Hilbert-Huang Transform and Spectrum Weighted Reconstruction Integration for Millimeter Wave Radar Based Heart Rate Detection

In recent years, health issues have been serious worldwide. As an important vital sign indicator to evaluate human health, heart rate (HR) detection has become an urgent need of society without disturbance and comfort. Traditional detection methods in medical institutions, such as photoplethysmography (PPG) and electrocardiography (ECG), though effective in providing real-time and accurate data, face limitations in terms of comfort and versatility. Advances in radar technologies make it possible to detect HR without contact. However, as the chest wall displacement caused by heartbeat is extremely weak, HR is easy to be overwhelmed by respiration harmonics, noise and clutter, so how to address the issue of the unknown environmental noise, and respiratory harmonic interference during the detection process are two critical challenges. To tackle the aforementioned challenges, in this paper, we propose a non-contact HR detection approach for millimeter wave radar based on Hilbert-Huang transform and spectrum weighted reconstruction to achieve accurate HR estimation without disturbance. The approach includes a micro-moving target location strategy and a heart rate reconstruction estimation strategy. In micro-motion target localization strategy, we first eliminate static clutter from the raw data. Then, building on the traditional Constant False Alarm Rate (CFAR) method, we design an adaptive CFAR approach that adjusts dynamically based on environmental noise thresholds, which reduces the impact of random noise and improves the sensitivity and accuracy of weak signal target detection during HR monitoring. In heartbeat signal reconstruction strategy, we first utilize the Hilbert-Huang Transform (HHT) for high-resolution time-frequency localization analysis, capturing transient features and variations of non-stationary and nonlinear signals such as heartbeats. By extracting the Intrinsic Mode Functions (IMFs) corresponding to the heart rate range and designing a spectral weighting reconstruction method, we segment and enhance the heart rate interval，which further suppresses respiratory harmonics and noise interference in the heartbeat signal, thereby improving the resolution of heart rate detection. Experiments were conducted in both laboratory and office settings using Texas Instruments IWR1843 millimeter-wave radar sensor. The performance of the proposed method under the different HR and the different individual states are investigated by extensive experiments. The results indicate that the proposed method could effectively suppress the respiration harmonics, noise and clutter interference, achieving superior heartbeat signal decomposition and reconstruction compared to existing methods. The average HR error is 1.16 BPM, significantly enhancing HR detection accuracy.