33 results about "Abnormal breathing" patented technology

The invention discloses a child abnormal breathing early warning and remote cooperation system for a family environment, and relates to the technical field of breathing data processing, and the child abnormal breathing early warning and remote cooperation system comprises a data acquisition module, a first data calculation module, a first data judgment module, a second data judgment module, a second data calculation module, a third data judgment module and a data output module. The system sequentially performs primary risk assessment, hidden risk mining and potential decompensation index calculation on the basis of a multi-modal signal processing technology by synchronously acquiring a breathing sound signal, a non-contact thoracoabdominal movement signal and a contact thoracoabdominal movement signal, so that graded early warning is realized, and a remote cooperation mechanism is automatically triggered according to the early warning level. According to the method, the hidden risk in the early stage of respiratory compensation of children can be effectively identified, and advanced early warning and timely intervention of abnormal respiration can be realized in a family environment.

The invention relates to the technical field of multi-parameter physiological signal monitoring, in particular to an intelligent laryngeal mask and respiration monitoring system which comprises a laryngeal mask body and a multi-parameter optical fiber measuring system. The optical fiber sensing unit is embedded into the laryngeal mask main body and comprises an acoustic sensing part and an environment sensing part with a functional coating; the signal processing unit demodulates and couples an original signal, combines physical parameters of the coating and inputs the original signal into a multi-parameter coupling correction model, and decouples independent and accurate multiple parameters; and the target thickness of the coating is determined through acoustic correction error minimization and comprehensive calculation that the sensitivity of each environmental parameter reaches the standard. The system can generate an abnormal ventilation alarm and feed back to control the breathing machine. Multi-parameter synchronous monitoring can be achieved, coating interference is eliminated, and breathing management safety and practicability are improved. In the actual application process, the abnormal breathing function caused by abnormal breathing machine state or abnormal laryngeal mask position of a user can be found in time, the purpose of early discovery and early intervention can be achieved, and malignant events are avoided.

The invention relates to the field of breath sound monitoring, and discloses a tracheal intubation device and a breath sound monitoring system.The tracheal intubation device comprises a tracheal catheter used for being inserted into an airway of a patient; the first sound detection piece is an optical fiber sound sensor, the first sound detection piece is installed on the tracheal catheter, the first sound detection piece is used for detecting breathing sound wave information in the airway of a patient, and the first sound detection piece is used for being in communication connection with the monitoring host so as to transmit the detected breathing sound wave information to the monitoring host. Therefore, the first sound detection piece is installed on the tracheal catheter, the first sound detection piece can be directly located in the internal environment of the airway of the patient, the accuracy of monitoring breathing sound and sound wave information of the patient through the tracheal intubation device can be improved, and the first sound detection piece transmits the breathing sound and sound wave information to the monitoring host in real time; the medical staff can monitor the breathing condition of the patient in real time through the monitoring host, find out abnormal breathing of the patient in time and carry out treatment.

The invention discloses a breathing equipment data processing method which comprises the following steps: synchronously collecting multichannel physiological signals through a built-in sensor of breathing equipment, carrying out noise processing and data standardization, and segmenting according to a time window to form data fragments; extracting breathing amplitude, frequency and morphological variability features of each data fragment, and combining to generate a multi-dimensional dynamic feature vector sequence; constructing a breathing state modeling system based on a hidden Markov model, training model parameters through a Baum-Welch algorithm, and calculating a prospective risk transition score by using a forward algorithm; and establishing an adaptive threshold strategy based on the risk transition score, carrying out structured labeling on the original respiration data, and generating processing data containing a prospective risk early warning label. Technical support can be provided for intelligent upgrading and personalized treatment of respiratory treatment equipment, and the accuracy, predictability and clinical practicability of respiratory anomaly detection are improved.

The invention discloses an intelligent ward risk assessment method and system, a product and a medium. The method comprises the following steps: synchronously monitoring breathing audio and mattress pressure signals of a patient, and identifying discrete action events with remarkable amplitude and duration from pressure data; further judging whether the breathing of the patient is abnormal or not before and after the action events occur, for example, the breathing rhythm is interrupted for a long time or the breathing cycle obviously deviates from a baseline; when the two conditions of the violent action and the abnormal breathing cooperatively occur, an effective cooperative event record is confirmed; according to the amplitude and the duration of the action in the event and the severity of the abnormal breathing, weighted summation is carried out through a preset risk weight, and a single event risk score is calculated; and accumulating all event risk scores in an evaluation period to obtain a comprehensive risk score, comparing the comprehensive risk score with a risk threshold, and outputting a risk level. By implementing the technical scheme provided by the invention, the accuracy of clinical compound action risk early warning of the patient is improved.

The present disclosure provides a method for assessing cardiorespiratory health. The method comprising: generating mechanical movement data associated with a torso of a subject's body using a plurality of ballistocardiography (BCG) sensors; filtering the mechanical movement data to extract respiratory signal waveforms and heart signal waveforms; decomposing the respiratory signal waveforms into thoracic, diaphragmatic, and abdominal breathing components based on waveform characteristics and anatomical location of the plurality of BCG sensors; analyzing phase differences between the thoracic, diaphragmatic, and abdominal breathing components over time to quantify a degree of synchrony or asynchrony; evaluating relative contributions of thoracic, diaphragmatic, and abdominal breathing components to an overall breathing effort; identifying a breathing abnormality based on the analysed phase differences and the evaluated relative contributions; and detecting a potential cardiorespiratory condition based on the identified breathing abnormality.

The present invention relates to an operating method of a lung monitoring system operated by at least one processor, the operating method comprising the steps of: acquiring a plurality of electrical impedance tomography (EIT) images of the chest of a patient and a plurality of respiratory sound intensity images generated on the basis of respiratory sound signals acquired from the chest of the patient; mapping, on the basis of a plurality of pieces of respiratory interval information acquired by binning a respiratory signal of the patient, an EIT image and a respiratory sound intensity image corresponding to each respiratory interval to each other; and inputting, into a respiratory anomaly prediction model, at least one pair of the EIT image and the respiratory sound intensity image included in each respiratory interval, to acquire a prediction result for respiratory anomalies of the patient.

The invention provides a radar non-inductive respiration monitoring system and method based on a passive wrinkled airbag, and the system comprises a sleep monitoring pillow and a respiration recognition and abnormity monitoring subsystem, the respiration recognition and abnormity monitoring subsystem comprises a signal monitoring and coupling module which processes original monitoring data, and transmits the processed data to the sleep monitoring pillow; a human body respiration direct measurement signal and an air bag coupling amplification signal are obtained; the state recognition and inversion module is used for carrying out in-place state recognition, sleeping posture recognition and airbag state inversion on the coupled double-domain fusion respiration candidate data to obtain state constraint data; the respiration reconstruction and feature extraction module is used for performing screening and feature extraction on the double-domain fusion respiration candidate data to obtain respiration feature data; and the abnormity identification and confirmation module is used for identifying the apnea candidate event and the abnormal event, obtaining an abnormity judgment result and carrying out graded alarm pushing. According to the invention, through the non-inductive pillow type structure and radar-air bag double-domain fusion perception, breathing abnormity intelligent monitoring with high robustness and low false alarm rate is realized.

The present application relates to the technical field of sleep intervention, in particular to an intelligent sleep instrument and a sleep instrument control method, the system comprising a respiratory mutation recognition module, a slow wave initiation linkage module, a pharyngeal cavity contraction evaluation module, an intervention section positioning module and a delay trigger control module. In the present application, the extreme points densely appearing in the continuous cycle are accurately positioned as the key nodes of respiratory abnormalities through the identification of the mutation of nasal respiratory flow, and the slow wave initiation stage closely related to the respiratory mutation is screened out and marked as the linkage period in combination with the power curve change trend of the electroencephalogram signal. The pharyngeal cavity state is dynamically evaluated through the change of acoustic echo in the period, the intelligent postponement and delay processing of the intervention behavior are realized through the comparison of the trigger point of the intervention action and the boundary of the platform section, the accuracy of the sleep intervention timeliness and the adaptability of the intervention execution are improved, and the respiratory event control effect and the sleep continuity are significantly improved.

The application discloses a mesh nebulizer. The nebulization control method of the mesh nebulizer comprises the following steps: acquiring breathing data of a user; determining a breathing state of the user and an end time point of exhalation of each breathing cycle based on the breathing data; the breathing state comprises stable breathing and abnormal breathing; and controlling a running state of a nebulization sheet based on the breathing state and the end time point of exhalation. According to the breathing state of the user and the end time point of exhalation, the running state of the nebulization sheet is accurately controlled, so that the drug utilization rate is improved, and intelligent and personalized nebulization treatment is realized.

The invention discloses a sleep quality monitoring and abnormal breathing early warning system based on a generative adversarial network, and particularly relates to the technical field of intelligent health monitoring. The system comprises a signal acquisition module, a data processing module, a generative adversarial network model library and an early warning module, the signal acquisition module is used for acquiring a breathing vibration signal, a body movement signal, a blood oxygen signal, a photoelectric volume pulse wave signal and a wrist body movement signal in parallel through a non-contact mattress piezoelectric sensor and an intelligent bracelet; the data processing module generates comparison data through a first processing method and a second processing method, and outputs a comprehensive feature vector after integration processing; a discriminator of the generative adversarial network model library adopts a multi-task learning architecture and outputs an abnormal index; and the early warning module realizes third-level early warning according to the abnormal index. Through multi-mode signal fusion processing and a non-contact monitoring mode, the accuracy, the reliability and the user experience of sleep breathing abnormity monitoring are effectively improved.

To provide a support surface moving device that can give the user a good breathing condition. [Solution] A support surface moving device that moves a support surface that supports the user's body comprises a respiratory abnormality detection unit that detects respiratory abnormalities of the user on the support surface, and a movement control unit that controls the movement of the support surface. The support surface has an upper body support section that supports the upper body of the user on the support surface, and the movement control unit moves the upper body support section based on a respiratory abnormality index that indicates the number of times the respiratory abnormality of the user has been detected by the respiratory abnormality detection unit within a predetermined period.

The invention discloses a breathing abnormity monitoring method and system based on electroencephalogram and pressure collaborative awareness, and relates to the field of biological data processing.The method comprises the steps that electroencephalogram signals and chest surface muscle mechanical vibration signals, synchronously collected in the breathing stage, of a user are obtained, and data preprocessing is conducted respectively; inputting the preprocessed electroencephalogram signal and the preprocessed chest vibration signal into a trained breathing type recognition model, and firstly, respectively inputting an electroencephalogram branch and a chest vibration branch in a double-branch structure to carry out feature extraction to obtain an electroencephalogram feature and a chest vibration feature; the electroencephalogram features and the chest vibration features are spliced through a splicing layer and then input into an ECA module for feature fusion to obtain fusion features, the fusion features pass through a full-connection classifier to obtain a corresponding breathing type, and if the breathing type is one of enzootic pneumonia, tachypnea, shortness of breath and apnea, it is determined that breathing is abnormal. The method solves the problem that characteristic similar states such as shallow and slow breathing and apnea are easy to misjudge.

The invention discloses a transmission device for assisting a small animal to autonomously breathe hyperpolarized gas, which is characterized in that a sealed breathing mask is used for replacing a trachea cannula, and stable pressure of atmospheric pressure at the breathing mask is used for replacing periodic pressure change under high-pressure forced ventilation at the cannula; the animal spontaneous respiration gas amount is used for replacing the manually set fixed gas amount, and the state of inhaled gas is controlled under spontaneous respiration to replace fully controlled mechanical ventilation in the respiration process; the invention further discloses a method for assisting the small animal in autonomously breathing the hyperpolarized gas, the device can supply the gas amount required by autonomously breathing to the experimental animal according to the self-breathing state of the experimental animal, and the gas amount inhaled by the experimental animal each time is determined by the self-breathing state, so that the experimental animal can autonomously breathe in a free state; lung expansion in an abnormal breathing state can be avoided, so that a more accurate ventilation function and qi-blood exchange function evaluation result is obtained.

This invention provides a method and system for detecting child sleep in a smart cockpit of a new energy vehicle. The sleep detection method includes: collecting respiratory signals from a target child sitting in a child seat, and verifying respiratory interruptions based on the respiratory signals to identify abnormal respiratory events; in response to the abnormal respiratory event to be verified, acquiring the target child's body surface temperature distribution information and the target child's body posture relative to the child seat; performing correlation analysis between the temporal characteristics of the abnormal respiratory event and the temporal temperature change sequence of the mouth and nose area in the body surface temperature distribution information to generate a suffocation risk verification result; and controlling the child seat to perform corresponding posture adjustments based on the risk level of the suffocation risk verification result. The technical problem solved by this invention is that existing child seats, as passive safety devices, have limited functionality and cannot handle the risk of abnormal breathing that occurs when a child is sleeping while traveling in a vehicle.

The invention relates to a breath pattern classification and evaluation method for chest and abdomen tumor patients before radiotherapy. The method comprises the steps that respiratory movement signals of the chest and the abdomen of a patient are collected, and original signal data reflecting changes of chest and abdomen respiratory movement along with time are obtained; preprocessing and feature extraction are conducted on the original signal data, a plurality of breathing feature parameters are obtained, and the feature parameters comprise a chest breathing movement amplitude parameter, an abdomen breathing movement amplitude parameter, breathing frequency, a chest and abdomen breathing synchronism index and a chest and abdomen breathing movement phase difference; inputting the characteristic parameters into a breathing mode classification model based on machine learning, and identifying and classifying the breathing mode of the patient to determine that the breathing mode of the patient belongs to an abdominal breathing mode, a thoracic breathing mode, a mixed breathing mode or other abnormal breathing modes; and evaluating the breathing mode of the patient based on the classification result and the characteristic parameters. According to the invention, the safety and effectiveness of the radiotherapy process are improved.

Embodiments disclosed herein improve digital stethoscopes and their application and operation. A first method detects of a respiratory abnormality using a convolution. A second method counts coughs for a patient. A third method predicts a respiratory event based on a detected trend. A fourth method forecasts characteristics of a future respiratory event. In a fifth embodiment, a base station is provided for a digital stethoscope.

The invention discloses a pig early disease abnormal behavior recognition system based on attitude angle trend and trajectory analysis of a low-power-consumption Bluetooth ear tag, and the system comprises an attitude angle collection terminal module, a trajectory behavior reconstruction module, a high-frequency sampling trigger module, an edge recognition and preprocessing module, and a fusion judgment and early warning output module. And a cloud platform collaboration module. The attitude angle acquisition terminal is used for acquiring three-axis attitude angle data (a yaw angle Yaw, a pitch angle Pitch and a roll angle Roll) of a pig and reducing energy consumption through an intermittent sampling mechanism; the trajectory reconstruction module constructs a virtual trajectory point string based on Bluetooth signals such as RSSI and the like, and is used for identifying spatial abnormal behavior characteristics; when the system recognizes a low-frequency attitude anomaly or a trajectory disturbance trend, a high-frequency sampling window is automatically triggered, and slight or short-time behavior changes such as local wandering, rotation behaviors and the like are further recognized; and finally, whether the pigs have behavior symptoms such as neurological diseases, abnormal breathing or highly pathogenic infectious diseases or not is judged through a fusion algorithm. The system has the capabilities of high recognition precision, low power consumption, long endurance and large-scale deployment, and can be widely applied to pig farm abnormal behavior early warning and animal health management scenes.

The invention relates to an automatic intervention device for preventing an upper airway from collapsing in sleep. The automatic intervention device comprises a physiological parameter monitoring module, a control module, a pneumatic driving module and a lower jaw advancing execution module which are connected in sequence, according to the device, by monitoring the oxyhemoglobin saturation in real time, the pneumatic driving module is automatically triggered when abnormal breathing is detected, so that an expandable driving body in the execution module is inflated and deformed, the lower jaw of a user is moved forwards instantly and accurately to open the airway, and the device is automatically reset after returning to normal; the problem of temporomandibular joint discomfort caused by continuous forward movement of a traditional oral appliance is solved, meanwhile, pure pneumatic and non-conductive design is adopted, air serves as a working medium, electrical risks near the mouth and the nose are thoroughly eradicated, and safety, comfort and reliability in unattended scenes such as family sleep are remarkably improved.

<pre>The Intelligent Monitoring System is an integrated solution for managing and monitoring chronic respiratory diseases. It combines a patient-facing mobile application with a platform dedicated to healthcare professionals, enabling continuous monitoring and collaborative management. The system collects physiological parameters—such as heart rate, oxygen saturation (SpO₂), and electrocardiogram (ECG) and photoplethysmogram (PPG) signals—using smartwatch sensors. It also records lung sounds via an electronic stethoscope and measures environmental data (air quality, temperature, and humidity) using specialized sensors and a public API. A smart inhaler is integrated to guide patients in its use, thereby boosting treatment adherence and ensuring greater therapeutic efficacy. By leveraging Edge AI, the system predicts key parameters like respiratory rate and analyzes lung sounds directly within the mobile app to detect respiratory abnormalities. Relying on edge computing, it offers real-time processing, reducing cloud dependency while maintaining high accuracy. The system helps patients quickly recognize critical symptoms, monitor vital signs and environmental triggers, and track treatment adherence. Data is shared instantly and securely with healthcare professionals via the platform, facilitating rapid interventions and reducing the need for emergency care. By integrating these features, the system promotes proactive respiratory disease management, prevents exacerbations, and improves patients' quality of life.< / pre>

The application provides a whole-process management method and system for non-contact microwave monitoring of physiological parameters, which comprises the following steps: collecting microwave signals of physiological parameters of a user, pre-processing and feature extraction processing the microwave signals to obtain human vital sign parameters; transmitting the human vital sign parameters to a terminal of the user or a nurse station in a wireless manner to evaluate the human vital sign parameters; comparing the evaluation results with a health rule library to generate corresponding health suggestions and push the health suggestions to the terminal of the user; when the user is in a night sleep state, analyzing the human vital sign parameters in the terminal by using a pre-trained model, judging whether the user is in an abnormal breathing state according to the analysis results; if yes, sending a control instruction to a corresponding sickbed of the user to control the sickbed to perform an angle adjustment operation; and specifically, the whole-process management method of collecting signs-state judgment-linkage intervention is used to realize real-time response and intervention on the health state of the patient.

The utility model relates to the technical field of pillows, and provides an intelligent sleep optimization pillow which comprises a pillow body, placing grooves used for placing arms are formed in the two ends of the pillow body, an arc-shaped supporting groove is formed in the top of the pillow body, a center groove is formed in the center of the top of the pillow body, and a height adjusting piece used for adjusting the height of the pillow body is arranged on the inner side of the pillow body. The air bag is in linkage with the miniature vacuum pump, the breathing frequency sensor and the built-in chip are combined, the breathing state of a user can be monitored in real time, when abnormal breathing is detected, the pillow is automatically inflated to raise the pillow body, the respiratory tract is expanded, snoring or apnea is relieved, and the sleep quality is improved.

The invention relates to the technical field of exercise monitoring, and discloses an exercise injury risk early warning system based on wearable sensor data, and the method comprises a collection module which is used for obtaining the exercise data of an athlete in a preset time period based on a wearable sensor; the heart rate judging module is used for judging whether the heart rate is abnormal or not and obtaining the heart rate abnormal degree; the breathing judgment module is used for calculating the fluctuation degree of the breathing frequency and obtaining the breathing abnormity degree according to the fluctuation degree; the muscle judgment module is used for extracting signal features of the electromyographic signals, judging whether the muscles are tired or not according to the signal features and obtaining the muscle fatigue degree according to the signal features; and the early warning module is used for calculating an athletic injury risk value according to the heart rate abnormality, the breathing abnormality and the muscle fatigue degree, and performing grade early warning according to the athletic injury risk value. According to the invention, comprehensive monitoring of the physiological state of the athlete is realized, and limitation of single index evaluation is avoided.

The invention discloses a non-inductive sleep respiration monitoring method based on a household infrared camera, and belongs to the field of image processing. The method comprises the following steps: in a bedroom scene, taking a household infrared camera as a unique video acquisition device to obtain a video sequence of a monitored person; automatically positioning a chest region of interest ROI through the target detection model; in the ROI, optical flow is combined with principal component analysis to extract the main direction of respiratory movement, and chest wall displacement time sequence signals are generated; then estimating a sleeping posture by using the human body key points, and executing projection compensation and same sleeping posture baseline normalization according to a camera-human body geometrical relationship to obtain a breathing amplitude; and extracting multi-dimensional features such as respiratory rate, respiratory amplitude, inspiration and expiration ratio, rhythmicity and signal quality in a fixed sliding window, inputting the multi-dimensional features into the lightweight time sequence model, and outputting judgment results of normal respiration, hypopnea, apnea and rhythm disorder. The scheme has the advantages of being low in cost, convenient to deploy, non-inductive to users and the like, and is suitable for family early screening and long-term monitoring of abnormal sleep breathing.

The invention relates to the technical field of wearable medical equipment, in particular to an intelligent respiration monitoring device and method based on multi-modal sensing and posture compensation. A thermistor array, a pressure differential sensor and an inertial measurement unit are used for synchronously collecting multi-source signals, a coupling analysis model of breathing signals and human body motion characteristics is constructed, an LSTM network is used for achieving time sequence modeling of a breathing mode, abnormal breathing detection under motion interference is carried out in combination with an isolated forest algorithm, and the detection accuracy is improved. The technical limitations that single-mode sensing is easily interfered by the environment and the false alarm rate of a complex action scene is high in a traditional scheme are broken through. Through multi-modal data fusion and edge computing architecture, real-time abnormal early warning is realized while respiratory rate detection precision is ensured, and a closed-loop processing link from signal acquisition and dynamic compensation to intelligent decision is formed.

The invention discloses a real-time breathing event detection method based on lightweight depth separable convolution and adaptive noise modeling. The method is implemented by the following steps: audio preprocessing and adaptive Me < l > spectrum feature extraction; carrying out adaptive noise modeling and suppression; constructing a lightweight deep separable convolutional network; a double-path attention mechanism; a multi-scale feature fusion strategy; and a real-time processing mechanism. According to the method, by adopting the deep separable convolution and lightweight network design, the calculation efficiency is greatly improved, the deployment requirements of mobile equipment and edge calculation are met, the system can realize smooth real-time detection on resource-limited platforms such as smart phones, tablet personal computers and wearable equipment due to the remarkable efficiency improvement, and the real-time performance of the system is improved. The method has excellent real-time processing capability, supports real-time analysis of continuous audio streams and has important clinical application value, and detection categories comprehensively cover main respiratory abnormality types in clinical diagnosis.

The invention discloses an abnormal breathing event detection method and device, equipment and a storage medium. The method comprises the following steps: acquiring an original vibration signal, wherein the original vibration signal comprises a vibration signal generated by a user and collected by a vibration sensor; preprocessing the original vibration signal to obtain a respiration signal in a required bandwidth range; generating an amplitude time sequence waveform corresponding to the respiratory signal; performing amplitude analysis on the amplitude time sequence waveform, and determining a baseline value and a drop value of the amplitude time sequence waveform; and based on the baseline value and the drop value, determining whether an abnormal breathing event exists in the original vibration signal. The breathing condition of the user in the sleep period can be noninductively detected, and the accuracy of abnormal breathing event detection is improved.

The invention provides an intelligent classification method and device for abnormal breath sound detection, equipment and a medium, and the method comprises the steps: carrying out the preprocessing of an original breath audio signal of each breath cycle, obtaining a one-dimensional time domain signal, and carrying out the length unified processing of the one-dimensional time domain signal; converting the one-dimensional time-domain signals subjected to the unified length processing into a two-dimensional time-frequency spectrogram, performing blank area removal processing on the two-dimensional time-frequency spectrogram, and determining a preprocessed two-dimensional time-frequency spectrogram; inputting the preprocessed two-dimensional time-frequency spectrogram into an anomaly detection model to carry out breath sound anomaly detection processing, and outputting a classification result corresponding to each breath cycle and a time-frequency region of abnormal sound; wherein in the training stage, a patient consistency constraint loss item based on a patient individual feature benchmark is introduced, so that the characterization of a plurality of respiratory cycle samples of the same patient tends to be consistent in a feature space. And the accuracy, the stability and the interpretability of breath sound anomaly detection are remarkably improved by utilizing the anomaly detection model.

A support surface movement device (30) that moves a support surface (520) for supporting the body of a user (U) comprises: a respiratory abnormality detection unit (36, 362) that detects a respiratory abnormality of the user on the support surface; and a movement control unit (37) that controls the movement of the support surface. The support surface has an upper body support part (521) for supporting the upper body of the user on the support surface. The movement control unit moves the upper body support part on the basis of a respiratory abnormality index (ABI) indicating the count of respiratory abnormalities of the user detected by the respiratory abnormality detection unit within a prescribed period.

The application discloses a sleep breathing abnormality non-inductive identification method based on a bed-based mechanical sensor, and belongs to the field of biomedical engineering, and aims to solve the problems of large interference of traditional monitoring equipment and low identification accuracy caused by the difficulty in fine separation of bed-based mechanical coupling signals. The method comprises the following steps: collecting original mechanical signals containing breathing and heartbeat depth coupling characteristics; filtering out interference and removing invalid data through multi-stage filtering and quality assessment; using a generative adversarial network to decouple the coupling signals and restore them into independent breathing and heartbeat waveforms; extracting the pathological feature vectors of the restored waveforms; and inputting the pathological feature vectors into a multi-scale feature fusion model to realize automatic classification of apnea events. The application realizes completely non-inductive non-contact monitoring, improves the separation accuracy of complex coupling signals and the robustness of pathological identification, and has high clinical application value.