High Accuracy Heartbeat Detection from CW-Doppler Radar Using Singular Value Decomposition and Matched Filter.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Remote Sensing

Background:

• Continuous wave Doppler radar sensors (CW-DRS) offer non-contact vital signs monitoring, crucial for home healthcare.
• Extracting cardiac signals from CW-DRS data is challenging due to superimposed respiratory and body motion signals.
• Existing methods struggle to isolate precise cardiac information from complex radar measurements.

Purpose of the Study:

• To develop a novel signal processing method for accurate heart rate extraction using CW-DRS.
• To address the challenge of separating cardiac signals from superimposed motion artifacts.
• To validate the proposed method across different subject postures.

Main Methods:

• A novel matched filter approach was developed.
• The method involves adaptive template generation using singular value decomposition (SVD) of a trajectory matrix.
• Signal reconstruction is achieved through convolution of the generated template with measurement signals.

Main Results:

• The proposed method demonstrated effectiveness in extracting heart rate from CW-DRS signals.
• For supine subjects, absolute errors in heart rate and standard deviation of heartbeat interval were 1.93±1.76 bpm and 57.0±28.1 s, respectively.
• For seated subjects, errors were 9.72±7.86 bpm and 81.3±24.3 s, indicating good performance across postures.

Conclusions:

• The novel matched filter method successfully extracts cardiac information from CW-DRS signals, even with superimposed motion.
• This technique enhances the feasibility of non-contact vital signs monitoring in home healthcare settings.
• The validated method offers improved accuracy for heart rate and heartbeat interval measurements in various subject positions.