Continuous Radar-Based Heart Rate Monitoring using Autocorrelation-Based Algorithm in Intensive Care Unit

Area of Science:

• Biomedical Engineering
• Medical Devices
• Signal Processing

Background:

• Continuous vital sign monitoring is crucial in Intensive Care Units (ICUs).
• Existing methods for heart rate monitoring can be invasive or prone to artifacts in critical care settings.
• Non-invasive, unobtrusive monitoring solutions are needed to improve patient care and reduce clinical burden.

Purpose of the Study:

• To develop and validate a radar-based algorithm for accurate, non-invasive heart rate monitoring in ICUs.
• To assess the performance of a 140 GHz Frequency-Modulated Continuous Wave (FMCW) radar system placed beneath hospital beds.
• To evaluate the algorithm's robustness against noise and motion artifacts common in ICU environments.

Main Methods:

• Utilized a 140 GHz FMCW radar system for data acquisition from 15 post-operative cardiac patients.
• Implemented a multi-stage algorithm involving data preprocessing, channel selection, heart rate estimation using autocorrelation, and post-processing.
• Validated radar-derived heart rates against electrocardiogram (ECG) ground truth data.

Main Results:

• Achieved a Mean Absolute Error (MAE) of 2.22 beats per minute (bpm) for heart rate estimation.
• Demonstrated overall system coverage of 66%, increasing to 98% during periods of sinus rhythm.
• Successfully mitigated noise and motion artifacts, ensuring reliable vital sign monitoring.

Conclusions:

• The developed radar-based algorithm offers a robust and accurate method for continuous, non-invasive heart rate monitoring in ICUs.
• This technology has significant potential to enhance patient care by providing reliable vital sign data in challenging clinical settings.
• The study validates the effectiveness of FMCW radar and the proposed algorithm in real-world hospital environments.