radarODE: An ODE-Embedded Deep Learning Model for Contactless ECG Reconstruction from Millimeter-Wave Radar

TL;DR

RadarODE bridges mechanical-to-electrical cardiac domain transformation by embedding ordinary differential equations as a morphological prior within an ODE-guided neural framework. It introduces a compact signal model for fine-grained cardiac radar features, a single-cycle ECG generator (SCEG) with an ODE solver, and a long-term reconstruction path that fuses temporal and morphological cues. Across a public mm-wave radar dataset, RadarODE achieves higher RMSE/PCC fidelity and lower MDR, especially under random body movement, compared with state-of-the-art baselines, and demonstrates improved R-peak timing and fine-grained cardiac event reconstruction. The results suggest practical potential for robust contactless ECG monitoring, with SST-based feature extraction and ODE priors contributing to improved robustness and interpretability.

Abstract

Radar-based contactless cardiac monitoring has become a popular research direction recently, but the fine-grained electrocardiogram (ECG) signal is still hard to reconstruct from millimeter-wave radar signal. The key obstacle is to decouple the cardiac activities in the electrical domain (i.e., ECG) from that in the mechanical domain (i.e., heartbeat), and most existing research only uses pure data-driven methods to map such domain transformation as a black box. Therefore, this work first proposes a signal model for domain transformation, and then a novel deep learning framework called radarODE is designed to fuse the temporal and morphological features extracted from radar signals and generate ECG. In addition, ordinary differential equations are embedded in radarODE as a decoder to provide morphological prior, helping the convergence of the model training and improving the robustness under body movements. After being validated on the dataset, the proposed radarODE achieves better performance compared with the benchmark in terms of missed detection rate, root mean square error, Pearson correlation coefficient with the improvement of 9%, 16% and 19%, respectively. The validation results imply that radarODE is capable of recovering ECG signals from radar signals with high fidelity and can be potentially implemented in real-life scenarios.

Figures (13)

• Figure 1: Relationships between cardiac mechanical and electrical activities within the same cardiac cycle.
• Figure 2: Overview of the radarODE framework for domain transformation: (a) Single-cycle ECG length (PPI) estimation; (b) Single-cycle ECG generator; (c) Long-term ECG reconstruction.
• Figure 3: Spectrograms obtained from the same radar signal: (a) Radar signal $\tilde{x}(t)$; (b) STFT result; (c) CWT result; (d) SST result with $T_1$, $T_2$ labeled, revealing concentrated energy around vibrations and clean background.
• Figure 4: Energy plot of the synchronous radar signals with different detected peaks: (a) Energy plot with high SNR and correct detection; (b) Energy plot with low SNR and wrong detection.
• Figure 5: Architecture of SCEG with SST segments as input and single-cycle ECG pieces as output.