Dynamic heart rate variability detection system and method based on MEMS IMU pectoral girdle

Abstract

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

Description

Technical Field

[0001] This invention relates to the field of medical electronics and wearable device technology, and in particular to a dynamic heart rate variability detection system and method based on a MEMS IMU chest strap. Background Technology

[0002] Heart rate variability (HRV) detection technology is crucial for cardiovascular health assessment, autonomic nervous system function monitoring, and exercise load management. Clinically, two main HRV detection technologies are used: traditional electrocardiogram (ECG) and photoplethysmography (PPG). Traditional ECG records cardiac electrical activity using electrodes attached to the body surface, accurately capturing the R wave peak, and is considered the "gold standard" for HRV calculation. This technology offers extremely high accuracy, providing clinicians with reliable time-domain, frequency-domain, and nonlinear HRV indicators, and is widely used in disease diagnosis and health assessment. PPG detection utilizes light sensors in wearable devices to acquire pulse wave signals by detecting fluctuations in light absorption rate caused by changes in blood volume under the skin, and then calculates heart rate and HRV accordingly. PPG technology requires no complex wiring, is easy to wear, and achieves long-term continuous acquisition of heart rate data at a lower cost and with greater user acceptance, making it widely used in smartwatches and fitness trackers. Both HRV detection technologies provide valuable physiological information and are important for stress level assessment, overtraining warnings, sleep quality analysis, and cardiovascular risk screening.

[0003] While HRV detection technology provides rich information on autonomic nervous system function, high-precision measurement in dynamic environments is necessary to meet real-world application requirements. Traditional static or quasi-static detection methods face significant challenges in this regard. ECG systems rely on stable electrode-skin contact, which is susceptible to signal interruption or increased artifacts during movement due to sweating or displacement. Furthermore, multi-lead wiring restricts user freedom of movement, making it unsuitable for routine dynamic monitoring. While PPG technology is easy to integrate, limb movement during active movements such as walking and running can cause strong motion artifacts, severely interfering with weak pulse wave signals and leading to a sharp drop in signal-to-noise ratio. This issue significantly increases the error in extracting the heart rate cycle under dynamic conditions in existing PPG systems, thus affecting the accuracy of calculating key HRV indicators such as SDNN and RMSSD. More critically, traditional methods lack effective motion interference suppression mechanisms, making it difficult to distinguish between genuine physiological variability and motion-induced spurious changes, increasing the risk of misinterpretation of HRV results. Although motion compensation techniques based on signal processing algorithms (such as adaptive filtering and independent component analysis) have been used to improve PPG signal quality, these methods are mostly post-correction methods with limited ability to recover signals with severe distortion. They are also highly dependent on the quality of the reference noise signal, have insufficient generalization ability, and are difficult to cope with complex and ever-changing real motion scenarios.

[0004] In recent years, inertial measurement unit (IMU) technology has shown new potential in the field of physiological signal monitoring. The cardiac impact mapping (BCG) method based on microelectromechanical systems (MEMS) accelerometers has demonstrated good performance in resting heartbeat detection, providing a new approach for non-invasive physiological monitoring. However, current dynamic HRV detection systems based on MEMS IMUs still have several limitations: First, most existing studies rely on accelerometer measurements of body vibrations, whose signals are easily submerged by macroscopic motion acceleration during dynamic activities, resulting in poor anti-interference capabilities and an inability to effectively separate weak cardiac mechanical activity signals. Second, existing systems struggle to maintain stable HRV detection accuracy during strenuous exercise (such as running) due to a lack of adaptive interference cancellation strategies for specific motion patterns. Third, existing methods lack robustness in detecting peak heartbeats, easily leading to missed or false detections when faced with motion-induced signal morphology variations and periodic perturbations, resulting in incomplete heartbeat sequences and severely impacting the reliability of subsequent HRV analysis.

[0005] ECG is the "gold standard" for clinical HRV detection. Its technical principle is as follows: conductive electrodes are attached to specific parts of the body (such as the chest, arms, and legs) to collect the bioelectrical signals generated during the heart's systole / diastole. After signal amplification and filtering (such as removing 50Hz power frequency interference), an ECG waveform containing the P wave, QRS complex, and T wave is formed. The heartbeat time is located based on the R wave in the QRS complex, the adjacent RR interval is calculated, and then the HRV index is obtained through time domain (such as SDNN, RMSSD) and frequency domain (such as LF, HF) analysis.

[0006] Traditional electrocardiogram (ECG) technology faces multiple bottlenecks in dynamic heart rate variability (HRV) monitoring. First, it has poor adaptability to various exercise scenarios. Sweating can easily cause electrode short circuits or signal crosstalk between adjacent electrodes, severely compromising the integrity of the ECG signal. Even with the use of textile or flexible electrodes to improve comfort, their fabrication requires complex integration of multiple layers of conductive materials and circuitry, resulting in high manufacturing difficulty and cost. Furthermore, it remains difficult to overcome electrode displacement and signal baseline drift caused by body movement during exercise, and motion artifacts remain significant. Second, operation and wear are limited. Accurate electrode placement and device adjustment usually rely on professionals, making it unsuitable for home use. Electrodes must fit tightly against the skin, and prolonged wear can easily cause skin allergies. The pulling sensation during exercise also severely impacts wearing comfort. More importantly, its dynamic detection capability is severely lacking. During high-intensity exercise such as running, ECG signal quality drops sharply, and the accuracy of R-wave localization decreases significantly (false negative rate exceeds 30%), making it impossible to reliably extract heartbeat timing and rendering HRV calculations ineffective. Therefore, this technology is basically limited to applications in resting states and cannot meet the needs of daily dynamic health monitoring.

[0007] PPG (Photodiode Photodetector) detects heart rate based on the difference in light absorption between blood and tissues such as muscles and bones. Its technical principle involves integrating a photodiode (emitter) and a photosensor (receiver) into a wearable device (such as a wristband or watch), which is attached to the skin surface of the wrist or fingers. The emitter emits specific wavelengths (such as 660nm red light or 940nm infrared light), and the receiver detects changes in light intensity after reflection / transmission through the skin. The heartbeat causes periodic fluctuations in blood volume, which in turn triggers periodic changes in light intensity. This change is converted into an electrical signal (pulse wave), and the heart rate is obtained by analyzing the pulse wave period. The core challenges currently facing wearable PPG technology are sensitivity to motion artifacts and limitations in applicable scenarios. During exercise, wrist movement causes relative displacement between the sensor and the skin, resulting in drastic drift in the light intensity signal. The amplitude of the motion artifacts can be 5 to 10 times that of the real heartbeat signal, severely interfering with signal quality. To suppress such interference, existing methods rely on complex frame models for frequency domain denoising. This not only results in complex algorithms and poor real-time performance, but also only estimates the average heart rate, failing to accurately identify the temporal location of each heartbeat, thus hindering accurate calculation of heart rate variability (HRV). Furthermore, interference from ambient light (such as sunlight and artificial light) forces devices to incorporate additional light-shielding structures, while skin oils and sweat alter local light reflection characteristics, further degrading signal quality. Consequently, the accuracy of HRV detection in resting states is only 75%–85%. Due to these factors, PPG technology is primarily limited to resting or low-dynamic scenarios, struggling to operate stably during moderate to high-intensity exercises such as jogging and running. Additionally, the sensor requires close contact with the skin, and prolonged wear can lead to poor contact and signal failure due to sweat accumulation, severely limiting its widespread application in dynamic health monitoring.

[0008] In summary, for dynamic HRV detection based on wearable devices, the core challenge of current technological development is to develop a comprehensive system that can utilize new sensing modalities with strong anti-interference capabilities, adaptively suppress interference based on motion state, and achieve high-integrity heartbeat sequence extraction. Summary of the Invention

[0009] To address the technical problems existing in the prior art, this invention proposes a dynamic heart rate variability detection system and method based on a MEMS IMU chest strap. By combining an inertial measurement module, adaptive filtering is used to suppress motion / respiratory interference, and secondary peak detection is used to avoid missed detection, thereby achieving accurate and stable detection of HRV under multiple states.

[0010] On the one hand, to achieve the above objectives, the present invention provides a dynamic heart rate variability detection system based on a MEMS IMU chest strap, comprising: Flexible chest strap: for wearing on the user's chest; Inertial Measurement Unit (MEMS IMU): Used to acquire the user's micro-vibration angular rate (GCG) signal and real-time step frequency; Control and transmission module: used to execute detection algorithms and wirelessly transmit data; Power supply module: Used to provide a stable power supply; The inertial measurement module, the control and transmission module, and the power supply module are all integrated into the flexible chest strap.

[0011] Preferably, the inertial measurement module includes a three-axis gyroscope and a three-axis accelerometer. The three-axis gyroscope is used to collect the micro-vibration angular rate (GCG) signal of the user's chest caused by the heartbeat, with a measurement range of 0-4000 dps. The three-axis accelerometer is used to detect the user's real-time step frequency and assist in judging the motion state.

[0012] Preferably, the processing flow of the control and transmission module includes: Based on the signals collected by the triaxial accelerometer, the step frequency is extracted by Fast Fourier Transform (FFT) analysis, and the user's motion state is determined based on the step frequency. Based on the motion state, an adaptive filtering strategy is dynamically selected to filter the GCG signal, and a secondary peak detection is performed on the filtered GCG signal to extract the heartbeat peak sequence. Heart rate variability (HRV) is calculated based on the aforementioned peak heart rate sequence.

[0013] Preferably, the adaptive filtering strategy includes: In all motion states, the GCG signal is filtered using a bandpass filter (BPF) with a passband frequency of 1 Hz to 9 Hz; During running, a band-stop filter (BEF) with a stopband frequency of 2.5Hz to 3.5Hz is used to filter out step frequency interference.

[0014] Preferably, the motion state includes a resting state, a jogging state, and a running state.

[0015] Preferably, the transfer function of the bandpass filter (BPF) is: ; In the formula, ξb The damping ratio; ωb The center frequency; s For Laplace variables, in the form of s = σ + jω , σ For the actual part, j For imaginary units, ω ω is the angular frequency, used to transform the differential / integral relationship in the time domain into an algebraic relationship in the complex frequency domain; Let be the transfer function of the bandpass filter.

[0016] Preferably, the secondary peak detection includes: First peak detection: The initial peak position is extracted from the filtered GCG signal using an adaptive threshold method; Secondary fine detection: Calculate the interval between adjacent preliminary peaks to identify whether there are any missed peaks. If the interval is greater than the preset average interval, re-detect and supplement peaks within the interval to form the final peak sequence.

[0017] Preferably, the system further includes an auxiliary verification module, which is used to synchronously acquire standard ECG signals and compare and verify them with GCG detection results.

[0018] On the other hand, to achieve the above objectives, the present invention also provides a detection method for implementing the aforementioned MEMS IMU chest strap-based dynamic heart rate variability detection system, comprising: GCG and acceleration signals are acquired through an inertial measurement module in a flexible chest strap; Based on the acceleration signal, step frequency is extracted through FFT analysis to determine the user's motion state; The GCG signal is filtered by dynamically selecting an adaptive filtering strategy based on the motion state. Perform secondary peak detection on the filtered GCG signal to extract the heartbeat peak sequence; Heart rate variability (HRV) is calculated based on the aforementioned peak heart rate sequence.

[0019] Preferably, the method further includes transmitting the HRV index to a smart terminal for display and storage via wireless transmission.

[0020] Compared with the prior art, the present invention has the following advantages and technical effects: (1) High detection accuracy and strong cross-state adaptability: Resting state: Heart rate detection accuracy >95%, HRV index average accuracy ≈92% (with ECG as the gold standard); Jogging state: The average accuracy of HRV index is ≈89%, with no obvious missed detections; Running mode: It can stably detect heart rate and HRV, solving the problem that existing technologies (ECG / PPG) cannot detect them under high-intensity exercise.

[0021] (2) Comfortable to wear and easy to operate: The flexible and elastic fabric carrier does not need to be in close contact with the skin (unlike existing ECG / PPG, which require skin contact), and there is no discomfort even after wearing it for a long time; Automatic sleep / wake-up function, no manual on / off required, results can be viewed by connecting to a mobile phone via Bluetooth, suitable for home and sports scenarios; (3) Low power consumption and good portability: The hardware is compact (PCB size 15mm×20mm), and the chest strap weighs less than 50g, making it comfortable to wear. Low power consumption design (sleep current <10μA, operating current <5mA), battery life ≥8 hours on a single charge, meeting all-day monitoring needs; (4) Strong anti-interference ability and good environmental adaptability: The sealed and waterproof design is unaffected by sweat and rain (existing ECG is affected by sweat, and PPG is affected by light reflection from sweat); the adaptive filtering strategy specifically filters out motion (step frequency) and breathing interference (breathing frequency 0.3-0.6Hz, which is filtered out by BPF), improving the signal-to-noise ratio (SNR). Attached Figure Description

[0022] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of a test platform with a chest strap and an electrocardiogram acquisition circuit according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the original signal under jogging conditions and the spectral density FFT of the original signal under jogging conditions according to an embodiment of the present invention. Figure 3 This is a schematic diagram of the filtered signal during jogging and the spectral density FFT of the filtered signal during jogging, according to an embodiment of the present invention. Figure 4 This is a schematic diagram of the original signal under running conditions and the spectral density FFT of the original signal under running conditions according to an embodiment of the present invention. Figure 5 This is a schematic diagram of the filtered signal during running and the spectral density FFT of the filtered signal during running, according to an embodiment of the present invention. Figure 6 This is a schematic diagram of the original GCG signal in the resting state and the filtered GCG signal and peak detection results in the resting state, according to an embodiment of the present invention. Figure 7 This is a schematic diagram of the original GCG signal under jogging conditions and the filtered GCG signal and peak detection results under jogging conditions, according to an embodiment of the present invention. Figure 8 This is a schematic diagram of the original GCG signal under fast running conditions and the filtered GCG signal and peak detection results under fast running conditions according to an embodiment of the present invention. Figure 9 This is a comparison diagram of GCG signal and electrocardiogram signal during jogging according to an embodiment of the present invention. Detailed Implementation

[0023] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0025] Existing ECG, BCG, and PPG technologies cannot reliably detect HRV during exercise (jogging, running) – ECG is affected by sweating and motion artifacts, BCG requires fixed equipment (such as mattress sensors) and cannot move, and PPG suffers from severe signal drift during motion.

[0026] Existing technologies have problems such as poor wearing comfort (ECG electrodes need to be close to the skin, PPG electrodes need to be tight), weak anti-interference ability (ambient light, sweat), and peak missed detection (heartbeat signals are masked during exercise).

[0027] The lack of a low-power, portable HRV detection solution that can operate across multiple states, including resting, jogging, and running, fails to meet the needs of daily health monitoring and exercise management.

[0028] Therefore, this embodiment proposes a dynamic heart rate variability detection system based on a MEMS IMU chest strap, such as... Figure 1 ,include: Flexible chest strap: for wearing on the user's chest; Sensor MEMS IMU module: used to acquire the user's micro-vibration angular rate (GCG) signal and real-time step frequency; Control and transmission module: used to execute detection algorithms and wirelessly transmit data; Power supply module: Used to provide a stable power supply; The inertial measurement module, the control and transmission module, and the power supply module are all integrated into the flexible chest strap.

[0029] Furthermore, the flexible chest strap uses elastic fabric as a carrier and integrates four core modules internally, encapsulated in a sealed waterproof structure (to prevent interference from sweat and rain). It can be worn close to the chest (without needing to be in direct contact with the skin). The specific components are as follows: Power supply module: Includes a polymer lithium battery (capacity 300mAh, battery life ≥8 hours) and a low dropout linear regulator (LDO) module. The LDO outputs a stable 3.3V voltage to power the entire system.

[0030] Inertial Measurement Unit (MEMS IMU): The core sensor is STMicroelectronics' LSM6DSR; Control and transmission module: It adopts a Bluetooth chip and also functions as a "microprogram controller (MCU)" - it is responsible for running the detection algorithm and transmitting the raw data or detection results (heart rate, HRV index) to the smartphone / terminal via Bluetooth 5.0. It also supports low power mode (standby current <10μA).

[0031] Furthermore, the inertial measurement module includes a three-axis gyroscope and a three-axis accelerometer. The three-axis gyroscope is used to collect the micro-vibration angular rate (GCG) signal of the user's chest caused by the heartbeat, with a measurement range of 0-4000 dps. The three-axis accelerometer is used to detect the user's real-time step frequency and assist in judging the motion state.

[0032] Specifically, the three-axis gyroscope is used to acquire the micro-vibration angular velocity signal (GCG signal) generated by the heartbeat in the chest. The measurement range is 0-4000 dps, the bias drift is ±0.005 dps / ℃, and the rate noise density is 5 mdps / √Hz (high performance mode) to ensure signal accuracy. Triaxial accelerometer: used to detect real-time step frequency and assist in judging motion status.

[0033] Furthermore, the processing flow of the control and transmission module includes: Based on the signals collected by the triaxial accelerometer, the step frequency is extracted by Fast Fourier Transform (FFT) analysis, and the user's motion state is determined based on the step frequency. Based on the motion state, an adaptive filtering strategy is dynamically selected to filter the GCG signal, and a secondary peak detection is performed on the filtered GCG signal to extract the heartbeat peak sequence. Heart rate variability (HRV) is calculated based on the aforementioned peak heart rate sequence.

[0034] Specifically, the sampling frequency of the MEMS IMU module is set to 128Hz (balancing signal integrity and low power consumption), the gyroscope only collects angular rate signals perpendicular to the human torso and parallel to the chest (where heartbeat micro-vibrations are most significant), and the triaxial accelerometer collects triaxial acceleration signals for state determination.

[0035] Furthermore, the adaptive filtering strategy includes: In all motion states, the GCG signal is filtered using a bandpass filter (BPF) with a passband frequency of 1 Hz to 9 Hz; During running, a band-stop filter (BEF) with a stopband frequency of 2.5Hz to 3.5Hz is used to filter out step frequency interference.

[0036] Specifically, the adaptive filtering strategy (suppressing motion / respiratory interference) includes: Step 1: Step frequency analysis - Perform a Fast Fourier Transform (FFT) on the signal collected by the accelerometer to extract the main peak in the frequency range of 0.5-3Hz. This peak corresponds to the step frequency in motion (0.4-1Hz for slow running and 2.5-3.5Hz for running).

[0037] Step 2: Filter Selection – Dynamically configure the filter based on the motion state determined by the step frequency: Universal for all conditions: Employs a 1Hz-9Hz bandpass filter (BPF) to filter out high-frequency noise (>9Hz, such as environmental vibrations) and low-frequency baseline drift (<1Hz, such as slow body tilt).

[0038] The exercise states include resting state, jogging state, and running state, such as Figures 2-8 .

[0039] Additional filtering during running: A 2.5Hz-3.5Hz band-stop filter (BEF) is used to filter out motion interference corresponding to the running cadence (this frequency band partially overlaps with the heartbeat signal). Resting / jogging state: No BEF required (jogging frequency of 0.4-1Hz, which has been filtered out by BPF; no cadence interference in the resting state).

[0040] Furthermore, the transfer function of the bandpass filter (BPF) is: ; In the formula, ξb The damping ratio; ωb The center frequency; s For Laplace variables, in the form of s = σ + jω , σ For the actual part, j For imaginary units, ω ω is the angular frequency, used to transform the differential / integral relationship in the time domain into an algebraic relationship in the complex frequency domain; The transfer function of the bandpass filter; in this embodiment, the damping ratio ξb =1.3, center frequency ωb =3Hz.

[0041] Furthermore, the secondary peak detection includes: First peak detection: The initial peak position is extracted from the filtered GCG signal using an adaptive threshold method; Secondary fine detection: Calculate the interval between adjacent preliminary peaks to identify whether there are any missed peaks. If the interval is greater than the preset average interval, re-detect and supplement peaks within the interval to form the final peak sequence.

[0042] Specifically, including: Peak detection: An adaptive thresholding method (threshold is the signal mean + 3 times the standard deviation) is used to extract the initial peak positions (Peaks_locs) from the filtered GCG signal.

[0043] Secondary fine detection (to solve missed detections in motion): Calculate the interval between adjacent preliminary peaks (AO-AO interval, AOAO-intervals=diff (Peaks_locs)) and the mean of the intervals (Mean-AOAO-interval=mean (AOAO-intervals)); for intervals that satisfy "AOAO-intervals>1.5×Mean-AOAO-interval" (this interval has missed peaks), re-detect possible peaks (Possible_peaks) within the interval, and select the peak "closest to the midpoint of the interval" as the supplementary peak (New_peak); add the supplementary peak to the preliminary peaks, sort them to obtain the final peaks (Final_peaks), ensuring that every heartbeat is detected.

[0044] Based on the AO-AO interval of the final peak value (analogous to the RR interval of ECG), the HRV index is calculated using three types of analysis methods: Time-domain metrics: AO-AO interval mean (AO-AOMEAN), population standard deviation (SDNN), root mean square deviation (RMSSD), number of heartbeats with an adjacent interval difference >50ms (NN50), percentage of NN50 (PNN50), standard deviation of adjacent interval difference (SDSD); Frequency domain parameters: Very low frequency energy (VLF, 0.0033-0.04Hz), low frequency energy (LF, 0.04-0.15Hz), high frequency energy (HF, 0.15-0.4Hz), total energy (TP), LF / HF ratio; Non-linear indicators: standard deviations of scatter plots, SD1 and SD2.

[0045] Furthermore, the system also includes an auxiliary verification module, which is used to synchronously acquire standard ECG signals and compare and verify them with GCG detection results.

[0046] Specifically, to verify accuracy, the auxiliary verification module synchronously acquires standard ECG signals, and the data is stored using LabVIEW software for comparison with GCG detection results, such as... Figure 9 .

[0047] Furthermore, in this embodiment, the sensor MEMS IMU module can be replaced with other models of "three-axis gyroscope + three-axis accelerometer" combination, such as ADI's ADXL345 (accelerometer) + ITG3200 (gyroscope) - as long as the following parameters are met: gyroscope measurement range ≥2000dps, bias drift ≤±0.01dps / ℃, accelerometer measurement range ±4g, the accuracy of GCG signal acquisition can be guaranteed without affecting subsequent filtering and peak detection.

[0048] The Bluetooth module can be replaced with a WiFi module (such as ESP8266) or an NFC module: WiFi is suitable for long-distance transmission (such as real-time monitoring of sports venues), while NFC is suitable for short-distance fast synchronization (such as synchronizing data with smartwatches). As long as the transmission rate is ≥1Mbps and the power consumption is ≤10mA, the overall system performance will not be affected.

[0049] This embodiment also provides a detection method for implementing the dynamic heart rate variability detection system based on a MEMS IMU chest strap as described in the claims, including: GCG and acceleration signals are acquired through an inertial measurement module in a flexible chest strap; Based on the acceleration signal, step frequency is extracted through FFT analysis to determine the user's motion state; The GCG signal is filtered by dynamically selecting an adaptive filtering strategy based on the motion state. Perform secondary peak detection on the filtered GCG signal to extract the heartbeat peak sequence; Heart rate variability (HRV) is calculated based on the aforementioned peak heart rate sequence.

[0050] Specifically, including: Wearing and activation: The user fixes the chest strap to the left chest area in front of the heart. After the accelerometer detects that the strap is being worn, the system automatically starts and connects to the smartphone via Bluetooth.

[0051] Status determination and signal acquisition: The accelerometer determines the motion status (resting / jogging / running) in real time, and the gyroscope synchronously acquires the angular rate signal and transmits it to the MCU.

[0052] Data preprocessing: The MCU performs FFT step frequency analysis and adaptive filtering (BPF±BEF) to output the denoised GCG signal.

[0053] Peak detection and HRV calculation: Perform one peak detection → two fine detections → calculate the AO-AO interval → obtain the HRV index.

[0054] Data output and verification: Bluetooth transmits heart rate and HRV indicators to a mobile app for display, and compares them with synchronously collected ECG results to ensure accuracy.

[0055] Low power consumption control: If the user removes the chest strap or remains still (without vital signs), the system will automatically go into sleep mode to reduce power consumption.

[0056] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A dynamic heart rate variability detection system based on a MEMS IMU chest strap, characterized in that, include: Flexible chest strap: for wearing on the user's chest; Inertial Measurement Unit (MEMS IMU): Used to acquire the user's micro-vibration angular rate (GCG) signal and real-time step frequency; Control and transmission module: used to execute detection algorithms and wirelessly transmit data; Power supply module: Used to provide a stable power supply; The inertial measurement module, the control and transmission module, and the power supply module are all integrated into the flexible chest strap.

2. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 1, characterized in that, The inertial measurement module includes a three-axis gyroscope and a three-axis accelerometer. The three-axis gyroscope is used to collect the micro-vibration angular rate (GCG) signal of the user's chest caused by the heartbeat, with a measurement range of 0-4000 dps. The three-axis accelerometer is used to detect the user's real-time step frequency and assist in judging the motion state.

3. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 2, characterized in that, The processing flow of the control and transmission module includes: Based on the signals collected by the triaxial accelerometer, the step frequency is extracted by Fast Fourier Transform (FFT) analysis, and the user's motion state is determined based on the step frequency. Based on the motion state, an adaptive filtering strategy is dynamically selected to filter the GCG signal, and a secondary peak detection is performed on the filtered GCG signal to extract the heartbeat peak sequence. Heart rate variability (HRV) is calculated based on the aforementioned peak heart rate sequence.

4. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 3, characterized in that, The adaptive filtering strategy includes: In all motion states, the GCG signal is filtered using a bandpass filter (BPF) with a passband frequency of 1 Hz to 9 Hz; During running, a band-stop filter (BEF) with a stopband frequency of 2.5Hz to 3.5Hz is used to filter out step frequency interference.

5. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 4, characterized in that, The exercise states include resting state, jogging state, and running state.

6. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 4, characterized in that, The transfer function of the bandpass filter (BPF) is: ； In the formula, ξb The damping ratio; ωb The center frequency; s For Laplace variables, in the form of s = σ + jω , σ For the actual part, j For imaginary units, ω ω is the angular frequency, used to transform the differential / integral relationship in the time domain into an algebraic relationship in the complex frequency domain; Let be the transfer function of the bandpass filter.

7. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 3, characterized in that, The secondary peak detection includes: First peak detection: The adaptive threshold method is used to extract the initial peak position from the filtered GCG signal; Secondary fine detection: Calculate the interval between adjacent preliminary peaks to identify whether there are any missed peaks. If the interval is greater than the preset average interval, re-detect and supplement peaks within the interval to form the final peak sequence.

8. The dynamic heart rate variability detection system based on a MEMS IMU chest strap according to claim 1, characterized in that, The system also includes an auxiliary verification module, which is used to synchronously acquire standard ECG signals and compare them with GCG detection results for verification.

9. A detection method for implementing the dynamic heart rate variability detection system based on a MEMS IMU chest strap as described in any one of claims 1-8, characterized in that, include: GCG and acceleration signals are acquired through an inertial measurement module in a flexible chest strap; Based on the acceleration signal, step frequency is extracted through FFT analysis to determine the user's motion state; The GCG signal is filtered by dynamically selecting an adaptive filtering strategy based on the motion state. Perform secondary peak detection on the filtered GCG signal to extract the heartbeat peak sequence; Heart rate variability (HRV) is calculated based on the aforementioned peak heart rate sequence.

10. The detection method according to claim 9, characterized in that, It also includes transmitting the HRV index to a smart terminal for display and storage via wireless transmission.

Images