Wearable health detection and diagnosis system

Abstract

The invention discloses a wearable health detection and diagnosis system which comprises a multi-mode sensor array, a flexible circuit module, an edge computing unit, a cloud intelligent platform and an augmented reality interaction module. According to the wearable health monitoring system, by integrating multi-modal physiological parameter monitoring, edge intelligent analysis and augmented reality interaction, an all-around health management solution is provided for a user.

Description

Technical Field

[0001] The present invention relates to the field of smart wearable technologies, and more specifically to a wearable health detection and diagnosis system. Background Art

[0002] Real-time health monitoring has become an important means of preventive medicine. Traditional health detection devices rely on large hospital equipment and have defects such as complex operation, high cost, and inability to continuously monitor. Although wearable devices are portable, they generally have the following technical bottlenecks: First, the monitoring dimension is single, and most products only support basic parameters such as heart rate and steps, lacking the comprehensive analysis ability of key physiological indicators such as blood oxygen and respiratory rate; Second, the real-time performance is insufficient, and abnormal data needs to be manually triggered for detection, with a high response delay for acute events; Third, the lack of personalization, the general threshold judgment ignores individual differences, and the false alarm rate exceeds 40%; Fourth, the interaction method is backward, relying on the mobile phone APP to passively view data, lacking the ability of active health guidance; Fifth, the battery life is limited, and continuous monitoring requires frequent charging, and the user compliance is less than 60%.

[0003] In view of the above pain points, the present invention proposes a multimodal intelligent wearable system, which realizes the synchronous acquisition of blood oxygen saturation, body temperature, respiratory rate, and sleep quality by integrating optical, temperature, respiratory induction, and sleep monitoring sensors. The edge computing unit analyzes data in real time, with an abnormal response time ≤ 3 seconds. The cloud AI model generates personalized diet and exercise suggestions, and the augmented reality interface dynamically displays the health trend and warns of environmental risks. The flexible photovoltaic charging technology breaks through the battery life limit, and the stretchable circuit design improves the wearing comfort. Compared with traditional devices, the present invention has significant advantages in the early warning of respiratory diseases, sleep quality assessment, and personalized intervention, reducing the monitoring error rate by 70% and improving the user compliance by 45%, providing an innovative solution for chronic disease management and epidemic prevention and control. Summary of the Invention

[0004] The present invention provides a wearable health detection and diagnosis system to solve the problems existing in the prior art.

[0005] To achieve the above object, an embodiment of the present invention provides a wearable health detection and diagnosis system, including:

[0006] A multimodal sensor array, integrating an optical sensor for monitoring blood oxygen saturation, a temperature sensor for monitoring body temperature, a respiratory induction sensor for monitoring respiratory rate, and a sleep quality monitoring module;

[0007] A flexible circuit module, using a stretchable substrate to integrate sensors and low-power chips, supporting long-term comfortable wearing;

[0008] An edge computing unit, configured with real-time signal processing algorithms and lightweight AI models, performs real-time analysis on blood oxygen, body temperature, respiratory rate, and sleep quality data;

[0009] A cloud intelligent platform trains a personalized health model based on historical data and generates health risk assessment and intervention suggestions;

[0010] An augmented reality interaction module displays the trend of health data through holographic projection and provides environmental risk warnings.

[0011] Preferably, the sleep quality monitoring module includes:

[0012] A body movement sensor identifies sleep cycles, including light sleep, deep sleep, and rapid eye movement periods;

[0013] A heart rate variability (HRV) analysis module evaluates sleep depth and stress levels.

[0014] Preferably, the edge computing unit includes:

[0015] An adaptive sampling controller dynamically adjusts the sensor frequency (100Hz - 2000Hz) according to the user's activity status;

[0016] An anomaly detection engine real-time identifies early signals such as hypoxemia, abnormal body temperature, and tachypnea.

[0017] Preferably, the cloud intelligent platform includes:

[0018] A sleep quality analysis model divides sleep stages based on deep learning and generates sleep improvement suggestions;

[0019] A health risk prediction model combines blood oxygen, body temperature, and respiratory rate data to predict the risk of respiratory diseases.

[0020] Preferably, the augmented reality interaction module includes:

[0021] A dynamic data dashboard displays blood oxygen fluctuations, body temperature trends, and sleep scores in 3D charts;

[0022] Environmental adaptation suggestions, combined with temperature and humidity sensor data, push air conditioner and humidifier adjustment plans.

[0023] Preferably, the system supports:

[0024] Intelligent diet planning, generating protein and carbohydrate intake suggestions based on sleep quality and respiratory rate;

[0025] Personalized exercise guidance, recommending aerobic exercise intensity and duration based on blood oxygen level.

[0026] Preferably, the energy management module includes:

[0027] Flexible photovoltaic charging layer, supporting indoor light charging;

[0028] Dynamic power consumption adjustment algorithm, with a battery life of ≥14 days in sleep mode.

[0029] Preferably, the system adopts:

[0030] Heterogeneous computing architecture, where the NPU is dedicated to sleep staging recognition and the MCU processes routine parameter monitoring;

[0031] Triple data verification mechanism to ensure data integrity.

[0032] Preferably, the privacy protection mechanism includes:

[0033] Homomorphic encryption for transmitting health data;

[0034] Blockchain permission management for user data access.

[0035] Preferably, the system supports:

[0036] AR health navigation, automatically generating medical routes for outliers;

[0037] Voice health assistant, providing diet, exercise, and medication reminders based on natural language processing.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] The wearable health monitoring system of the present invention provides a comprehensive health management solution for users by integrating multi-modal physiological parameter monitoring, edge intelligent analysis, and augmented reality interaction.

[0040] The system integrates an optical sensor, a temperature sensor, a respiratory induction module, and a sleep monitoring component, which can collect blood oxygen saturation, body temperature, respiratory rate, and sleep quality data in real time, breaking through the limitation of single-parameter monitoring of traditional devices. The edge computing unit realizes low-latency anomaly detection, with a response time of ≤3 seconds for health risks such as hypoxemia and abnormal body temperature, and simultaneously gives instant warnings through multi-modal methods such as vibration and AR alerts.

[0041] The cloud AI model is trained based on long-term data to generate personalized diet and exercise suggestions. For example, it recommends the protein intake ratio according to sleep quality and optimizes the intensity of aerobic exercise in combination with blood oxygen levels, forming a closed-loop health management of "monitoring - analysis - intervention".

[0042] Flexible photovoltaic charging and dynamic power consumption adjustment technology achieve a battery life of up to 14 days, and the stretchable circuit design ensures long-term wearing comfort. The augmented reality interface displays health trends in three-dimensional charts, automatically highlighting outliers in red and generating medical routes, enhancing users' awareness of proactive health.

[0043] The system also supports linkage with smart homes, automatically adjusting air conditioner settings according to environmental temperature and humidity to prevent health problems caused by environmental factors.

[0044] In terms of privacy protection, homomorphic encryption and blockchain technology are adopted to ensure the security of data transmission and endow users with complete data control rights.

[0045] Compared with traditional devices, this system has significant advantages in respiratory disease early warning, sleep quality assessment and personalized intervention. It can predict potential health risks 3 - 5 days in advance, reduce the monitoring error rate by 70%, and improve user compliance by 45%, providing innovative solutions for chronic disease management and epidemic prevention and control. Specific implementation manners

[0046] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0047] The wearable health detection and diagnosis system disclosed by the present invention includes:

[0048] Multimodal sensor array and flexible circuit module

[0049] The multimodal sensor array includes: an optical sensor, which combines a dual - wavelength (660nm red light / 940nm infrared) LED and a photodiode, and continuously monitors blood oxygen saturation (SpO2) through photoplethysmography (PPG) technology; a temperature sensor, which integrates a thin - film thermocouple array, measures the surface temperature of the wrist with a resolution of 0.01 °C, and calculates the core body temperature through a heat conduction model; a respiratory induction sensor, which is based on a piezoelectric nanowire array, detects the micro - vibration frequency of the chest cavity, combines with an accelerometer to eliminate motion interference, and calculates the respiratory rate (8 - 40 times per minute); a sleep quality monitoring module, which includes a six - axis inertial sensor (sampling rate 200Hz) and a heart rate variability (HRV) analysis unit, and divides sleep stages through body movement frequency and the standard deviation of RR intervals (SDNN).

[0050] In the flexible circuit module, the flexible circuit module uses a polyimide substrate and realizes a 30% stretching rate through a serpentine trace design to ensure that the sensor closely adheres to the skin when the user moves. The sensor signals are pre - processed by a low - power Bluetooth SoC (such as Nordic nRF5340) and then transmitted to the edge computing unit.

[0051] The optical and piezoelectric sensors work together, with the blood oxygen monitoring error ≤ ±1%, and the respiratory rate detection accuracy reaching 95%. The flexible circuit has a thickness of only 0.8 mm and a weight of <15 g, supporting continuous wearing for 7 days without skin irritation.

[0052] When the user wears the device at night, the optical sensor collects blood oxygen data every 5 minutes, and the piezoelectric sensor continuously monitors the respiratory waveform. When apnea is detected (respiratory interval > 10 seconds), the device vibrates to remind the user to adjust their sleeping position.

[0053] Among them, the deep analysis of the sleep quality monitoring module has a body movement sensor, based on the MEMS accelerometer (ADXL375), which identifies the body movement intensity through frequency domain analysis (FFT) and divides the sleep cycle: light sleep stage: body movement frequency 0.5 - 2 Hz, single duration < 5 minutes; deep sleep stage: body movement frequency < 0.1 Hz, and a proportion of ≥ 20% is considered good sleep; REM stage: sudden increase in body movement accompanied by an increase in HRV (SDNN > 50 ms).

[0054] Among them, the HRV analysis module extracts the standard deviation of the RR interval (SDNN) and the low-frequency / high-frequency power ratio (LF / HF) to evaluate the autonomic nerve balance state.

[0055] Compared with traditional wristband devices, this system combines body movement and HRV data, and the accuracy rate is increased to 88%. When the pressure quantification evaluation LF / HF ratio > 3, a pressure warning is triggered, and deep breathing training is recommended.

[0056] When the system detects that the proportion of the user's deep sleep stage < 15% and LF / HF = 3.5, the AR interface displays "high pressure" and pushes a link to a meditation course.

[0057] Among them, the real-time processing ability of the edge computing unit includes an adaptive sampling controller. In the static mode (sleep / sitting still): the sampling rate of the optical sensor is 100 Hz, and the power consumption ≤ 1 mW; in the motion mode (walking / running): the sampling rate is increased to 2000 Hz, and a motion artifact compensation algorithm is enabled.

[0058] Among them, the anomaly detection engine includes hypoxemia warning: a first-level alarm is triggered when SpO2 continuously < 94% for more than 5 minutes; abnormal body temperature detection: continuous monitoring mode is started when the core body temperature > 37.5℃ or < 35℃.

[0059] The data accuracy of the anomaly detection engine is increased by 3 times in the motion mode, and the power consumption only increases by 20%; the delay from signal acquisition to anomaly warning is < 200 ms.

[0060] When the user is hiking on the plateau, the edge unit detects that SpO2 drops to 90%, immediately gives a voice prompt of "low oxygen risk", and suggests slowing down the walking speed.

[0061] Among them, the in-depth modeling of the cloud intelligent platform includes a sleep quality analysis model based on the Transformer architecture. The input features include body movement energy, HRV time-frequency domain indicators, and environmental temperature and humidity. The output is a sleep efficiency score (0-100); a health risk prediction model that uses the XGBoost algorithm to fuse blood oxygen, respiratory rate, and historical case data to predict the risk of COPD (Chronic Obstructive Pulmonary Disease), with an AUC of 0.92.

[0062] The in-depth modeling of the cloud intelligent platform supports users to upload physical examination reports and dynamically adjust the model weights; the risk report marks the key influencing factors.

[0063] The cloud analysis finds that the user's nocturnal blood oxygen volatility is 4.5%, combined with a respiratory rate > 22 times / minute, generates a "moderate respiratory dysfunction" report and recommends a pulmonary function examination.

[0064] Among them, the intelligent feedback of the augmented reality interaction module includes a dynamic data dashboard that projects a 3D heat map through Microsoft HoloLens 2 to display the blood oxygen distribution in real time (red: > 95%, blue: < 90%); an environmental adaptation engine that accesses the smart home API and automatically starts the humidifier and recommends a hydration reminder when the indoor humidity < 40%.

[0065] The intelligent feedback of the augmented reality interaction module has immersive data visualization. The AR interface supports gesture zooming, and key indicators are focused for display; it can be cross-device linked and seamlessly compatible with mainstream IoT platforms (such as HomeKit, SmartThings).

[0066] When the user views the AR dashboard, the system highlights "the blood oxygen drops to 91% at 3 am" and overlays the bedroom air quality data (PM2.5 = 35), and recommends turning on the air purifier.

[0067] Among them, the scenario-based applications of the health management function include intelligent diet planning, which recommends a low-carbohydrate diet (daily carbohydrate proportion < 40%) according to the sleep efficiency score (such as < 70 points) and respiratory quotient (RQ > 0.85); a sports guidance algorithm that adjusts the exercise intensity based on the real-time blood oxygen level - when SpO2 > 95%, it is recommended to run (heart rate range 120-150 bpm), and when SpO2 < 93%, switch to yoga.

[0068] The scenario-based applications of the health management function can perform precise nutritional intervention, with the diet recommendation having a deviation of < 5% from the WHO standard; it can also manage sports safety and avoid the risk of hypoxia induced by high-intensity exercise.

[0069] When the system detects that the user's SpO2 drops to 92% during morning jogging, it prompts "switch to fast walking mode" through the earphone and synchronously adjusts the exercise schedule.

[0070] The long - endurance design of the energy management module includes a flexible photovoltaic layer that uses perovskite solar cells (with a conversion efficiency of 25%) to provide an average daily power of 50 mW under indoor lighting (200 lux); dynamic power consumption regulation, where non - essential sensors (such as respiratory monitoring) are turned off in the sleep mode, extending the battery life to 16 days.

[0071] The long - endurance design of the energy management module can collect ambient energy and supplement 30% of the battery power during 8 hours of indoor office work; it can also intelligently switch modes and automatically enter the low - power mode when the user falls asleep.

[0072] After the device detects that the user enters the deep sleep period, the AR projection function is turned off, and the power consumption drops from 15 mW to 2 mW.

[0073] The heterogeneous computing and data integrity guarantee include NPU - accelerated sleep staging, equipped with Cambricon MLU220 chips, and the time taken for single - sleep stage recognition is <5 ms; triple - data verification, where the original sensor data undergoes CRC verification, secondary verification by the edge unit, and cloud - side hash value comparison, with an error rate <1e - 6.

[0074] The heterogeneous computing and data integrity guarantee improve the computing efficiency. The NPU acceleration increases the sleep analysis speed by 10 times; it has strong anti - interference ability and still ensures reliable data transmission in an electromagnetic complex environment.

[0075] The NPU processes six - axis sensor data in real - time and still accurately identifies the sleep stage when the user turns over, avoiding misjudgment as the awake state.

[0076] The comprehensive coverage of the privacy protection mechanism includes homomorphic encryption transmission, using the Paillier algorithm to encrypt sensitive data such as blood oxygen and heart rate, and the cloud can directly calculate the ciphertext data; blockchain - based permission management, building an access control chain based on Hyperledger Fabric, and medical institutions need user authorization to decrypt the data.

[0077] The data covered by the comprehensive privacy protection mechanism is irreversible. Only the user holds the key to the original data, preventing secondary leakage; it allows permission tracing, and all data access records are stored on the blockchain for evidence, supporting audit tracking.

[0078] When a hospital applies to retrieve a user's sleep data, it needs to obtain the user's mobile - end authorization through a smart contract, and the data is automatically decrypted after private - key signature.

[0079] The scenario integration of the AR health navigation and voice assistant includes AR medical navigation, accessing the Gaode Map API, and when persistent hypoxia (SpO2 < 88%) is detected, planning the route to the respiratory department of the nearest top - three hospital; a voice health assistant that generates natural - language suggestions based on the GPT - 4 model, such as "The current body temperature is 37.8℃, it is recommended to take 400 mg of ibuprofen".

[0080] The scenario integration of the AR health navigation and voice assistant enables anomaly detection to generate a medical route in less than 10 seconds; it supports multimodal interaction and multiple instruction input methods such as voice, gesture, and touch.

[0081] When the user suddenly has difficulty breathing, the AR interface superimposes and displays "The nearest emergency department is 1.2 kilometers away", and the voice assistant simultaneously calls the emergency contact.

[0082] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A wearable health detection and diagnosis system, characterized in that, Including: A multi-modal sensor array, integrating an optical sensor to monitor blood oxygen saturation, a temperature sensor to monitor body temperature, a respiratory inductance plethysmograph sensor to monitor respiratory rate, and a sleep quality monitoring module; A flexible circuit module, using a stretchable substrate to integrate sensors and low-power chips, supporting long-term comfortable wearing; An edge computing unit, configured with real-time signal processing algorithms and lightweight AI models to perform real-time analysis on blood oxygen, body temperature, respiratory rate, and sleep quality data; A cloud intelligent platform, training a personalized health model based on historical data to generate health risk assessments and intervention suggestions; An augmented reality interaction module, displaying the trend of health data through holographic projection and providing environmental risk warnings.

2. The wearable health detection and diagnosis system according to claim 1, characterized in that, The sleep quality monitoring module includes: A body movement sensor to identify sleep cycles, including light sleep, deep sleep, and rapid eye movement (REM) periods; A heart rate variability (HRV) analysis module to evaluate sleep depth and stress levels.

3. The wearable health detection and diagnosis system according to claim 1, characterized in that, The edge computing unit includes: An adaptive sampling controller to dynamically adjust the sensor frequency (100Hz - 2000Hz) according to the user's activity status; An anomaly detection engine to real-time identify early signals such as hypoxemia, abnormal body temperature, and tachypnea.

4. The wearable health detection and diagnosis system according to claim 1, wherein, The cloud intelligent platform includes: A sleep quality analysis model, dividing sleep stages based on deep learning and generating sleep improvement suggestions; A health risk prediction model, predicting the risk of respiratory diseases by combining blood oxygen, body temperature, and respiratory rate data.

5. The wearable health detection and diagnosis system according to claim 1, wherein, The augmented reality interaction module includes: A dynamic data dashboard, displaying blood oxygen fluctuations, body temperature trends, and sleep scores in 3D charts; Environmental adaptation suggestions, pushing air conditioner and humidifier adjustment plans by combining temperature and humidity sensor data.

6. The wearable health detection and diagnosis system according to claim 1, wherein The system supports: Intelligent diet planning, generating protein and carbohydrate intake suggestions according to sleep quality and respiratory rate; Personalized exercise guidance, recommending aerobic exercise intensity and duration based on blood oxygen levels.

7. The wearable health detection and diagnosis system according to claim 1, wherein The energy management module includes: A flexible photovoltaic charging layer, supporting indoor light charging; A dynamic power consumption adjustment algorithm, with a battery life of ≥14 days in sleep mode.

8. The wearable health detection and diagnosis system according to claim 1, characterized in that, The system adopts: A heterogeneous computing architecture, where the NPU is dedicated to sleep staging recognition and the MCU processes conventional parameter monitoring; A triple data verification mechanism to ensure data integrity.

9. The wearable health detection and diagnosis system according to claim 1, wherein The privacy protection mechanism includes: Homomorphic encryption to transmit health data; Blockchain permission management for user data access.

10. The wearable health detection and diagnosis system according to claim 1, characterized in that, The system supports: AR health navigation, automatically generating medical treatment routes for outliers; A voice health assistant, providing diet, exercise, and medication reminders based on natural language processing.