Non-invasive lvef monitoring method and system based on millimeter wave radar chest displacement signal

Abstract

The application provides a non-invasive LVEF monitoring method and system based on a millimeter wave radar chest displacement signal, comprising the following steps: S1, millimeter wave radar signal emission; S2, chest micro-motion signal collection; S3, chest micro-motion signal preprocessing; S4, vital sign separation; S5, heart failure feature extraction; S6, risk assessment; and S7, result output. The application provides the creative combination of radar sensing technology and medical diagnosis, and constructs a non-contact and continuous monitoring heart function evaluation system. Compared with traditional LVEF detection methods (such as echocardiography, cardiac MRI, etc.), breakthrough improvements are realized in clinical practicability, data continuity and technical popularity.

Description

Technical Field

[0001] This invention belongs to the field of intelligent medical monitoring, specifically relating to a non-invasive LVEF monitoring method and system based on millimeter-wave radar chest displacement signals. Background Technology

[0002] Heart failure (HF) is a serious and life-threatening chronic disease. Early detection and dynamic monitoring are crucial for preventing deterioration and guiding treatment. Current HF monitoring methods mainly fall into two categories: contact sensor monitoring and non-contact physiological signal detection, each with different system structures and technical implementations. However, existing technologies have limitations in terms of user comfort, measurement accuracy, power consumption control, and clinical applicability, making it difficult to meet the practical needs of high-precision, low-cost, and non-invasive real-time monitoring in scenarios such as community healthcare and remote health management.

[0003] Contact-based monitoring solutions primarily utilize electrocardiogram (ECG) patches, electrode chest straps, etc., which are attached to the surface of the human body to collect physiological signals such as heart rate and respiratory rate in real time. A typical system includes sensing electrodes, front-end signal conditioning circuitry, analog-to-digital conversion module, microcontroller (MCU), communication module (such as Bluetooth / WiFi), and power supply system.

[0004] However, this type of technology has the following drawbacks:

[0005] (1) Poor user comfort: As described in Chinese patent CN110123345A, prolonged application of ECG patches can easily lead to skin allergies or local inflammation, reducing user compliance;

[0006] (2) Activity restriction: As described in US Patent US2022156789, the chest strap sensor restricts the user's activity and is not suitable for the elderly or those who need to wear it for a long time;

[0007] (3) High usage and maintenance costs: Most surface mount products are disposable consumables that need to be replaced frequently, increasing the burden of use;

[0008] (4) Sensitive to wearing position: The device needs to fit a specific area, and even a slight deviation in the wearing method will seriously affect the signal quality.

[0009] Non-contact monitoring systems remotely sense minute physiological activities of the human body through optical, acoustic, or electromagnetic waves, resulting in higher user acceptance. Common technologies include:

[0010] 1) Optical camera-based solution: This solution uses ordinary or infrared cameras to acquire images of subtle color changes in human skin or chest rise and fall, and then estimates heart rate and respiratory rate.

[0011] Defects: As shown in international patent WO2023108122, this solution is extremely sensitive to ambient light. The detection accuracy decreases under low light or strong backlight conditions, and the camera poses privacy risks, limiting its application in home or bedroom scenarios.

[0012] 2) Ultrasound-based detection scheme: Low-frequency physiological parameters such as respiration are obtained by emitting low-frequency ultrasound waves to the target and analyzing the changes in its echoes.

[0013] Defects: As disclosed in Japanese Patent JP2024078931, the effective distance of ultrasonic detection is usually less than 1.5 meters, which makes it difficult to support flexible deployment requirements such as bedside and wall, and the signal stability is significantly affected by environmental structure interference.

[0014] 3) Millimeter-wave radar-based solution: Using FMCW (Frequency Modulated Continuous Wave) millimeter-wave radar to capture micro-motion signals in the human chest cavity for non-contact heart rate and respiration monitoring. It has strong penetration and all-weather characteristics and has become a research hotspot in recent years.

[0015] Disadvantages: Commercial radar modules have high power consumption (e.g., the TI IWR6843 solution consumes more than 5W), making it difficult to adapt to low-power devices; hardware resolution is limited, and signal acquisition is sensitive to attitude and size, causing fluctuations in measurement stability; the algorithm is complex and has poor adaptability to the deployment environment, and a standardized solution has not yet been formed. Summary of the Invention

[0016] To address at least one of the technical problems in the background art described above, the object of this invention is to provide [the solution / solution].

[0017] The present invention solves the technical problem by adopting the following technical solution:

[0018] A non-invasive LVEF monitoring method based on millimeter-wave radar thoracic cavity displacement signals includes the following steps:

[0019] Step S1: Millimeter-wave radar signal transmission;

[0020] Step S2: Acquisition of thoracic cavity micro-motion signals;

[0021] Step S3: Preprocessing of thoracic cavity micromotion signals;

[0022] Step S4, separation of vital signs;

[0023] Step S5, Heart Failure Feature Extraction;

[0024] Step S6, Risk Assessment;

[0025] Step S7, output the result.

[0026] Furthermore, in step S1, the millimeter-wave radar signal transmission uses a linear frequency modulated continuous wave with a frequency slope of 0.5. ;

[0027] Antenna layout optimization: The formula for calculating antenna spacing is as follows:

[0028]

[0029]

[0030] in, d is the radar wavelength; d is the antenna spacing.

[0031] Furthermore, in step S2, the thoracic cavity micromotion signal includes: heart rate cycle, amplitude, and time-frequency characteristics.

[0032] Further, in step S3, the preprocessing of the thoracic cavity micromotion signal includes:

[0033] Phase demodulation:

[0034]

[0035] Displacement transformation:

[0036]

[0037] in, d(t) is the radar wavelength, and d(t) is the thoracic displacement signal.

[0038] Further, in step S4, vital signs are separated, including:

[0039] Separate heartbeat signals;

[0040] The heart rate cycle T is located based on a dynamic threshold, as shown in the following formula:

[0041]

[0042] in, This represents the i-th peak heart rate moment;

[0043] The segmented heartbeat cycle signal is used for feature extraction.

[0044] Furthermore, in step S5, the method for extracting heart failure features includes:

[0045] Biomarkers associated with heart failure were extracted from the pure signal, including the following key features:

[0046] LVEF, which uses displacement waveforms to invert changes in ventricular volume;

[0047] E / A ratio, used to analyze diastolic function;

[0048] HRV, heart rate variability, assesses autonomic nervous system status.

[0049] Furthermore, in step S6, heart failure risk stratification is performed by integrating LVEF, E / A ratio, and HRV data.

[0050] A non-invasive LVEF monitoring system based on millimeter-wave radar chest cavity displacement signals includes: a millimeter-wave radar signal transmission module, a chest cavity micro-motion signal acquisition module, a chest cavity micro-motion signal preprocessing module, a vital signs separation module, a heart failure feature extraction module, a risk assessment module, and a result output module;

[0051] The millimeter-wave radar signal transmitting module is used to transmit linear frequency modulated continuous waves with a frequency slope of 0.5. ;

[0052] The chest cavity micro-motion signal acquisition module is used to acquire the heartbeat cycle, amplitude, and time-frequency characteristics;

[0053] The preprocessing module for the thoracic cavity micro-motion signal is used to perform phase demodulation and displacement conversion on the thoracic cavity micro-motion signal;

[0054] The vital signs separation module is used to separate heartbeat signals;

[0055] The heart failure feature extraction module is used to extract heart failure-related biomarkers from pure signals;

[0056] The risk assessment module integrates LVEF, E / A ratio, and HRV data to stratify heart failure risk.

[0057] The result output module is used to output the non-invasive LVEF monitoring results.

[0058] The beneficial effects of this invention are:

[0059] This invention addresses the three core pain points of traditional LVEF testing—continuity, accessibility, and cost—through an innovative combination of non-contact sensing and artificial intelligence algorithms. It not only improves the accuracy of clinical decision-making but also advances cardiac function assessment from "occasional hospital testing" to a new paradigm of "routine monitoring," demonstrating significant medical value and socio-economic benefits. Attached Figure Description

[0060] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] refer to Figure 1 This invention provides a non-invasive LVEF monitoring method based on millimeter-wave radar thoracic cavity displacement signals, comprising the following steps:

[0063] Step S1: Millimeter-wave radar signal transmission;

[0064] Step S2: Acquisition of thoracic cavity micro-motion signals;

[0065] Step S3: Preprocessing of thoracic cavity micromotion signals;

[0066] Step S4, separation of vital signs;

[0067] Step S5, Heart Failure Feature Extraction;

[0068] Step S6, Risk Assessment;

[0069] Step S7, output the result.

[0070] To further optimize the technical solution, in step S1, the millimeter-wave radar signal transmission adopts a linear frequency modulated continuous wave with a frequency slope of 0.5. ;

[0071] The raw echo signal received by a radar sensor (such as a 60GHz FMCW radar) is shown in the following formula:

[0072]

[0073] in The original echo signal, f represents the amplitude of the echo signal. c For carrier frequency, This includes phase modulation caused by micromovements in the thoracic cavity.

[0074] Antenna layout optimization: The formula for calculating antenna spacing is as follows:

[0075]

[0076]

[0077] in, d is the radar wavelength; d is the antenna spacing.

[0078] To eliminate radar signal interference caused by body movements (such as turning over or coughing), accelerometer data is used to compensate for the displacement signal.

[0079] The mathematical modeling method is as follows:

[0080] Calculate displacement by integral of acceleration:

[0081]

[0082] Z-axis acceleration data (unit: mm / s²); Sampling frequency (unit: Hz); This refers to the illusory displacement caused by motion.

[0083] Compensation formula:

[0084]

[0085] This is a false displacement caused by motion; This is the original thoracic cavity displacement signal; This is the empirical coupling coefficient (usually taken as 0.8).

[0086] Implementation steps: (1) Collect triaxial accelerometer data (focus on the Z-axis, i.e., perpendicular to the chest cavity). (2) Perform a second integration on the acceleration signal to obtain the motion artifact displacement. (3) Subtract the motion artifact from the original displacement to obtain the pure chest cavity displacement signal.

[0087] To further optimize the technical solution, in step S2, the thoracic cavity micromotion signal includes: heartbeat cycle, amplitude, and time-frequency characteristics.

[0088] To filter out interference from respiratory signals on heartbeat detection (respiratory frequency is typically 0.1~0.5Hz, which modulates the heartbeat signal), an adaptive notch filter is used to suppress the respiratory fundamental frequency and its harmonics.

[0089] Filter Design (MATLAB Example):

[0090] Setting parameters:

[0091] Respiratory rate:

[0092] Normalized cutoff frequency

[0093] bandwidth (Control filter steepness)

[0094] Transfer function (second-order IIR notch filter):

[0095]

[0096] coefficient ,coefficient ,Depend on generate.

[0097] Zero-phase filtering:

[0098] filtered_signal = filtfilt(b, a, displacement); bidirectional filtering eliminates phase distortion.

[0099] A deep attenuation (e.g., -30dB) is created at the respiratory rate, while preserving the heartbeat signal (>0.8Hz).

[0100] Further optimize the technical solution. In step S3, the preprocessing of the thoracic cavity micromotion signal includes:

[0101] Phase demodulation:

[0102]

[0103] Displacement transformation:

[0104]

[0105] in, d(t) is the radar wavelength, and d(t) is the thoracic displacement signal.

[0106] Further optimize the technical solution. In step S4, vital signs separation includes:

[0107] Separate heartbeat signals;

[0108] The heart rate cycle T is located based on a dynamic threshold, as shown in the following formula:

[0109]

[0110] in, This represents the i-th peak heart rate moment;

[0111] The segmented heartbeat cycle signal is used for feature extraction.

[0112] Further optimization of the technical solution: In step S5, the method for extracting heart failure features includes:

[0113] Biomarkers associated with heart failure were extracted from the pure signal, including the following key features:

[0114] LVEF (Leakage Frame Ejection) is used to invert ventricular volume changes through displacement waveforms. Its function is to extract cardiac volume changes from raw signals (such as radar or ultrasound data) using algorithms and calculate the ejection fraction. Input: Heartbeat signal (time-domain waveform or radar displacement data). Output: LVEF value (percentage). Located at the core, it directly receives heartbeat signals and provides foundational data for subsequent risk assessment.

[0115] Heartbeat signal, function: provides raw heartbeat waveform or chest displacement signal. Data source: Radar sensor (non-contact monitoring, extracting chest cavity micro-motions). Electrocardiogram (ECG) or photoplethysmography (PPG). Output: Heartbeat cycle, amplitude, and time-frequency characteristics. As an input layer, it connects to the LVEF calculation and HRV assessment modules. The heartbeat signal is the fundamental data source for all analyses, simultaneously supplying three downstream modules.

[0116] The E / A ratio analyzes diastolic function. Function: Assessing cardiac diastolic function (by the ratio of the E peak to the A peak in the mitral valve flow spectrum). Input: Hemodynamic characteristics in the heartbeat signal (requires Doppler ultrasound or radar inversion). Output: E / A ratio (normal >1; abnormal indicates diastolic dysfunction). It is calculated in parallel with LVEF and both are input to the risk stratification module. LVEF and the E / A ratio assess systolic and diastolic function respectively, complementing each other rather than being sequential.

[0117] HRV (Heart Rate Variability) assesses autonomic nervous system status. Function: Analyzes autonomic nervous system regulation through heart rate variability (HRV). Input: RR interval (time difference between adjacent heartbeats) of the heartbeat signal. Output: HRV indices such as SDNN and RMSSD (reflecting sympathetic / parasympathetic balance). It shares heartbeat signal data with the LVEF calculation module, and its output is independent but participates in the overall risk assessment. HRV assessment provides an additional dimension (neuromotor regulation), jointly determining the risk level with cardiac function indicators.

[0118] This invention designs and implements a closed-loop process from signal acquisition to clinical decision-making, balancing efficiency and comprehensiveness.

[0119] Feature extraction methods include:

[0120] Time-domain characteristics (direct statistics): Heart rate variability (HRV); Standard deviation of peak interval;

[0121] Frequency domain characteristics (FFT transform):

[0122]

[0123] Extract the energy ratio of low frequency (LF) and high frequency (HF); among which, For the frequency domain, f is the heart rate.

[0124] Model parameters are adjusted based on individual patient differences (such as chest wall thickness and heart rate variability), and linear regression calibration is performed based on baseline clinical data (such as echocardiographic LVEF values).

[0125] 1) Feature extraction:

[0126] Extracting time-frequency domain features from historical radar data (such as heart rate amplitude, HRV).

[0127] Corresponding clinical measurement value (e.g., LVEF%).

[0128] 2) Regression calibration:

[0129]

[0130] w is the feature weight vector, and b is the bias term (the focus of calibration).

[0131] 3) Parameter update:

[0132] Fit w and b using the least squares method:

[0133]

[0134] The implementation steps are as follows: Query the database to obtain the patient's historical characteristics and clinical labels. Train a linear regression model and optimize the bias term b. Inject the calibrated b into the main model (e.g., the base_score parameter in XGBoost).

[0135] To further optimize the technical solution, in step S6, heart failure risk stratification is performed by integrating LVEF, E / A ratio, and HRV data.

[0136] The present invention also provides a non-invasive LVEF monitoring system based on millimeter-wave radar chest displacement signals, comprising: a millimeter-wave radar signal transmission module, a chest micro-motion signal acquisition module, a chest micro-motion signal preprocessing module, a vital signs separation module, a heart failure feature extraction module, a risk assessment module, and a result output module;

[0137] The millimeter-wave radar signal transmitting module is used to transmit linear frequency modulated continuous waves with a frequency slope of 0.5. ;

[0138] The chest cavity micro-motion signal acquisition module is used to acquire the heartbeat cycle, amplitude, and time-frequency characteristics;

[0139] The preprocessing module for the thoracic cavity micro-motion signal is used to perform phase demodulation and displacement conversion on the thoracic cavity micro-motion signal;

[0140] The vital signs separation module is used to separate heartbeat signals;

[0141] The heart failure feature extraction module is used to extract heart failure-related biomarkers from pure signals;

[0142] The risk assessment module integrates LVEF, E / A ratio, and HRV data to stratify heart failure risk.

[0143] The result output module is used to output the non-invasive LVEF monitoring results.

[0144] This invention addresses the three core pain points of traditional LVEF testing—continuity, accessibility, and cost—through an innovative combination of non-contact sensing and artificial intelligence algorithms. It not only improves the accuracy of clinical decision-making but also advances cardiac function assessment from "occasional hospital testing" to a new paradigm of "routine monitoring," demonstrating significant medical value and socio-economic benefits.

[0145] The advantages compared with traditional LVEF detection methods are shown in Table 1:

[0146] Table 1

[0147]

[0148] The improved clinical outcomes brought about by this invention are as follows:

[0149] (1) Early warning capability of heart failure

[0150] Traditional limitations: Ultrasound / MRI cannot be used frequently, and it is easy to miss the diagnosis of compensated heart failure (the stage of slow decline in LVEF).

[0151] By analyzing continuous LVEF trends (e.g., triggering an alert if daily fluctuations exceed 5%), deterioration of cardiac function can be detected 3-6 months earlier than traditional methods.

[0152] By combining respiratory-heart rate coupling indexes, diastolic heart failure (HFpEF, traditional LVEF often shows "normal") can be identified.

[0153] (2) Precision treatment guidance

[0154] Medication adjustment: Dynamic LVEF data can optimize the dosage of drugs such as beta-blockers and ARNI (e.g., initiating intensive therapy when LVEF < 40%).

[0155] Postoperative monitoring: After cardiac resynchronization therapy (CRT), the efficacy is assessed in real time to reduce unnecessary electrode adjustments.

[0156] (3) Reduce medical costs

[0157] Reduce hospitalizations: Through early intervention, the hospitalization rate for acute exacerbations of heart failure can be reduced by 30%.

[0158] Alternative to some ultrasound examinations: For patients in the stable phase, radar monitoring can replace 50% of follow-up ultrasound examinations.

[0159] This invention has social value and industry impact, including: Tiered healthcare: empowering community hospitals to conduct cardiac function screening, alleviating pressure on tertiary hospitals. Aging population management: providing unmanned monitoring solutions for home-based elderly care, reducing nursing costs. Medical insurance cost control: reducing the high treatment costs of end-stage heart failure through preventative medicine.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions 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 invention.

Claims

1. A non-invasive LVEF monitoring system based on millimeter-wave radar thoracic displacement signals, characterized in that, include: The system includes a millimeter-wave radar signal transmission module, a chest cavity micro-motion signal acquisition module, a chest cavity micro-motion signal preprocessing module, a vital signs separation module, a heart failure feature extraction module, a risk assessment module, and a result output module. The millimeter-wave radar signal transmitting module is used to transmit linear frequency modulated continuous waves with a frequency slope of 0.

5. ; The chest cavity micro-motion signal acquisition module is used to acquire the heartbeat cycle, amplitude, and time-frequency characteristics; The preprocessing module for the thoracic cavity micro-motion signal is used to perform phase demodulation and displacement conversion on the thoracic cavity micro-motion signal to obtain the thoracic cavity displacement signal; The vital signs separation module is used to separate heartbeat signals; The heart failure feature extraction module is used to extract heart failure-related biomarkers from pure signals; The risk assessment module integrates LVEF, E / A ratio, and HRV data to stratify heart failure risk. The result output module is used to output the non-invasive LVEF monitoring results; The method for using a non-invasive LVEF monitoring system based on millimeter-wave radar chest displacement signals includes the following steps: Step S1: Millimeter-wave radar signal transmission; Step S2: Acquisition of thoracic cavity micro-motion signals; Step S3: Preprocessing of thoracic cavity micromotion signals; Step S4, separation of vital signs; Step S5, Heart Failure Feature Extraction; Step S6, Risk Assessment; Step S7, output the result; In step S1, the millimeter-wave radar signal transmission uses a linear frequency modulated continuous wave with a frequency slope of 0.

5. ; Antenna layout optimization: The formula for calculating antenna spacing is as follows: in, d is the radar wavelength; d is the antenna spacing. In step S4, vital signs are separated, including: Separate heartbeat signals; The heart rate cycle T is located based on a dynamic threshold, as shown in the following formula: in, This represents the i-th peak heart rate moment; The segmented heartbeat cycle signal is used for feature extraction. In step S5, the methods for extracting heart failure features include: Biomarkers associated with heart failure were extracted from the pure signal, including the following key features: LVEF, which uses displacement waveforms to invert changes in ventricular volume; E / A ratio, used to analyze diastolic function; HRV, heart rate variability, assesses autonomic nervous system status.

Images