Related Experiment Video
Updated: Aug 6, 2026
08:21
Wideband Optical Detector of Ultrasound for Medical Imaging Applications
Published on: May 11, 2014
MM-FGHM: Fine-Grained Heartbeat Monitoring Using MIMO Millimeter-Wave Radar
IEEE Journal of Biomedical and Health Informatics
|July 21, 2026
Summary
This study introduces a contactless radar system (MM-FGHM) for detailed heartbeat waveform reconstruction and cardiac metric estimation. The system demonstrates high accuracy and robustness for non-contact health monitoring.
Area of Science:
- Biomedical Engineering
- Signal Processing
- Health Informatics
Background:
- Radar systems are increasingly used for health monitoring due to accessible technology and advanced algorithms.
- Accurate reconstruction of fine-grained heartbeat waveforms is crucial for comprehensive cardiac activity analysis, surpassing single heart rate values.
- Effective heartbeat waveform reconstruction relies on optimal signal preprocessing and network architecture selection.
Purpose of the Study:
- To introduce MM-FGHM, a novel contactless radar-based system for fine-grained heartbeat waveform reconstruction and cardiac metric estimation.
- To develop and validate an advanced deep learning architecture for extracting detailed cardiac features from radar signals.
- To enhance the precision of heartbeat reconstruction through a specifically designed joint loss function.
Main Methods:
- Development of MM-FGHM, a contactless radar-based health monitoring system.
- Design of ResED-Net, a dual-stream network integrating ResNet with an encoder-decoder architecture for feature extraction from 3D Range-Angle-Time matrices.
- Implementation of a joint loss function to improve reconstruction accuracy.
Main Results:
- MM-FGHM successfully achieved high-accuracy fine-grained heartbeat waveform reconstruction and cardiac metric estimation.
- The ResED-Net architecture effectively extracted cardiac features from both real and imaginary parts of radar data.
- Experiments involving 16 users across diverse configurations confirmed the system's robustness and generalization capabilities.
Conclusions:
- MM-FGHM offers a promising solution for reliable non-contact health monitoring through advanced radar signal processing.
- The proposed ResED-Net and joint loss function significantly contribute to precise heartbeat waveform reconstruction.
- The system's validated performance highlights its potential for widespread application in remote cardiac health assessment.
