AI-powered wearable respiratory monitoring device with multimodal sensing
A wearable device with integrated multimodal sensing and in-device AI processing addresses portability and data reliability issues, providing real-time and secure respiratory monitoring.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Utility models
- Current Assignee / Owner
- マルワ オバイヤ
- Filing Date
- 2026-03-09
- Publication Date
- 2026-05-22
AI Technical Summary
Conventional respiratory monitoring devices are limited by portability, complexity, cost, and reliance on single sensing methods, leading to incomplete data, motion artifacts, and privacy concerns, making them unsuitable for continuous and reliable monitoring outside clinical settings.
A wearable respiratory monitoring device integrating multiple sensing methods (strain and bioimpedance) with in-device processing and motion compensation, and utilizing AI for real-time data analysis and secure wireless communication.
Enables continuous, accurate, and reliable respiratory monitoring with reduced latency and enhanced data privacy, suitable for diverse environments.
Smart Images
Figure 0003255934000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a wearable physiological monitoring device, and more particularly, to a wearable respiratory monitoring device configured to detect and classify respiratory patterns in real time using multimodal sensing and in-device artificial intelligence processing, and relates to a device applicable to clinical applications, home healthcare, and portable medical applications.
Background Art
[0002] The background description includes information that may be useful for understanding the present invention. Any information provided in this specification is not admitted to be prior art or relevant to the present invention. Nor is any specifically or implicitly cited publication admitted to be prior art.
[0003] Respiratory monitoring plays an important role in clinical diagnosis, chronic disease management, sleep assessment, and exercise capacity assessment. By accurately tracking respiratory patterns, it becomes possible to detect conditions such as apnea, tachypnea, and dyspnea at an early stage. However, since conventional hospital devices have limitations in portability, operability, and availability, it is still difficult to continuously monitor respiration outside a controlled environment.
[0004] Conventional respiratory monitoring systems such as spirometers and polysomnography devices are generally large and require a clinical environment under the supervision of medical staff. These systems are not suitable for continuous or portable use. Moreover, their complexity, high cost, and low portability limit daily monitoring in a home healthcare environment and reduce the practicality of long-term physiological assessment.
[0005] Many existing wearable devices rely on a single sensing method, which can result in incomplete or unreliable data. Motion artifacts, posture changes, and environmental noise often distort respiratory signals. Furthermore, reliance on cloud-based processing can lead to latency and privacy concerns, reducing responsiveness in real-time health monitoring and compromising data security.
[0006] Therefore, there is a need for a compact, wearable respiratory monitoring device that can integrate multiple sensing methods and perform in-device processing. Such a system should ideally minimize operational interference, classify respiratory events in real time, and enable secure wireless communication, thereby providing accurate and reliable respiratory assessment under diverse operating conditions. [Brief explanation of the drawing]
[0007] [Figure 1] A perspective view of a wearable respiratory monitoring device 100 configured as a flexible chest band is shown. The device 100 comprises a strain sensor 101, a bioimpedance sensor 102, an output and interface module 103, a motion compensation sensor 104, an embedded processing unit 105, a power management unit 106, and a wireless communication module 107. These components are integrated within the band, enabling compact and self-contained respiratory monitoring. [Modes for carrying out the invention]
[0008] The embodiments shown in the attached drawings will be described in detail below.
[0009] This invention relates to a wearable respiratory monitoring device 100 configured as a flexible band worn on the user's chest. The device 100 enables continuous and real-time respiratory monitoring in clinical, home, and portable settings. The band structure maintains a stable position during various activities and ensures sufficient sensor contact for reliable acquisition of respiratory data even under dynamic physiological conditions.
[0010] The device 100 includes a strain sensor 101 integrated within a flexible band to detect mechanical deformation caused by chest expansion and contraction. The strain sensor 101 generates an electrical signal proportional to respiratory movement. This signal represents respiratory amplitude and respiratory rate, serving as a primary respiratory indicator for further analysis and enabling accurate monitoring both at rest and during activity.
[0011] The bioimpedance sensor 102 is positioned on the band to measure changes in chest impedance that occur during inspiration and expiration. The bioimpedance sensor 102 is configured to be in contact with the skin to ensure stable signal acquisition. Impedance-based measurements complement strain sensor-derived data, and multimodal physiological sensing improves the characterization of the respiratory cycle and enhances overall detection reliability.
[0012] To reduce interference caused by movement, the device 100 is equipped with a motion compensation sensor 104 that detects user movement and changes in posture. The motion compensation sensor 104 generates motion data to distinguish between artifacts caused by physical activity and actual respiratory signals. This improves measurement stability, even when the device is portable or under exercise conditions, and enables accurate respiratory analysis.
[0013] The signals generated from the strain sensor 101, the bioimpedance sensor 102, and the motion compensation sensor 104 are input to the embedded processing unit 105. The embedded processing unit 105 performs preprocessing, including filtering and artifact reduction. This extracts valid respiratory waveforms, improves signal quality before classification processing, and increases the reliability of respiratory event detection.
[0014] The embedded processing unit 105 further includes an in-device artificial intelligence engine that analyzes multimodal data input. Extracted parameters such as frequency, amplitude, and variability are evaluated to classify the respiratory state. Real-time processing enables rapid detection of anomaly patterns without relying on cloud computing, resulting in reduced latency and improved data privacy.
[0015] The power management unit 106 is electrically connected to control the energy supply to each functional component within the device 100. The power management unit 106 ensures stable operation during long-term monitoring, optimizes energy consumption, and extends the operating time of the device while maintaining continuous sensing, processing, and communication functions.
[0016] The output and interface module 103 provides alerts to the user when abnormal respiratory events are detected. The wireless communication module 107 transmits processed respiratory data to an external device via Bluetooth communication. This configuration enables real-time notifications to the user and secure data synchronization, supporting remote monitoring and clinical evaluation as needed. The embodiments described above are illustrative of the present invention and do not limit it. Various modifications, equivalent substitutions, and improvements within the scope of the principle of the present invention are included within the scope of the present invention.
Claims
1. A wearable respiratory monitoring device 100 using multimodal sensing, A strain sensor 101 generates electrical signals corresponding to chest expansion and contraction during respiration, A bioimpedance sensor 102 detects impedance changes related to inhalation and exhalation, A motion compensation sensor 104 generates motion data indicating the user's movements, An embedded processing unit 105 connected to the strain sensor 101, the bioimpedance sensor 102, and the motion compensation sensor 104, A power management unit 106 controls the power supplied to the device 100, A wireless communication module 107 transmits respiratory data to an external device via Bluetooth communication. Equipped with, A wearable respiratory monitoring device characterized in that the embedded processing unit 105 integrates multimodal sensor inputs to establish a respiratory reference state and detects abnormal respiratory events in real time.
2. The apparatus according to claim 1, wherein the flexible wearable band is formed of an elastic fiber material consisting of natural or synthetic fibers.
3. The apparatus according to claim 1, wherein the strain sensor 101 includes a piezoresistive element whose resistance value changes in accordance with the mechanical deformation of the band.
4. The apparatus according to claim 1, wherein the bioimpedance sensor 102 includes at least one pair of skin-contact electrodes for measuring changes in chest impedance.
5. The apparatus according to claim 1, wherein the motion compensation sensor 104 includes a triaxial accelerometer that detects movement in three orthogonal axes.
6. The apparatus according to claim 1, wherein the embedded processing unit 105 performs adaptive filtering processing to suppress motion artifacts using data from the motion compensation sensor 104.
7. The apparatus according to claim 1, wherein the embedded processing unit 105 uses a neural network classification model to identify the respiratory state.
8. The apparatus according to claim 1, wherein the power management unit 106 dynamically adjusts energy consumption in order to extend the operating time.
9. The apparatus according to claim 1, wherein the wireless communication module 107 transmits respiratory parameters to a mobile device using the Bluetooth Low Energy protocol.