信号；从每个所述第一电信号中获取对应的第一呼吸信号，得到多个所述第一电信号对应的多个所述第一呼吸信号。
在一个可能的实施方式中，所述获取模块，还被设置为通过压电传感 器获取用户处于睡眠状态下的睡眠电信号。
在一个可能的实施方式中，所述获取模块，还被设置为根据所述第一电信号的频度范围对每个所述第一电信号进行滤波处理，得到用户呼吸时对应的多个第一呼吸信号。
数据处理模块，具体被设置为根据用户呼吸时所产生信号的时域特征，对多个所述第一呼吸信号进行处理，去除每个所述第一呼吸信号的高斯误差，得到处理后的多个所述第一呼吸信号；将处理后的每个所述第一呼吸信号执行短时傅里叶变换，确定对应的频域特征；基于每个所述频域特征确定对应的所述用户的呼吸频率；从每个所述呼吸频率中确定对应的所述第一呼吸信号的多个参数信息；对每个所述第一呼吸信号的多个参数信息进行向量化处理，得到每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量。
本实施例提供的睡眠状态检测模型构建的装置可以是如图4中所示的睡眠状态检测模型构建的装置，可执行如图1-2中睡眠状态检测模型构建的方法的所有步骤，进而实现图1-2所示睡眠状态检测模型构建的方法的技术效果，具体请参照图1-2相关描述，为简洁描述，在此不作赘述。
图5为本公开实施例提供的一种用户睡眠状态检测装置的结构示意图，具体包括：
获取模块501，被设置为获取用户处于睡眠状态下的睡眠电信号；
数据处理模块502，被设置为从所述睡眠电信号中提取多个第一呼吸信号；
所述数据处理模块502，还被设置为从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；
确定模块503，被设置为将多个所述多维特征向量输入至睡眠状态检 测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
本实施例提供的用户睡眠状态检测装置可以是如图5中所示的用户睡眠状态检测装置，可执行如图3中睡眠状态检测模型构建的方法的所有步骤，进而实现图3所示用户睡眠状态检测方法的技术效果，具体请参照图3相关描述，为简洁描述，在此不作赘述。
图6为本公开实施例提供的一种睡眠检测设备的结构示意图，图6所示的电子设备600包括：至少一个处理器601、存储器602、至少一个网络接口604和其他用户接口603。睡眠检测设备600中的各个组件通过总线系统605耦合在一起。可理解，总线系统605被设置为实现这些组件之间的连接通信。总线系统605除包括数据总线之外，还包括电源总线、控制总线和状态信号总线。但是为了清楚说明起见，在图6中将各种总线都标为总线系统605。
其中，用户接口603可以包括显示器、键盘或者点击设备(例如，鼠标，轨迹球(trackball)、触感板或者触摸屏等。
可以理解，本公开实施例中的存储器602可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(Read-Only Memory，ROM)、可编程只读存储器(Programmable ROM，PROM)、可擦除可编程只读存储器(Erasable PROM，EPROM)、电可擦除可编程只读存储器(Electrically EPROM，EEPROM)或闪存。易失性存储器可以是随机存取存储器(Random Access Memory，RAM)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(Static RAM，SRAM)、动态随机存取存储器(Dynamic RAM，DRAM)、同步动态随机存取存储器(Synchronous DRAM，SDRAM)、双倍数据速率同步动态随机存取存储器(Double Data Rate SDRAM，DDRSDRAM)、增强型同步动态随机存取存储器(Enhanced SDRAM，ESDRAM)、同步连接动态随机存取存储器(Synch  link DRAM，SLDRAM)和直接内存总线随机存取存储器(Direct Rambus RAM，DRRAM)。本文描述的存储器602旨在包括但不限于这些和任意其它适合类型的存储器。
在一些实施方式中，存储器602存储了如下的元素，可执行单元或者数据结构，或者他们的子集，或者他们的扩展集：操作系统6021和应用程序6022。
其中，操作系统6021，包含各种系统程序，例如框架层、核心库层、驱动层等，被设置为实现各种基础业务以及处理基于硬件的任务。应用程序6022，包含各种应用程序，例如媒体播放器(Media Player)、浏览器(Browser)等，被设置为实现各种应用业务。实现本公开实施例方法的程序可以包含在应用程序6022中。
在本公开实施例中，通过调用存储器602存储的程序或指令，具体的，可以是应用程序6022中存储的程序或指令，处理器601被设置为执行各方法实施例所提供的方法步骤，例如包括：
获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
在一个可能的实施方式中，获取用户处于睡眠状态下的睡眠电信号；从所述睡眠信号中截取多个预设时长的第一电信号；从每个所述第一电信号中获取对应的第一呼吸信号，得到多个所述第一电信号对应的多个所述第一呼吸信号。
在一个可能的实施方式中，通过压电传感器获取用户处于睡眠状态下 的睡眠电信号。
在一个可能的实施方式中，根据所述第一电信号的频度范围对每个所述第一电信号进行滤波处理，得到用户呼吸时对应的多个第一呼吸信号。
在一个可能的实施方式中，根据用户呼吸时所产生信号的时域特征，对多个所述第一呼吸信号进行处理，去除每个所述第一呼吸信号的高斯误差，得到处理后的多个所述第一呼吸信号；将处理后的每个所述第一呼吸信号执行短时傅里叶变换，确定对应的频域特征；基于每个所述频域特征确定对应的所述用户的呼吸频率；从每个所述呼吸频率中确定对应的所述第一呼吸信号的多个参数信息；对每个所述第一呼吸信号的多个参数信息进行向量化处理，得到每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量。
或，
获取用户处于睡眠状态下的睡眠电信号；从所述睡眠电信号中提取多个第一呼吸信号；从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
在一个可能的实施方式中，将所述睡眠状态发送至终端设备，以使所述终端设备恩显示所述睡眠状态。
上述本公开实施例揭示的方法可以应用于处理器601中，或者由处理器601实现。处理器601可能是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器601中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器601可以是通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field Programmable  Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本公开实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本公开实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件单元组合执行完成。软件单元可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器602，处理器601读取存储器602中的信息，结合其硬件完成上述方法的步骤。
可以理解的是，本文描述的这些实施例可以用硬件、软件、固件、中间件、微码或其组合来实现。对于硬件实现，处理单元可以实现在一个或多个专用集成电路(Application Specific Integrated Circuits，ASIC)、数字信号处理器(Digital Signal Processing，DSP)、数字信号处理设备(DSPDevice，DSPD)、可编程逻辑设备(Programmable Logic Device，PLD)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)、通用处理器、控制器、微控制器、微处理器、被设置为执行本公开所述功能的其它电子单元或其组合中。
对于软件实现，可通过执行本文所述功能的单元来实现本文所述的技术。软件代码可存储在存储器中并通过处理器执行。存储器可以在处理器中或在处理器外部实现。
本实施例提供的睡眠检测设备可以是如图6中所示的睡眠检测设备，可执行如图1-2中睡眠状态检测模型构建的方法和图3中用户睡眠状态检测方法的所有步骤，进而实现图1-2所示睡眠状态检测模型构建的方法和图3中用户睡眠状态检测方法的技术效果，具体请参照图1-2和图3相关描述，为简洁描述，在此不作赘述。
本公开实施例还提供了一种存储介质(计算机可读存储介质)。这里 的存储介质存储有一个或者多个程序。其中，存储介质可以包括易失性存储器，例如随机存取存储器；存储器也可以包括非易失性存储器，例如只读存储器、快闪存储器、硬盘或固态硬盘；存储器还可以包括上述种类的存储器的组合。
当存储介质中一个或者多个程序可被一个或者多个处理器执行，以实现上述在睡眠检测设备侧执行的自动打版方法。
所述处理器被设置为执行存储器中存储的睡眠状态检测模型构建和用户睡眠状态检测程序，以实现以下在睡眠检测设备侧执行的睡眠状态检测模型构建的方法和用户睡眠状态检测方法的步骤：
获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
在一个可能的实施方式中，获取用户处于睡眠状态下的睡眠电信号；从所述睡眠信号中截取多个预设时长的第一电信号；从每个所述第一电信号中获取对应的第一呼吸信号，得到多个所述第一电信号对应的多个所述第一呼吸信号。
在一个可能的实施方式中，通过压电传感器获取用户处于睡眠状态下的睡眠电信号。
在一个可能的实施方式中，根据所述第一电信号的频度范围对每个所述第一电信号进行滤波处理，得到用户呼吸时对应的多个第一呼吸信号。
在一个可能的实施方式中，根据用户呼吸时所产生信号的时域特征，对多个所述第一呼吸信号进行处理，去除每个所述第一呼吸信号的高斯误 差，得到处理后的多个所述第一呼吸信号；将处理后的每个所述第一呼吸信号执行短时傅里叶变换，确定对应的频域特征；基于每个所述频域特征确定对应的所述用户的呼吸频率；从每个所述呼吸频率中确定对应的所述第一呼吸信号的多个参数信息；对每个所述第一呼吸信号的多个参数信息进行向量化处理，得到每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量。
或，
获取用户处于睡眠状态下的睡眠电信号；从所述睡眠电信号中提取多个第一呼吸信号；从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
在一个可能的实施方式中，将所述睡眠状态发送至终端设备，以使所述终端设备恩显示所述睡眠状态。
专业人员应该还可以进一步意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本公开的范围。
结合本文中所公开的实施例描述的方法或算法的步骤可以用硬件、处理器执行的软件模块，或者二者的结合来实施。软件模块可以置于随机存储器(RAM)、内存、只读存储器(ROM)、电可编程ROM、电可擦除可编程ROM、寄存器、硬盘、可移动磁盘、CD-ROM、或技术领域内所公 知的任意其它形式的存储介质中。
以上所述的具体实施方式，对本公开的目的、技术方案和有益效果进行了进一步详细说明，所应理解的是，以上所述仅为本公开的具体实施方式而已，并不用于限定本公开的保护范围，凡在本公开的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本公开的保护范围之内。

Claims (11)

  1. 一种睡眠状态检测模型构建的方法，包括：
    获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；
    根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；
    将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
  2. 根据权利要求1所述的方法，其中，所述获取用户处于睡眠状态下的多个第一呼吸信号，包括：
    获取用户处于睡眠状态下的睡眠电信号；
    从所述睡眠信号中截取多个预设时长的第一电信号；
    从每个所述第一电信号中获取对应的第一呼吸信号，得到多个所述第一电信号对应的多个所述第一呼吸信号。
  3. 根据权利要求2所述的方法，其中，所述获取用户处于睡眠状态下的睡眠电信号，包括：
    通过压电传感器获取用户处于睡眠状态下的睡眠电信号。
  4. 根据权利要求3所述的方法，其中，所述从每个所述第一电信号中获取对应的第一呼吸信号，得到多个所述第一电信号对应的多个所述第一呼吸信号，包括：
    根据所述第一电信号的频度范围对每个所述第一电信号进行滤波处理，得到用户呼吸时对应的多个第一呼吸信号。
  5. 根据权利要求4所述的方法，其中，所述根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量， 得到多个所述第一呼吸信号对应的多个所述多维特征向量，包括：
    根据用户呼吸时所产生信号的时域特征，对多个所述第一呼吸信号进行处理，去除每个所述第一呼吸信号的高斯误差，得到处理后的多个所述第一呼吸信号；
    将处理后的每个所述第一呼吸信号执行短时傅里叶变换，确定对应的频域特征；
    基于每个所述频域特征确定对应的所述用户的呼吸频率；
    从每个所述呼吸频率中确定对应的所述第一呼吸信号的多个参数信息；
    对每个所述第一呼吸信号的多个参数信息进行向量化处理，得到每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量。
  6. 一种用户睡眠状态检测方法，包括：
    获取用户处于睡眠状态下的睡眠电信号；
    从所述睡眠电信号中提取多个第一呼吸信号；
    从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；将多个所述多维特征向量输入至如权利要求1-5任一项所构建的睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
  7. 根据权利要求6所述的方法，其中，所述方法还包括：
    将所述睡眠状态发送至终端设备，以使所述终端设备恩显示所述睡眠状态。
  8. 一种睡眠状态检测模型构建的装置，包括：
    获取模块，被设置为获取用户处于睡眠状态下的多个第一呼吸信号，以及所述睡眠状态对应的标签；
    数据处理模块，被设置为根据每个所述第一呼吸信号中的多个参数信息，确定每个所述第一呼吸信号对应的多维特征向量，得到多个所述第一呼吸信号对应的多个所述多维特征向量；
    训练模块，被设置为将多个所述多维特征向量输入至初始模型中，进行深度学习训练，直至所述初始模型的输出结果与所述标签的相似度大于或等于设定阈值，则确定所述初始模型训练完成，将训练好的初始模型作为睡眠状态检测模型。
  9. 一种用户睡眠状态检测装置，包括：
    获取模块，被设置为获取用户处于睡眠状态下的睡眠电信号；
    数据处理模块，被设置为从所述睡眠电信号中提取多个第一呼吸信号；
    所述数据处理模块，还被设置为从每个所述第一呼吸信号中确定多个参数信息，以及基于多个所述参数信息，生成每个所述第一呼吸信号对应的多维特征向量，进而得到多个所述第一呼吸信号对应的多个所述多维特征向量；
    确定模块，被设置为将多个所述多维特征向量输入至如睡眠状态检测模型中，以使所述睡眠状态检测模型输出所述用户的睡眠状态。
  10. 一种睡眠检测设备，包括：处理器和存储器，所述处理器被设置为执行所述存储器中存储的睡眠状态检测模型构建和用户睡眠状态检测程序，以实现权利要求1～5中任一项所述的睡眠状态检测模型构建方法或权利要求6～7中任一项所述的用户睡眠状态检测方法。
  11. 一种存储介质，所述存储介质存储有一个或者多个程序，所述一个或者多个程序可被一个或者多个处理器执行，以实现权利要求1～5中任一项所述的睡眠状态检测模型构建方法或权利要求6～7中任一项所述的用户睡眠状态检测方法。
PCT/CN2020/140494 2020-06-08 2020-12-29 睡眠状态检测模型构建、睡眠状态检测方法及装置 WO2021248882A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010517041.3 2020-06-08
CN202010517041.3A CN111814830B (zh) 2020-06-08 2020-06-08 睡眠状态检测模型构建、睡眠状态检测方法及装置

Publications (1)

Publication Number Publication Date
WO2021248882A1 true WO2021248882A1 (zh) 2021-12-16

Family

ID=72846489

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/140494 WO2021248882A1 (zh) 2020-06-08 2020-12-29 睡眠状态检测模型构建、睡眠状态检测方法及装置

Country Status (2)

Country Link
CN (1) CN111814830B (zh)
WO (1) WO2021248882A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116030534A (zh) * 2023-02-22 2023-04-28 中国科学技术大学 睡眠姿态模型的训练方法和睡眠姿态识别方法
CN116808391A (zh) * 2023-06-29 2023-09-29 北京理工大学 一种基于生理信号解码的睡眠唤醒方法和系统
CN117077013A (zh) * 2023-10-12 2023-11-17 之江实验室 一种睡眠纺锤波检测方法、电子设备、介质

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111814830B (zh) * 2020-06-08 2024-07-09 珠海格力电器股份有限公司 睡眠状态检测模型构建、睡眠状态检测方法及装置
CN113040710B (zh) * 2021-03-01 2022-05-17 珠海格力电器股份有限公司 一种睡眠状态分析方法及装置
CN115120837A (zh) * 2022-06-27 2022-09-30 慕思健康睡眠股份有限公司 基于深度学习的睡眠环境调节方法、系统、装置及介质
CN117017233B (zh) * 2023-10-08 2024-01-09 华南师范大学 一种睡眠呼吸暂停检测方法、装置、设备以及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140313913A1 (en) * 2011-08-16 2014-10-23 The University Of Utah Research Foundation Monitoring breathing via signal strength in wireless networks
CN106388818A (zh) * 2016-09-21 2017-02-15 广州视源电子科技股份有限公司 睡眠状态监测模型的特征信息提取方法和系统
CN110623665A (zh) * 2019-09-26 2019-12-31 川北医学院 一种智能睡眠时相检测与睡眠质量评估系统及方法
US20200100728A1 (en) * 2018-10-01 2020-04-02 Koninklijke Philips N.V. Systems and methods for using breath events in sleep staging
CN111012132A (zh) * 2019-12-17 2020-04-17 珠海格力电器股份有限公司 基于枕头的睡眠状态调整方法、装置及智能枕头
CN111814830A (zh) * 2020-06-08 2020-10-23 珠海格力电器股份有限公司 睡眠状态检测模型构建、睡眠状态检测方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20180075832A (ko) * 2016-12-27 2018-07-05 바이텔스 주식회사 수면 상태 모니터링 방법 및 장치
CN109259733A (zh) * 2018-10-25 2019-01-25 深圳和而泰智能控制股份有限公司 一种睡眠中呼吸暂停检测方法、装置和检测设备
CN111035367B (zh) * 2019-12-31 2021-05-18 华南师范大学 一种用于判断睡眠呼吸暂停的信号检测系统
CN111227792B (zh) * 2020-01-10 2022-11-01 京东方科技集团股份有限公司 呼吸暂停的检测方法、系统、电子设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140313913A1 (en) * 2011-08-16 2014-10-23 The University Of Utah Research Foundation Monitoring breathing via signal strength in wireless networks
CN106388818A (zh) * 2016-09-21 2017-02-15 广州视源电子科技股份有限公司 睡眠状态监测模型的特征信息提取方法和系统
US20200100728A1 (en) * 2018-10-01 2020-04-02 Koninklijke Philips N.V. Systems and methods for using breath events in sleep staging
CN110623665A (zh) * 2019-09-26 2019-12-31 川北医学院 一种智能睡眠时相检测与睡眠质量评估系统及方法
CN111012132A (zh) * 2019-12-17 2020-04-17 珠海格力电器股份有限公司 基于枕头的睡眠状态调整方法、装置及智能枕头
CN111814830A (zh) * 2020-06-08 2020-10-23 珠海格力电器股份有限公司 睡眠状态检测模型构建、睡眠状态检测方法及装置

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116030534A (zh) * 2023-02-22 2023-04-28 中国科学技术大学 睡眠姿态模型的训练方法和睡眠姿态识别方法
CN116808391A (zh) * 2023-06-29 2023-09-29 北京理工大学 一种基于生理信号解码的睡眠唤醒方法和系统
CN116808391B (zh) * 2023-06-29 2023-12-26 北京理工大学 一种基于生理信号解码的睡眠唤醒方法和系统
CN117077013A (zh) * 2023-10-12 2023-11-17 之江实验室 一种睡眠纺锤波检测方法、电子设备、介质
CN117077013B (zh) * 2023-10-12 2024-03-26 之江实验室 一种睡眠纺锤波检测方法、电子设备、介质

Also Published As

Publication number Publication date
CN111814830A (zh) 2020-10-23
CN111814830B (zh) 2024-07-09

Similar Documents

Publication Publication Date Title
WO2021248882A1 (zh) 睡眠状态检测模型构建、睡眠状态检测方法及装置
US10765399B2 (en) Programmable electronic stethoscope devices, algorithms, systems, and methods
CN108388912A (zh) 基于多传感器特征优化算法的睡眠分期方法
US11948690B2 (en) Pulmonary function estimation
Barton et al. Data reduction for cough studies using distribution of audio frequency content
JP2016529966A (ja) 咳の検知用、分析用および通信用のプラットフォーム
WO2021208656A1 (zh) 睡眠风险预测方法、装置和终端设备
CN114391807B (zh) 睡眠呼吸障碍分析方法、装置、设备及可读介质
WO2016168979A1 (zh) 一种生命体征分析方法与系统
Cohen-McFarlane et al. Comparison of silence removal methods for the identification of audio cough events
Doheny et al. Estimation of respiratory rate and exhale duration using audio signals recorded by smartphone microphones
Palaniappan et al. A telemedicine tool to detect pulmonary pathology using computerized pulmonary acoustic signal analysis
Shukla et al. An efficient heart sound segmentation approach using kurtosis and zero frequency filter features
CN111685731B (zh) 睡眠数据处理方法、装置、设备及存储介质
CN112155560A (zh) 基于实时心冲击信号的呼吸暂停检测方法及系统
Gahane et al. An analytical review of heart failure detection based on IoT and machine learning
Gong et al. Accurate cirrhosis identification with wrist-pulse data for mobile healthcare
US11918346B2 (en) Methods and systems for pulmonary condition assessment
EP4017351A1 (en) A method and apparatus for processing asthma patient cough sound for application of appropriate therapy
McLane et al. Comprehensive analysis system for automated respiratory cycle segmentation and crackle peak detection
Sharan et al. Automated Cough Sound Analysis for Detecting Childhood Pneumonia
Chattopadhyay et al. Sleep apnea diagnosis by DWT-based kurtosis, radar and histogram analysis of electrocardiogram
Pessoa et al. Automated respiratory sound analysis
KR102650641B1 (ko) 수면 단계 예측 장치 및 방법
US20230240641A1 (en) System and method for determining an auscultation quality metric

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20940085

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20940085

Country of ref document: EP

Kind code of ref document: A1