Lightweight multi-lead sleep monitor and data processing method

Abstract

The application provides a light-weight multi-lead sleep monitor and a data processing method. The light-weight multi-lead sleep monitor is designed by multiplexing a wearing structure, an electrode structure, an acquisition channel, a sensor and a data processing flow, so that the same wearing component or the same acquisition signal can undertake multiple sleep monitoring functions, thereby reducing the number of independent patches, exposed wires, sensors and wearing components, while still being able to acquire data required for sleep monitoring, such as sleep staging, respiratory events, blood oxygen, heart rate, body position, body movement and snoring sound.

Description

Technical Field

[0001] This invention relates to the field of neuromodulation technology, specifically to a lightweight polysomnography device and a data processing method. Background Technology

[0002] Polysomnography (PSG) is currently an important clinical technique for evaluating sleep structure, sleep-disordered breathing events, and related physiological abnormalities. Routine PSG typically involves placing EEG electrodes, EMG electrodes, EMG electrodes, respiratory airflow sensors, chest and abdominal breathing bands, blood oxygen sensors, position sensors, and snoring sensors on multiple sites on the subject's head, face, chest, abdomen, fingers, and lower limbs to collect various physiological signals, including EEG, EMG, heart rate, blood oxygenation, respiratory airflow, chest and abdominal movements, position, body movement, and snoring.

[0003] However, traditional polysomnography (PSG) devices typically suffer from problems such as a large number of electrode patches, complex lead wires, scattered wearing components, and cumbersome connection relationships. To reduce the complexity of wearing them, some existing portable sleep monitoring devices have adopted improvements by reducing the number of leads, reducing the types of sensors, or using wireless transmission. Current portable sleep monitoring devices usually rely on the detection results of a single sensor or a single channel as the primary basis for signal processing. For example, respiratory events are often determined by changes in nasal airflow or blood oxygenation, snoring events are often collected by a separate microphone, and body position and movement are often identified by a separate posture sensor.

[0004] Therefore, there is still a need in the existing technology for a lightweight polysomnography solution suitable for home sleep monitoring. This solution should be able to reduce the number of electrode patches, exposed wires, and independent sensors while retaining the key information required for sleep staging, respiratory event recognition, snoring recognition, and body position and movement analysis. Furthermore, it should be able to comprehensively evaluate and fuse the signal quality of different channels to improve the stability and effectiveness of sleep monitoring results. Summary of the Invention

[0005] The purpose of this invention is to provide a lightweight polysomnography device and data processing method, which solves the problems of traditional polysomnography devices having a large number of electrode patches, complex exposed leads, cumbersome wearing and operation, and low effective data collection rate in home scenarios. It also addresses the problems of existing portable sleep monitoring devices having a large number of independent sensors, insufficient reliability of single-channel judgment, and the tendency for motion artifacts or low-quality signals to cause misjudgments.

[0006] The first aspect of this invention provides a lightweight polysomnography device, comprising:

[0007] A flexible headband module includes a flexible band body, a plurality of bioelectric electrodes disposed on the flexible band body, and a flexible conductive circuit electrically connected to the plurality of bioelectric electrodes. A head-mounted main unit module is fixed to the flexible headband module and electrically connected to the plurality of bioelectric electrodes through the flexible conductive lines; the flexible headband module serves as both the fixing structure for the head-mounted main unit module and the load-bearing structure for the bioelectric electrodes and the flexible conductive lines. The bioelectric multiplexing acquisition module, integrated within the head-mounted host module, is configured to form multiple virtual leads based on different electrode combinations of the same set of bioelectric electrodes, so as to synchronously acquire and output electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) signals. At least one external multiplexing sensor module is communicatively connected to the head-mounted host module for acquiring physiological or motor posture signals of the subject independent of the bioelectrical signals; The system also includes a multi-source fusion analysis module, which is communicatively connected to the bioelectric multiplexing acquisition module and the external multiplexing sensing module. It is configured to perform fusion analysis on the electroencephalogram (EEG), electrooculogram (EOG), electromyogram (EMG), and physiological or motor posture signals, and output sleep monitoring results.

[0008] In a second aspect, the present invention provides a lightweight polysomnography data processing method, applied to the lightweight polysomnography monitor described above, the method comprising: Receive mixed bioelectric signals collected from the same set of bioelectric electrodes on the flexible headband module; Based on the preset lead combination relationship, the mixed bioelectric signals are differentially combined to construct multiple virtual leads in order to separate the electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) signals. The system receives physiological signals and motion posture signals of the subject from the external multiplexing sensing module, wherein the physiological signals of the subject include at least photoplethysmography (PPG) signals and nasal airflow pressure signals. Feature extraction is performed on each signal separately, including extracting the frequency energy features of the EEG signal, the eye movement features of the EEG signal, the muscle tone features of the EMG signal, and the blood oxygenation features of the photoplethysmography (PPG) signal and the respiratory features of the nasal airflow pressure signal. Signal quality is assessed based on the electrode impedance, noise level, or waveform continuity of each signal channel to obtain a quality score for each signal channel. Multi-source fusion correction is performed based on the quality score to output sleep monitoring results including sleep stages and respiratory events. Specifically, the multi-source fusion correction includes: when performing sleep stages, if the quality score of the bioelectric channel is low, the confidence of the EEG signal in the sleep stage results for the corresponding time period is reduced, and the analysis weights of body movement characteristics, heart rate variability characteristics, and photoplethysmography (PPG) signal characteristics are increased accordingly; when recognizing respiratory events, if the nasal airflow pressure signal changes abnormally and the motion posture signal is detected simultaneously indicating large body movement or loosening of the tested equipment, the confidence of the abnormal nasal airflow pressure signal as a respiratory event is reduced.

[0009] This invention provides a lightweight polysomnography (PSG) monitor and data processing method. A flexible headband integrates bioelectric electrodes, conductive circuitry, and a head-mounted main unit. A bioelectric multiplexing acquisition module constructs virtual leads for EEG, EEG, and EMG based on different differential combinations of the same set of electrodes. Simultaneously, an optical multiplexing sensing module (ring-type) calculates blood oxygen, heart rate, heart rate variability, and respiratory modulation characteristics from the same photoplethysmography (PPG) signal. A respiratory multiplexing acquisition module identifies respiratory events and snoring events simultaneously from a single nasal airflow pressure signal through frequency band separation. An inertial multiplexing sensing module detects body position, body movement, and motion artifacts. Finally, a multi-source fusion analysis module assesses the quality of signals from each channel and dynamically adjusts the analysis weights to achieve multi-source fusion correction. While significantly reducing the number of independent electrode patches, exposed wires, and sensors, it can still comprehensively collect core monitoring data such as sleep stages, respiratory events, snoring, and body position and movement, reducing the complexity of wearing it and interference with sleep. At the same time, through signal quality assessment and fusion correction, it effectively suppresses motion artifacts and misjudgments caused by single-channel signal quality degradation, improving the stability and reliability of home sleep monitoring.

[0010] The above description is merely an overview of the technical solution disclosed herein. In order to better understand the technical means of this disclosure and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic diagram of the system structure of the lightweight polysomnography monitor of the present invention; Figure 2 This is a schematic diagram of the structure of the head-mounted main unit module and the flexible headband module of the present invention; Figure 3 This is a schematic diagram of lead multiplexing in the bioelectric multiplexing acquisition module of the present invention; Figure 4 This is a schematic diagram of the structure of the ring-type optical multiplexing sensing module of the present invention; Figure 5 This is a schematic diagram of the respiratory multiplexing acquisition module of the present invention; Figure 6 This is a flowchart of the lightweight polysomnography monitoring method of the present invention; Figure 7 This is a flowchart of the multi-source signal fusion analysis of the present invention. Detailed Implementation

[0013] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. Various corresponding modifications and alterations can be made by those skilled in the art without departing from the spirit and essence of the present invention. Specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. For those skilled in the art, equivalent substitutions or conventional modifications made to the structure, connection relationships, signal processing methods, or specific parameters in the following embodiments without departing from the concept of the present invention should all fall within the scope of protection of the present invention.

[0014] This invention provides a lightweight polysomnography (PSG) monitor and its detection method. The lightweight PSG monitor utilizes a reusable design for its wearing structure, electrode structure, acquisition channels, sensors, and data processing flow. This allows the same wearing component or the same acquisition signal to perform multiple sleep monitoring functions, thereby reducing the number of independent patches, exposed wires, sensors, and wearing components. Simultaneously, it can still collect the data required for sleep monitoring, such as sleep stages, respiratory events, blood oxygen saturation, heart rate, body position, body movement, and snoring.

[0015] In this invention, "lightweighting" is not limited to simply reducing the number of hardware components, but also includes at least one of the following technical meanings: the flexible headband module simultaneously serves as a fixing structure, an electrode support structure, and a flexible conductive circuit support structure; the same set of bioelectric electrodes forms virtual leads related to electroencephalography (EEG), electrooculography (EOG), and electromyography (EMG) through different lead combinations; the same photoplethysmography (PPG) pulse wave signal is used for blood oxygen, heart rate, heart rate variability, and respiratory modulation analysis; the same inertial signal is used for body position recognition, body movement detection, wearing status judgment, and motion artifact marking; and the same nasal airflow pressure signal is used for respiratory event recognition and snoring event recognition.

[0016] The lightweight polysomnography device described in this invention can be used for home sleep monitoring, initial screening of sleep apnea, sleep quality assessment, chronic disease follow-up management, and remote medical auxiliary analysis.

[0017] Example 1 This embodiment provides a lightweight polysomnography device.

[0018] like Figure 1 As shown, the lightweight polysomnography monitor 1 includes a head-mounted main unit module 10, a flexible headband module 20, a bioelectric multiplexing acquisition module 30, an optical multiplexing sensing module 40, an inertial multiplexing sensing module 50, a respiratory multiplexing acquisition module 60, and a multi-source fusion analysis module 70.

[0019] The head-mounted main unit module 10 is worn on the user's forehead and houses functional components such as main control, signal acquisition, data processing, wireless communication, power supply, and storage. The flexible headband module 20 connects to the head-mounted main unit module 10, securing it to the user's head and serving as a support structure for both bioelectric electrodes and flexible conductive circuits. The bioelectric multiplexing acquisition module 30 uses the same set of bioelectric electrodes on the flexible headband module 20 to form multiple virtual leads for acquiring EEG, EEG, and EMG signals. The optical multiplexing sensing module 40 is worn on the user's finger and acquires PPG signals. The inertial multiplexing sensing module 50 acquires data related to the user's posture, body movement, and wearing status. The respiratory multiplexing acquisition module 60 acquires nasal airflow pressure signals. The multi-source fusion analysis module 70 performs preprocessing, feature extraction, signal quality assessment, sleep staging, respiratory event recognition, snoring event recognition, body position and movement analysis, and report generation on the data acquired by each module.

[0020] Therefore, this embodiment does not achieve weight reduction by simply reducing the number of monitoring items, but rather by reusing the structure and signals, enabling a small number of wearable components to obtain the core information required for polysomnography monitoring.

[0021] 1. Headset-mounted main unit module like Figure 1 and Figure 2 As shown, the head-mounted main unit module 10 includes a main unit housing 11, a main control circuit board 12, a bioelectric analog front end 13, a signal processing unit 14, a wireless communication unit 15, a power supply unit 16, a storage unit 17, and a status indication unit 18.

[0022] The main unit housing 11 can be an arc-shaped housing adapted to the curvature of the user's forehead. A flexible cushioning layer can be provided on the side of the main unit housing 11 closest to the skin to reduce the pressure on the forehead area during sleep. The two sides of the main unit housing 11 are connected to the flexible headband module 20, so that the head-mounted main unit module 10 can be stably held in the user's forehead area.

[0023] The main control circuit board 12 is housed inside the main unit housing 11. The bioelectric analog front end 13, signal processing unit 14, wireless communication unit 15, power supply unit 16, and storage unit 17 can be mounted on the main control circuit board 12. The status indicator unit 18 is located on the outside of the main unit housing 11 or in the visible area of ​​the main unit housing 11, and is used to display statuses such as power-on, data acquisition in progress, low battery, communication abnormality, wearing abnormality, or sensor detachment.

[0024] The bioelectric analog front-end 13 receives bioelectric signals transmitted from the flexible headband module 20 and performs differential amplification, filtering, and analog-to-digital conversion on the bioelectric signals. The signal processing unit 14 processes the acquired data locally. The wireless communication unit 15 communicates with the mobile terminal 80 or the server 90. The power supply unit 16 supplies power to the head-mounted host module 10 and related acquisition modules. The storage unit 17 stores raw monitoring data, feature data, event data, signal quality data, and sleep monitoring reports.

[0025] In one embodiment, the head-mounted host module 10 employs a low-power operating mode. This low-power operating mode may include: reducing the sampling frequency of some channels when the signal is stable; turning off or reducing the operating frequency of the wireless communication unit 15 when wireless communication is idle; storing monitoring data in the storage unit 17 when the mobile terminal 80 or server 90 is temporarily unavailable; and maintaining low sensor drive power when the quality of the optical or respiratory signals meets requirements. These methods can extend the continuous monitoring time of the device.

[0026] 2. Flexible headband module like Figure 2 As shown, the flexible headband module 20 includes a flexible band body 21, a flexible conductive line 22, a forehead electrode 23, a side temporal electrode 24, a reference electrode 25, a grounding electrode 26, and a head circumference adjustment structure 27.

[0027] The flexible band 21 can be made of elastic fabric, silicone, thermoplastic elastomer, or other flexible materials suitable for long-term skin contact. The flexible conductive lines 22 can be embedded inside the flexible band 21, or located on the side of the flexible band 21 closest to or furthest from the skin. The flexible conductive lines 22 are electrically connected to the forehead electrode 23, the temporal electrode 24, the reference electrode 25, and the ground electrode 26, respectively, and transmit the signals collected by each electrode to the head-mounted main unit module 10.

[0028] The forehead electrode 23 is positioned on the flexible headband module 20 corresponding to the user's forehead. The forehead electrode 23 may include a first forehead electrode and a second forehead electrode, positioned on the left and right sides of the forehead, respectively. The lateral temporal electrode 24 is positioned on the flexible headband module 20 corresponding to the user's temporal side. The reference electrode 25 may be positioned behind the user's ear, near the mastoid process, or on the posterior side of the headband. The grounding electrode 26 may be positioned on the side or back of the headband.

[0029] The head circumference adjustment structure 27 can be a Velcro, buckle, slide, elastic tightening structure or knob adjustment structure, used to adjust the tightness of the flexible band 21 around the user's head, so that each electrode maintains stable contact with the skin.

[0030] In this embodiment, the flexible headband module 20 simultaneously serves as the fixation function of the head-mounted main unit module 10, the arrangement function of the bioelectric electrodes, the load-bearing function of the flexible conductive lines 22, and the electrode contact pressure adjustment function. Compared with traditional polysomnography devices, this structure can reduce the number of independently pasted electrodes and exposed leads, which is beneficial to improving the wearing convenience and sleep comfort in the home environment.

[0031] 3. Bioelectricity Multiplexing Acquisition Module like Figure 3 As shown, the bioelectric multiplexing acquisition module 30 includes an electrode interface unit 31, a lead selection unit 32, a differential amplification unit 33, a filtering unit 34, an analog-to-digital conversion unit 35, and an electrode contact detection unit 36.

[0032] Electrode interface unit 31 is connected to forehead electrode 23, temporal electrode 24, reference electrode 25, and ground electrode 26, respectively. Lead selection unit 32 can be implemented using an analog switch matrix, multiplexer, or configurable analog front end, and is used to select different electrode combinations from multiple electrodes according to preset lead combination relationships to form multiple virtual leads.

[0033] In one embodiment, the lead selection unit 32 uses the differential signal between the first frontal electrode and the reference electrode 25 as the first EEG lead, the differential signal between the second frontal electrode and the reference electrode 25 as the second EEG lead, the differential signal between the first frontal electrode and the second frontal electrode as the electrooculogram (EOG) lead, and the differential signal between the frontal electrode 23 and the lateral temporal electrode 24 as the electromyographic lead for the frontalis muscle or temporalis muscle.

[0034] In another embodiment, when the flexible headband module 20 also includes a mandibular extension electrode, the lead selection unit 32 can also use the differential signal between the mandibular extension electrode and the reference electrode 25 as a mandibular electromyographic lead to assist in rapid eye movement sleep phase recognition, bruxism event recognition, or muscle tone change analysis.

[0035] The differential amplifier unit 33 amplifies the signals of each virtual lead. The filter unit 34 sets the corresponding filter frequency band according to the purpose of different virtual leads. The analog-to-digital converter unit 35 converts analog bioelectrical signals into digital bioelectrical signals. The electrode contact detection unit 36 ​​detects the contact status between each electrode and the user's skin.

[0036] The electrode contact detection unit 36 ​​can determine whether the electrode contact is normal by impedance detection, signal saturation detection, baseline drift detection, or noise intensity detection. When an abnormal electrode contact is detected, the head-mounted host module 10 can prompt the user to adjust the wearing position through the status indicator unit 18 or the mobile terminal 80; at the same time, the multi-source fusion analysis module 70 reduces the weight of the corresponding virtual lead in subsequent analysis.

[0037] With the above structure, the same set of bioelectric electrodes can be reused as multiple virtual leads to acquire related signals of electroencephalography (EEG), electrooculography (EOG), and electromyography (EMG), thereby preserving the key bioelectric information required for sleep stages while reducing the number of electrodes.

[0038] 4. Optical multiplexing sensor module like Figure 4 As shown, the optical multiplexing sensing module 40 can be a ring-shaped structure. The optical multiplexing sensing module 40 includes a ring housing 401, a light-emitting unit 402, a photoelectric receiving unit 403, a PPG acquisition circuit 404, a fingertip inertial sensor 405, and a ring communication unit 406.

[0039] The ring housing 401 is worn on the user's finger. The light-emitting unit 402 may include a red light-emitting device and an infrared light-emitting device, and may further include a green light-emitting device. The photoelectric receiving unit 403 is used to receive light signals reflected or transmitted through finger tissue. The PPG acquisition circuit 404 is used to amplify, filter, and perform analog-to-digital conversion on the photoelectric signal output by the photoelectric receiving unit 403 to obtain a PPG signal. The fingertip inertial sensor 405 is used to detect the finger movement state or the ring wearing state. The ring communication unit 406 is used to transmit PPG data and fingertip movement data to the head-mounted host module 10, the mobile terminal 80, or the server 90.

[0040] In this embodiment, the PPG signal acquired by the optical multiplexing sensing module 40 is not limited to calculating blood oxygen saturation, but can also be used to calculate heart rate, heart rate variability, pulse wave amplitude changes, PPG respiratory modulation characteristics, and hypoperfusion state. The fingertip inertial sensor 405 is used to determine ring loosening, large finger movements, or PPG motion artifacts.

[0041] For example, when the fingertip inertial sensor 405 detects rapid finger movement and a sudden change in the PPG waveform, the multi-source fusion analysis module 70 marks the corresponding time period as a low-confidence PPG time period and reduces the confidence of blood oxygenation drop events or heart rate abnormality events within that time period.

[0042] 5. Inertial multiplexing sensor module The inertial multiplexing sensor module 50 includes a three-axis accelerometer and / or a gyroscope. The inertial multiplexing sensor module 50 can be located within the head-mounted main unit module 10, within the optical multiplexing sensor module 40, or both.

[0043] The inertial multiplexing sensor module 50, located within the head-mounted main unit module 10, is used to detect head posture, changes in sleep position, turning over events, headband loosening, and motion artifacts. The inertial multiplexing sensor module 50, located within the optical multiplexing sensor module 40, is used to detect finger movements, ring loosening, and PPG artifacts.

[0044] In one implementation, a triaxial accelerometer determines the user's position based on changes in the direction of gravity. When the user changes from a supine to a lateral position, the attitude angle output by the triaxial accelerometer changes, and the multi-source fusion analysis module 70 records this change as a position change event. When this position change event has a temporal correlation with changes in the frequency of respiratory events, the multi-source fusion analysis module 70 outputs a position-related respiratory risk warning.

[0045] In another implementation, when the triaxial accelerometer detects a short-term large-amplitude head movement, the multi-source fusion analysis module 70 marks the EEG signal, EOS signal, and EMG signal within the corresponding time period as high-risk motion artifacts, so as to avoid misjudging motion artifacts as awakening events or EMG events.

[0046] 6. Respiratory Reuse Acquisition Module like Figure 5 As shown, the respiratory multiplexing acquisition module 60 includes a nasal cannula 61, a pressure sensor 62, a pressure acquisition circuit 63, and a respiratory signal processing unit 64.

[0047] A nasal cannula 61 is positioned near the user's nostrils to sense changes in airflow pressure during inhalation and exhalation. A pressure sensor 62 is connected to the nasal cannula 61 to convert changes in nasal airflow pressure into an electrical signal. A pressure acquisition circuit 63 amplifies, filters, and performs analog-to-digital conversion on the electrical signal output from the pressure sensor 62. A respiratory signal processing unit 64 extracts the respiratory cycle, respiratory amplitude, airflow waveform flattening characteristics, and high-frequency vibration characteristics from the nasal airflow pressure signal.

[0048] In this embodiment, the nasal airflow pressure signal acquired by the respiratory multiplexing acquisition module 60 is used for both respiratory event recognition and snoring event recognition. Specifically, the low-frequency components of the nasal airflow pressure signal are used to extract information related to inspiration, expiration, apnea, and hypoventilation; the high-frequency vibration components in the nasal airflow pressure signal are used to identify snoring or upper airway vibration. Therefore, this embodiment can complete snoring event recognition based on the same nasal airflow pressure signal without the need for a separate snoring microphone.

[0049] In an optional implementation, the respiratory multiplexing acquisition module 60 can also be connected to a chest-abdominal breathing belt. The chest-abdominal breathing belt is used to acquire respiratory motion signals from the chest and abdomen to help differentiate between obstructive apnea, central apnea, and mixed apnea.

[0050] Specifically, the respiratory signal processing unit 64 can perform dual-channel parallel separation processing on the raw nasal airflow pressure signal Praw(t) output by the pressure acquisition circuit 63 to obtain baseline respiratory airflow information and high-frequency snoring information from the same pressure signal.

[0051] The first channel is used for airflow feature extraction of respiratory events. The respiratory signal processing unit 64 can process the original nasal airflow pressure signal Praw(t) using a low-pass filter to obtain a low-frequency respiratory waveform Presp(t) that characterizes the baseline respiratory airflow changes. The feature extraction unit 73 calculates the respiratory baseline based on the temporal amplitude change of Presp(t). When the waveform amplitude of Presp(t) decreases by more than a preset proportion within a preset duration, the respiratory event identification unit 76 marks this time period as a candidate event for apnea; when the waveform amplitude of Presp(t) decreases to the level of the hypoventilation judgment condition, and is accompanied by a decrease in blood oxygen, heart rate changes, or awakening-related features, the respiratory event identification unit 76 marks this time period as a candidate event for low-pass ventilation.

[0052] The second channel is used for high-frequency snoring feature extraction. The respiratory signal processing unit 64 can use a bandpass filter to process the original nasal airflow pressure signal Praw(t) and separate the high-frequency pressure fluctuation signal Psnore(t) caused by upper airway soft tissue vibration. Subsequently, the feature extraction unit 73 can calculate the short-time energy envelope E(t) of Psnore(t) based on a sliding time window to determine whether there are snoring candidate segments.

[0053] In one specific implementation, when the amplitude of the energy envelope E(t) exceeds the dynamic environmental noise background threshold Thbg and the duration is within the preset snoring duration range, the snoring event identification unit 77 initially marks it as a snoring candidate segment. To reduce misjudgments caused by mattress friction, turning over, or nasal cannula collision, the multi-source fusion analysis module 70 performs time-domain correspondence verification between the snoring candidate segment and the low-frequency respiratory waveform Presp(t). If the snoring candidate segment has a preset degree of temporal overlap with the inspiratory phase of Presp(t), it is determined to be a valid snoring event; otherwise, it is marked as a noise candidate segment or a low-confidence snoring segment.

[0054] Through the above dual-channel processing method, this embodiment can simultaneously obtain information related to apnea, hypopnea, and snoring using a single nasal airflow pressure channel, which helps to reduce the configuration of independent snoring sensors and reduce installation errors between channels.

[0055] 7. Multi-source fusion analysis module like Figure 7 As shown, the multi-source fusion analysis module 70 can be located within the head-mounted host module 10, or partially within the mobile terminal 80 or the server 90. The multi-source fusion analysis module 70 includes a preprocessing unit 71, a virtual lead construction unit 72, a feature extraction unit 73, a signal quality assessment unit 74, a sleep staging unit 75, a respiratory event recognition unit 76, a snoring event recognition unit 77, a body position and movement analysis unit 78, and a report generation unit 79.

[0056] The preprocessing unit 71 is used to filter, correct baselines, suppress power frequency interference, synchronize time, and remove outliers from bioelectric signals, PPG signals, inertial signals, and nasal airflow pressure signals.

[0057] The virtual lead construction unit 72 is used to construct EEG leads, EOG leads, and EMG leads based on different electrode combination signals output by the lead selection unit 32.

[0058] The feature extraction unit 73 is used to extract brainwave frequency band energy, eye movement features, electromyographic tension features, blood oxygen features, heart rate features, heart rate variability features, respiratory amplitude features, snoring vibration features, and body position and movement features.

[0059] The signal quality assessment unit 74 is used to score the signal quality of each channel. The quality score may include bioelectrical quality score, PPG quality score, respiratory quality score, and inertial quality score. The signal quality assessment unit 74 can determine the signal quality of each channel based on electrode impedance, PPG perfusion index, nasal airflow waveform continuity, motion amplitude, signal saturation state, and sensor detachment state.

[0060] The sleep staging unit 75 is used to determine the user's sleep stages based on electroencephalogram (EEG), eye movement (EMG), electromyogram (EMG), body movement (BPM), and heart rate variability (HRV). Sleep stages may include wakefulness, light sleep, deep sleep, and rapid eye movement (REM) sleep, and may be further divided into W (wake) sleep, N1 (sleep) sleep, N2 (sleep) sleep, N3 (sleep) sleep, and REM sleep.

[0061] The respiratory event recognition unit 76 is used to recognize apnea events, hypoventilation events, and decreased blood oxygenation events based on nasal airflow pressure signals, changes in blood oxygen, changes in heart rate, body position information, and optional chest and abdominal respiratory effort signals.

[0062] The snoring event recognition unit 77 is used to recognize snoring events based on the high-frequency vibration components in the nasal airflow pressure signal.

[0063] The body position and movement analysis unit 78 is used to identify supine, left lateral, right lateral, prone, turning over, large body movements and micro body movements based on inertial signals.

[0064] The report generation unit 79 is used to generate a sleep monitoring report. The sleep monitoring report may include total sleep duration, sleep efficiency, sleep stage results, apnea-hypopnea index, blood oxygen saturation index, lowest blood oxygen saturation, average heart rate, number of snoring sounds, body position distribution, body position-related respiratory event indicators, and signal quality evaluation.

[0065] Example 2 This embodiment provides a lightweight polysomnography method. The method can be applied to the lightweight polysomnography monitor 1 described in Embodiment 1.

[0066] like Figure 6 and Figure 7 As shown, the lightweight polysomnography method includes the following steps.

[0067] S1. Wear the lightweight polysomnography device and complete the initialization.

[0068] The headband module 10 is worn on the user's forehead area, and the flexible headband module 20 is wrapped around the user's head, so that the forehead electrode 23 is located in the user's forehead area, the lateral temporal electrode 24 is located in the user's temporal area, and the reference electrode 25 and the ground electrode 26 are located behind the ear, near the mastoid process, on the side of the head, or in the occipital area, respectively.

[0069] The optical multiplexing sensor module 40 is worn on the user's finger, so that the light-emitting unit 402 and the photoelectric receiving unit 403 correspond to the skin of the finger. The nasal cannula 61 of the respiratory multiplexing acquisition module 60 is placed near the user's nostrils, so that the nasal cannula 61 can collect the changes in airflow pressure generated during inhalation and exhalation.

[0070] After wearing the device, the head-mounted main unit module 10 is activated. The head-mounted main unit module 10 initializes the bioelectric multiplexing acquisition module 30, the optical multiplexing sensing module 40, the inertial multiplexing sensing module 50, and the respiratory multiplexing acquisition module 60.

[0071] The initialization process includes: detecting electrode contact status, detecting ring wearing status, detecting nasal airflow pressure sensor status, detecting battery power, detecting storage space, establishing a communication connection with mobile terminal 80 or server 90, and completing time synchronization of each channel.

[0072] When any module fails to initialize, the head-mounted host module 10 will issue a prompt through the status indicator unit 18 or the mobile terminal 80, prompting the user to adjust the wearing position or reconnect the corresponding module.

[0073] S2. Collect multi-source sleep monitoring signals and construct virtual leads.

[0074] After initialization, the lightweight polysomnography monitor 1 enters the monitoring state and continuously collects multi-source sleep monitoring signals during the user's sleep.

[0075] Specifically, the bioelectric multiplexing acquisition module 30 acquires bioelectric signals through the forehead electrode 23, the lateral temporal electrode 24, the reference electrode 25, and the ground electrode 26. The lead selection unit 32 generates multiple virtual leads according to a preset lead combination relationship. The multiple virtual leads include at least EEG virtual leads, Eoptometry virtual leads, and EMG virtual leads.

[0076] In one embodiment, a first EEG virtual lead is formed between the first frontal electrode and the reference electrode 25; a second EEG virtual lead is formed between the second frontal electrode and the reference electrode 25; an Ophthalmic virtual lead is formed between the first frontal electrode and the second frontal electrode; and an electromyographic virtual lead is formed between the frontal electrode 23 and the lateral temporal electrode 24.

[0077] The optical multiplexing sensor module 40 acquires PPG signals through the light-emitting unit 402 and the photoelectric receiving unit 403, and obtains pulse wave data through the PPG acquisition circuit 404. The inertial multiplexing sensor module 50 acquires acceleration data, angular velocity data, and attitude data of the head and / or fingers. The respiratory multiplexing acquisition module 60 acquires nasal airflow pressure signals through the nasal cannula 61 and the pressure sensor 62.

[0078] In an optional implementation, chest breathing effort signals and abdominal breathing effort signals can also be collected via a chest and abdomen breathing belt, or leg movement signals can be collected via a leg sensor.

[0079] Through the above steps, the same set of bioelectric electrodes can be reused in multiple bioelectric signal channels, thereby reducing the number of electrodes while achieving multichannel sleep signal acquisition.

[0080] S3. Perform preprocessing, feature extraction, and signal quality assessment on multi-source signals.

[0081] The preprocessing unit 71 in the multi-source fusion analysis module 70 preprocesses the acquired multi-source signals.

[0082] For bioelectrical signals, preprocessing may include bandpass filtering, power frequency interference suppression, baseline drift correction, outlier removal, and sampling rate unification. For PPG signals, preprocessing may include DC drift removal, pulse peak detection, outlier pulse wave removal, low perfusion detection, and motion artifact labeling. For nasal airflow pressure signals, preprocessing may include respiratory waveform smoothing, respiratory cycle detection, pressure zero-point correction, and high-frequency vibration component extraction. For inertial signals, preprocessing may include gravity component estimation, attitude angle calculation, motion intensity calculation, and initial screening of body movement events.

[0083] After preprocessing, the signals from each channel are unified onto the same time axis for multi-source fusion analysis.

[0084] Feature extraction unit 73 extracts features from various preprocessed signals. For EEG virtual leads, it extracts energy in different frequency bands, frequency-to-energy ratios, slow-wave features, spindle wave candidate features, and signal complexity features. For EEG virtual leads, it extracts rapid eye movement features, slow eye movement features, eye movement amplitude, eye movement frequency, and eye movement duration. For EMG virtual leads, it extracts root mean square value of EMG, average power of EMG, muscle tone reduction features, and transient EMG burst features. For PPG signals, it extracts blood oxygen saturation, heart rate, heart rate variability, pulse wave amplitude changes, PPG respiratory modulation features, hypoperfusion features, and blood oxygen decrease candidate events. For nasal airflow pressure signals, it extracts respiratory cycle, respiratory amplitude, airflow descent ratio, airflow waveform flattening features, high-frequency vibration features, apnea candidate events, low-frequency climate candidate events, and snoring candidate events. For inertial signals, it extracts features related to supine, left lateral, right lateral, prone, turning over, gross body movement, microbody movement, loosening of the device, and motion artifacts.

[0085] The signal quality assessment unit 74 scores the signal quality of each channel based on the contact status, continuity, noise level, motion interference, and waveform rationality. For the bioelectric channel, the signal quality score is determined based on electrode impedance, baseline drift, signal saturation, power frequency interference, and motion artifacts. For the PPG channel, the signal quality score is determined based on perfusion index, pulse peak stability, light intensity saturation, ring wearing status, and finger movement intensity. For the respiratory channel, the signal quality score is determined based on the continuity of the nasal airflow pressure waveform, the rationality of the respiratory cycle, nasal cannula dislodgement characteristics, and the stability of the pressure signal amplitude. For the inertial channel, the signal quality score is determined based on the continuity of sensor output, the rationality of attitude changes, and motion amplitude changes.

[0086] When the signal quality score of a certain channel is lower than the preset threshold, the multi-source fusion analysis module 70 reduces the weight of that channel in subsequent analysis and marks the corresponding time period as a low-confidence time period.

[0087] S4. Perform sleep stage analysis, respiratory events analysis, snoring events analysis, and postural and physical activity analysis.

[0088] The sleep staging unit 75 classifies users' sleep based on EEG characteristics, EEG characteristics, EMG characteristics, body movement characteristics, heart rate variability characteristics, and signal quality scores for each channel.

[0089] In one implementation, the sleep staging unit 75 outputs wakefulness, light sleep, deep sleep, and REM sleep. In another implementation, the sleep staging unit 75 outputs W stage, N1 stage, N2 stage, N3 stage, and REM stage. The sleep staging unit 75 can employ a rule-based model, a machine learning model, or a deep learning model. When the quality of the EEG virtual leads is high, the sleep staging unit 75 increases the weight of EEG features in sleep staging; when the EEG virtual leads are affected by motion artifacts, the sleep staging unit 75 increases the relative weight of electrooculography (EOG) features, electromyography (EMG) features, body movement features, and heart rate variability (HRV) features.

[0090] For example, when the virtual electrooculography (EOG) lead detects rapid eye movement (REM) characteristics, the virtual electromyography (EMG) lead detects decreased muscle tone characteristics, and the EEG characteristics match those of REM sleep, the sleep staging unit 75 classifies the corresponding time period as REM sleep. When the inertial signal detects continuous body movement and the EEG signal exhibits high-frequency, low-amplitude characteristics, the sleep staging unit 75 classifies the corresponding time period as wakefulness or an arousal event.

[0091] The respiratory event recognition unit 76 identifies apnea and hypoventilation events based on nasal airflow pressure signals, changes in blood oxygenation, changes in heart rate, body position information, and optional chest and abdominal respiratory effort signals. When the nasal airflow pressure signal significantly decreases or disappears within a preset time, the respiratory event recognition unit 76 identifies it as a candidate apnea event. When the nasal airflow pressure signal decreases by a certain proportion and is accompanied by a decrease in blood oxygenation, changes in heart rate, awakening-related characteristics, or changes in sleep stages, the respiratory event recognition unit 76 identifies it as a candidate hypoventilation event.

[0092] When a candidate event of sleep apnea or a candidate event of low-flux climate corresponds temporally with an event of decreased blood oxygenation, the respiratory event identification unit 76 increases the confidence level of the respiratory event. When an abnormal nasal airflow pressure signal is accompanied by a gross motor event, the respiratory event identification unit 76 decreases the confidence level of the abnormality as a real respiratory event.

[0093] The snoring event recognition unit 77 identifies snoring events based on the high-frequency vibration component in the nasal airflow pressure signal. When the high-frequency vibration component is time-dependent on the respiratory cycle, and the duration, amplitude, or frequency of the high-frequency vibration component meets preset conditions, the snoring event recognition unit 77 identifies that time period as a snoring candidate event.

[0094] In an optional implementation, when the system receives a chest and abdominal breathing effort signal, the breathing event identification unit 76 further classifies the apnea event based on the chest and abdominal breathing effort signal. If nasal airflow disappears but chest and abdominal breathing effort still exists, it is determined to be a candidate event for obstructive apnea; if nasal airflow disappears and chest and abdominal breathing effort disappears simultaneously, it is determined to be a candidate event for central apnea; if both characteristics are present, it is determined to be a candidate event for mixed apnea.

[0095] The body position and movement analysis unit 78 identifies the user's sleeping position and movement events based on the acceleration and angular velocity data output by the inertial multiplexing sensor module 50. The body position may include supine, left lateral, right lateral, and prone positions. The movement events may include turning over events, gross body movement events, and micro body movement events.

[0096] When the body position and movement analysis unit 78 detects that respiratory events are concentrated in the supine position, the multi-source fusion analysis module 70 marks the corresponding respiratory events as position-related respiratory events. When respiratory events are significantly reduced in the lateral position, the report generation unit 79 can generate position-related prompts in the sleep monitoring report. When body movement events occur simultaneously with abnormalities in EEG, EEG, or PPG, the multi-source fusion analysis module 70 marks the corresponding time period as a motion artifact time period and lowers the confidence level of the corresponding abnormal event.

[0097] S5. Perform multi-source fusion correction.

[0098] The multi-source fusion analysis module 70 performs fusion correction on sleep stage results, respiratory event results, snoring event results, blood oxygenation drop events, body position and movement events, and signal quality scores.

[0099] The fusion correction may include at least one of the following: Increase the confidence level of the respiratory event when the nasal airflow pressure signal indicates a respiratory event, the PPG signal indicates a decrease in blood oxygen, and there are no obvious body movement artifacts during this period. When nasal airflow pressure signals are abnormal, but inertial signals indicate gross movement or nasal cannula loosening, reduce the confidence level of this abnormality as a respiratory event. When the PPG channel quality score is below the threshold, the weight of the blood oxygenation decline event in the respiratory event determination is reduced. When the quality of the bioelectric channel is low, the confidence level of the sleep stage results for the corresponding time period is reduced, and the weights of body movement characteristics, heart rate variability characteristics, and PPG characteristics are increased. When the characteristics of rapid eye movement (REM), low tone of muscle (EMG), and electroencephalography (EEG) are consistent, the confidence level of REM sleep stage determination is improved. When snoring events occur simultaneously with flattened airflow waveforms, increasing upper airway resistance increases the confidence level of the relevant cues.

[0100] Through the above-mentioned fusion correction, the risk of misjudgment by a single channel can be reduced while minimizing the number of hardware components, thereby improving the stability of lightweight sleep monitoring results.

[0101] S6. Generate a sleep monitoring report and provide abnormal alerts and data uploads.

[0102] The report generation unit 79 generates a sleep monitoring report based on the results of the multi-source fusion analysis.

[0103] The sleep monitoring report may include total sleep duration, sleep efficiency, sleep latency, number of awakenings, sleep stage ratio, REM sleep ratio, deep sleep ratio, apnea-hypopnea index, blood oxygen saturation index, lowest blood oxygen, average blood oxygen, average heart rate, heart rate variability index, number of snoring events, body position distribution, number of supine breathing events, number of lateral breathing events, signal quality score, and effective monitoring duration.

[0104] The sleep monitoring report can be stored in the storage unit 17 or transmitted to the mobile terminal 80 or the server 90 via the wireless communication unit 15. The mobile terminal 80 can display the sleep monitoring report, and the server 90 can be used for remote management, doctor viewing, or long-term follow-up analysis.

[0105] When the multi-source fusion analysis module 70 detects a preset abnormal condition, the head-mounted host module 10 or the mobile terminal 80 can issue an abnormality prompt. The preset abnormal condition may include blood oxygen continuously falling below a preset threshold, frequent occurrence of apnea events, abnormal heart rate, sensor detachment, severe signal quality degradation, or insufficient battery power.

[0106] When wireless communication conditions are good, the head-mounted host module 10 uploads the monitoring data to the server 90 in real time or in batches. When wireless communication conditions are unstable, the head-mounted host module 10 temporarily stores the monitoring data in the storage unit 17 and continues to upload it after communication is restored.

[0107] Example 3 This embodiment provides a preferred lightweight configuration method.

[0108] In this embodiment, the lightweight polysomnography monitor 1 includes a head-mounted main unit module 10, a flexible headband module 20, a ring-type optical multiplexing sensor module 40, a respiratory multiplexing acquisition module 60, and an inertial multiplexing sensor module 50 disposed within the head-mounted main unit module 10.

[0109] The flexible headband module 20 is equipped with two forehead electrodes 23, one lateral temporal electrode 24, one reference electrode 25, and one ground electrode 26. Through the bioelectric multiplexing acquisition module 30, two virtual EEG leads, one virtual EOG lead, and one virtual EMG lead can be formed based on these electrodes.

[0110] The ring-type optical multiplexing sensor module 40 is used to acquire PPG signals and output blood oxygen saturation, heart rate, heart rate variability and PPG respiratory modulation characteristics.

[0111] The respiratory multiplexing acquisition module 60 is used to acquire nasal airflow pressure signals and extract respiratory cycle, respiratory amplitude, apnea candidate events, low-pass climate candidate events and snoring candidate events from the same nasal airflow pressure signal.

[0112] The inertial multiplexing sensor module 50 is used to identify body position, turning over, body movement, and loosening of clothing.

[0113] This embodiment can obtain multidimensional sleep monitoring information such as electroencephalography (EEG), electrooculography (EOG), electromyography (EMG), blood oxygen saturation, heart rate, respiration, body position, body movement, and snoring using only a few of the aforementioned components. Compared with traditional polysomnography devices, this embodiment reduces the number of independent electrode patches, chest and abdominal straps, leg movement electrodes, and exposed leads, making it more suitable for home sleep monitoring and long-term follow-up scenarios.

[0114] Example 4 This embodiment provides an enhanced configuration method.

[0115] Based on Embodiment 3, the lightweight polysomnography monitor 1 also includes a chest and abdominal breathing effort module and a leg movement monitoring module.

[0116] The thoracic-abdominal breathing effort module includes a chest breathing band and an abdominal breathing band for collecting chest and abdominal breathing movements. The thoracic-abdominal breathing effort module is wirelessly connected to the head-mounted main unit module 10 or connected via a flexible cable.

[0117] The leg movement monitoring module includes inertial sensors or electromyographic electrodes placed on the lower limbs to detect periodic limb movements, leg electromyographic activity, and the correlation between limb movements and respiratory events.

[0118] In this enhanced configuration, the respiratory event recognition unit 76 can further distinguish between obstructive, central, and mixed respiratory events based on nasal airflow pressure signals and chest and abdominal respiratory effort signals. The body position and movement analysis unit 78 can identify periodic limb movement events based on signals output by the leg movement monitoring module.

[0119] This enhanced configuration is suitable for scenarios requiring higher monitoring accuracy or clinical diagnostic assistance.

[0120] Through the above technical solution, the present invention has at least the following technical effects: First, by setting the flexible headband module 20 to simultaneously serve as the fixed structure of the head-mounted main unit module 10, the carrier structure of the bioelectric electrodes, and the carrier structure of the flexible conductive lines, and by constructing EEG, EOG, and EMG virtual leads based on the same set of bioelectric electrodes through the bioelectric multiplexing acquisition module 30, this invention can obtain multiple types of bioelectric signals required for sleep staging while reducing the number of independent electrode patches and exposed lead wires, thereby reducing the complexity of wearing and improving the wearing comfort and effective acquisition rate in home sleep monitoring scenarios.

[0121] Secondly, the PPG signal, inertial signal, and nasal airflow pressure signal are multiplexed by the optical multiplexing sensing module 40, the inertial multiplexing sensing module 50, and the respiratory multiplexing acquisition module 60, respectively. The multi-source fusion analysis module 70 combines the signal quality assessment results to perform fusion correction on sleep stages, respiratory events, snoring events, and body position and movement events. This invention can improve the stability of sleep monitoring results and reduce the risk of misjudgment caused by single channel dropout, motion artifacts, or low-quality signals while reducing the configuration of independent sensors.

[0122] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0123] In this disclosure, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The block diagrams of devices, apparatuses, devices, and systems involved in this disclosure are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as "comprising," "including," "having," etc., are open-ended terms meaning "including but not limited to," and are used interchangeably with them. The terms "or" and "and" as used herein refer to the terms "and / or," and are used interchangeably with them unless the context clearly indicates otherwise. The term "such as" as used herein refers to the phrase "such as but not limited to," and is used interchangeably with it.

[0124] Additionally, as used herein, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not imply that the described example is preferred or better than other examples.

[0125] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.

[0126] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.

[0127] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0128] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A lightweight polysomnography monitor, characterized in that, include: A flexible headband module includes a flexible headband body, a plurality of bioelectric electrodes disposed on the flexible headband body, and a flexible conductive circuit electrically connected to the plurality of bioelectric electrodes. A head-mounted main unit module is fixed to the flexible headband module and electrically connected to the multiple bioelectric electrodes through the flexible conductive lines; The flexible headband module serves as both the fixing structure for the head-mounted main unit module and the load-bearing structure for the bioelectric electrodes and flexible conductive circuits. The bioelectric multiplexing acquisition module, integrated within the head-mounted host module, is configured to form multiple virtual leads based on different combinations of the multiple bioelectric electrodes, so as to synchronously acquire and output electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) signals. At least one external multiplexing sensor module is communicatively connected to the head-mounted host module for acquiring physiological or motor posture signals of the subject independent of bioelectrical signals; The system also includes a multi-source fusion analysis module, which is communicatively connected to the bioelectric multiplexing acquisition module and the external multiplexing sensing module. It is configured to perform fusion analysis on the electroencephalogram (EEG), electrooculogram (EOG), electromyogram (EMG), and physiological or motor posture signals, and output sleep monitoring results.

2. The lightweight polysomnography monitor according to claim 1, characterized in that, The plurality of bioelectric electrodes include a forehead electrode, a lateral temporal electrode, a reference electrode, and a ground electrode; the forehead electrode is disposed on the flexible headband module at the position corresponding to the user's forehead, the lateral temporal electrode is disposed at the position corresponding to the user's temporal side, the reference electrode is disposed at the position corresponding to the user's behind the ear, near the mastoid process, or at the posterior side of the head, and the ground electrode is disposed at the position corresponding to the user's side or back of the head.

3. The lightweight polysomnography monitor according to claim 2, characterized in that, The bioelectric multiplexing acquisition module includes a lead selection unit; the forehead electrode includes a first forehead electrode and a second forehead electrode. The lead selection unit is configured to: extract the differential signal between the first frontal electrode and the reference electrode as a virtual EEG lead; extract the differential signal between the first frontal electrode and the second frontal electrode as a virtual Ophthalmic lead; and extract the differential signal between the frontal electrode and the lateral temporal electrode as a virtual electromyographic lead.

4. The lightweight polysomnography device according to any one of claims 1 to 3, characterized in that, The external multiplexing sensing module includes an optical multiplexing sensing module, which includes a light-emitting unit and a photoelectric receiving unit worn on the user's limb end for collecting photoplethysmography (PPG) signals. The multi-source fusion analysis module is further configured to calculate blood oxygen saturation, heart rate variability, and respiratory modulation characteristics based on the same photoplethysmography pulse wave signal.

5. The lightweight polysomnography monitor according to claim 4, characterized in that, The optical multiplexing sensing module is further equipped with a fingertip inertial sensor; The multi-source fusion analysis module is further configured to mark or correct the confidence attenuation of motion artifacts in the photoplethysmography signal based on the motion signal collected by the fingertip inertial sensor.

6. The lightweight polysomnography device according to any one of claims 1 to 3, characterized in that, The external multiplexing sensing module includes a respiratory multiplexing acquisition module, which includes a nasal cannula and a pressure sensor for acquiring a single nasal airflow pressure signal. The multi-source fusion analysis module is configured to extract low-frequency airflow descent features to identify respiratory events and extract high-frequency airflow vibration features to identify snoring events by performing frequency band separation on the single nasal airflow pressure signal.

7. The lightweight polysomnography device according to any one of claims 1 to 3, characterized in that, The external multiplexing sensing module includes an inertial multiplexing sensing module, which is disposed within the head-mounted main unit module or worn independently on the user's body, and is used to collect acceleration signals and / or angular velocity signals. The multi-source fusion analysis module is configured to identify at least one of the user's sleeping position, turning over event, body movement event, and motion artifact event based on the acceleration signal and / or angular velocity signal.

8. The lightweight polysomnography monitor according to claim 1, characterized in that, The multi-source fusion analysis module includes a signal quality assessment unit; the signal quality assessment unit is used to score the quality of the signals output by the bioelectric multiplexing acquisition module and the external multiplexing sensing module respectively; the multi-source fusion analysis module is used to dynamically adjust the analysis weight of each signal in sleep stage and / or respiratory event identification based on the quality score.

9. The lightweight polysomnography monitor according to claim 5, characterized in that, The at least one external multiplexing sensing module includes an optical multiplexing sensing module, an inertial multiplexing sensing module, and a respiratory multiplexing acquisition module; the multi-source fusion analysis module is further configured to perform multi-source fusion correction: when the nasal airflow pressure signal indicates a respiratory event, and the photoplethysmography signal simultaneously indicates a decrease in blood oxygenation, and the inertial multiplexing sensing module does not detect significant body motion artifacts, the confidence level of the respiratory event determination is increased.

10. A lightweight data processing method for polysomnography, characterized in that, The method, applied to the lightweight polysomnography device as described in any one of claims 1-9, comprises: The system receives mixed bioelectric signals from the same set of bioelectric electrodes on the flexible headband module; according to the preset lead combination relationship, it performs differential combination on the mixed bioelectric signals to construct multiple virtual leads, so as to separate the electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) signals. The system receives physiological signals and motion posture signals of the subject from the external multiplexing sensing module, wherein the physiological signals of the subject include at least photoplethysmography (PPG) signals and nasal airflow pressure signals. Feature extraction is performed on each signal: frequency energy features of the EEG signal, eye movement features of the EOS signal, muscle tone features of the EMG signal, blood oxygenation features of the photoplethysmography (PPG) signal, and respiratory features of the nasal airflow pressure signal are extracted. Signal quality is assessed based on the electrode impedance, noise level, or waveform continuity of each signal channel to obtain a quality score for each signal channel. Multi-source fusion correction is performed based on the quality score to output sleep monitoring results including sleep stages and respiratory events. Specifically, the multi-source fusion correction includes: when performing sleep stages, if the quality score of the bioelectric channel is low, the confidence of the EEG signal in the sleep stage results for the corresponding time period is reduced, and the analysis weights of body movement characteristics, heart rate variability characteristics, and photoplethysmography (PPG) signal characteristics are increased accordingly; when recognizing respiratory events, if the nasal airflow pressure signal changes abnormally and the motion posture signal is detected simultaneously indicating large body movement or loosening of the tested equipment, the confidence of the abnormal nasal airflow pressure signal as a respiratory event is reduced.

Images