CN-122004818-A - Dynamic heart rate variability detection system and method based on MEMS IMU chest strap

CN122004818ACN 122004818 ACN122004818 ACN 122004818ACN-122004818-A

Abstract

The invention discloses a dynamic heart rate variability detection system and method based on an MEMS IMU chest strap, wherein the system comprises a flexible chest strap, an inertia measurement module, a control and transmission module and a power supply module, wherein the flexible chest strap is used for being worn on the chest of a user, the inertia measurement module is used for collecting micro-vibration angular rate GCG signals and real-time step frequencies of the user, the control and transmission module is used for executing a detection algorithm and wirelessly transmitting data, and the power supply module is used for providing a stable power supply, and the inertia measurement module, the control and transmission module and the power supply module are integrated in the flexible chest strap. The invention has high detection accuracy and strong cross-state adaptability.

Inventors

DENG SONGQING
WANG DONGYU
CHEN RUN
ZHENG JU

Assignees

中山大学附属第一医院

Dates

Publication Date: 20260512
Application Date: 20260202

Claims (10)

1. A dynamic heart rate variability detection system based on MEMS IMU chest straps, comprising: a flexible chest strap for wearing on the chest of a user; the inertial measurement module MEMS IMU is used for collecting micro-vibration angular rate GCG signals of a user and real-time step frequency; The control and transmission module is used for executing a detection algorithm and wirelessly transmitting data; The power supply module is used for providing a stable power supply; the inertial measurement module, the control and transmission module and the power supply module are integrated in the flexible chest strap.
2. The dynamic heart rate variability detection system based on the MEMS IMU chest strap according to claim 1, wherein the inertial measurement module comprises a three-axis gyroscope and a three-axis accelerometer, wherein the three-axis gyroscope is used for collecting micro-vibration angular rate GCG signals generated by the chest of a user due to heart beating, the measurement range is 0-4000dps, and the three-axis accelerometer is used for detecting real-time step frequency of the user and assisting in judging the motion state.
3. The MEMS IMU chest strap-based dynamic heart rate variability detection system of claim 2, wherein the control and transmission module process flow comprises: According to signals acquired by the triaxial accelerometer, step frequencies are analyzed and extracted through Fast Fourier Transform (FFT), and the motion state of a user is judged based on the step frequencies; Dynamically selecting an adaptive filtering strategy according to the motion state to carry out filtering processing on the GCG signal, and carrying out secondary peak detection on the filtered GCG signal to extract a heartbeat peak sequence; A heart rate variability HRV indicator is calculated based on the sequence of heart beat peaks.
4. A MEMS IMU chest strap-based dynamic heart rate variability detection system according to claim 3, wherein the adaptive filtering strategy comprises: under all motion states, adopting a band-pass filter (BPF) with the passband frequency of 1Hz to 9Hz to filter the GCG signal; In the running state, a band elimination filter BEF with the stop band frequency of 2.5Hz to 3.5Hz is adopted to filter out the step frequency interference.
5. The MEMS IMU chest strap-based dynamic heart rate variability detection system of claim 4, wherein the motion states comprise a resting state, a jogging state, and a running state.
6. The MEMS IMU chest strap-based dynamic heart rate variability detection system of claim 4, wherein the transfer function of the band pass filter BPF is: ; Wherein, xi b is damping ratio, ωb is central frequency, s is Laplacian variable, s=sigma+jω, σ is real part, j is imaginary unit, ω is angular frequency, which is used to convert the differential/integral relationship of time domain into algebraic relationship of complex frequency domain; is the transfer function of the band pass filter.
7. A MEMS IMU chest strap-based dynamic heart rate variability detection system according to claim 3, wherein the secondary peak detection comprises: primary peak detection, namely extracting a primary peak position from the filtered GCG signal by adopting an adaptive threshold method; and (3) secondary fine detection, namely calculating the interval between adjacent primary peaks, identifying whether a missed detection peak exists or not, and if the interval is larger than a preset average interval, re-detecting and supplementing the peak in the interval to form a final peak sequence.
8. The MEMS IMU chest strap-based dynamic heart rate variability detection system of claim 1, further comprising an auxiliary verification module for synchronously acquiring standard ECG signals and comparing the same with GCG detection results.
9. A detection method for implementing a MEMS IMU chest strap-based dynamic heart rate variability detection system according to any one of claims 1-8, comprising: collecting GCG signals and acceleration signals through an inertia measurement module in the flexible chest strap; based on the acceleration signal, step frequency is extracted through FFT analysis, and the motion state of a user is judged; dynamically selecting an adaptive filtering strategy to filter the GCG signal according to the motion state; Performing secondary peak detection on the filtered GCG signal, and extracting a heartbeat peak sequence; A heart rate variability HRV indicator is calculated based on the sequence of heart beat peaks.
10. The detection method according to claim 9, further comprising sending the HRV indicator to an intelligent terminal for display and storage by wireless transmission.

Description

Dynamic heart rate variability detection system and method based on MEMS IMU chest strap Technical Field The invention relates to the technical field of medical electronics and wearable equipment, in particular to a dynamic heart rate variability detection system and method based on MEMS IMU chest bands. Background Heart rate variability (HEART RATE Variability, HRV) detection techniques are critical to cardiovascular health assessment, autonomic nervous system function monitoring, and motor load management. Two HRV detection techniques are mainly applied clinically, namely traditional Electrocardiographic (ECG) detection and photoplethysmogram (PPG) detection. The traditional electrocardiogram detection records the heart electrical activity through electrodes attached to the body surface, can accurately capture the R peak value, and is a gold standard for HRV calculation. The technical precision is extremely high, reliable time domain, frequency domain and nonlinear HRV indexes can be provided for clinicians, and the method is widely applied to disease diagnosis and health assessment. The PPG detection uses a photosensor in the wearable device to obtain a pulse wave signal by detecting the fluctuation of light absorption rate caused by the change of blood volume under the skin, and calculates the heart rate and HRV according to the pulse wave signal. The PPG technology does not need complex wiring, is convenient to wear, realizes long-term continuous collection of heart rate data with lower cost and user acceptance, and is widely applied to intelligent watches and bracelets. Both HRV detection techniques can provide valuable physiological information, and have important values for stress level assessment, excessive training and early warning, sleep quality analysis and cardiovascular risk screening. Although HRV detection techniques can provide rich autonomic function information, high-precision measurements in dynamic environments are required to meet real-world application requirements. Conventional static or quasi-static detection methods face serious challenges in this respect. The ECG system relies on stable contact of electrodes and skin, signal interruption or artifact increase is easily caused by sweating and displacement in the movement process, and meanwhile, the multi-lead wiring limits the freedom of movement of a user, so that the ECG system is difficult to be suitable for daily dynamic monitoring. Although the PPG technology is convenient to integrate, in active motion states such as walking, running and the like, limb shaking can cause strong motion artifacts, weak pulse wave signals are seriously disturbed, and the signal to noise ratio is sharply reduced. The problem causes the heartbeat cycle extraction error of the existing PPG system under the dynamic condition to be obviously increased, thereby affecting the calculation accuracy of key HRV indexes such as SDNN, RMSSD and the like. More importantly, the traditional method lacks an effective motion interference suppression mechanism, so that the real physiological variability is difficult to distinguish from the pseudo-variation introduced by motion, and the risk of misjudgment of the HRV result is increased. Although motion compensation techniques based on signal processing algorithms (e.g. adaptive filtering, independent component analysis) have been used to improve PPG signal quality, most of these methods correct after the fact, have limited signal recovery capability for severe distortion, are highly dependent on the quality of the reference noise signal, have insufficient generalization capability, and are difficult to cope with complex and variable real motion scenes. In recent years, an Inertial Measurement Unit (IMU) technology has new potential in the field of physiological signal monitoring, and a Ballistocardiogram (BCG) method based on a micro-electromechanical system (MEMS) accelerometer has good performance in heartbeat detection in a resting state, so that a new thought is provided for non-inductive physiological monitoring. However, the current MEMS IMU-based dynamic HRV detection system still has various limitations that firstly, most of the existing researches rely on accelerometers to measure body vibration, signals of the existing researches are easily submerged by macroscopic motion acceleration in dynamic activities, the anti-interference capability is poor, weak heart mechanical activity signals cannot be effectively separated, secondly, the existing systems are difficult to maintain stable HRV detection precision in a severe motion (such as running) state because of lack of an adaptive interference elimination strategy aiming at a specific motion mode, and thirdly, the existing methods have insufficient robustness in a heartbeat peak detection link, and are easy to generate omission detection or false detection when facing to signal morphology variation and periodic disturbance caused by motion, so that heartbeat