Frame Registration for Motion Compensation in Imaging Photoplethysmography

Area of Science:

• Biomedical Engineering
• Optical Sensing
• Physiological Monitoring

Background:

• Imaging photoplethysmography (iPPG) assesses microcirculation and cardiovascular signs using optical sensors.
• Motion artifacts are a significant challenge in iPPG signal acquisition.
• Developing robust motion compensation techniques is crucial for accurate iPPG analysis.

Purpose of the Study:

• To evaluate the effectiveness of silicone palm casts for pre-testing iPPG motion compensation algorithms.
• To establish a performance baseline for image registration schemes.
• To isolate motion-induced noise from physiological signals in iPPG.

Main Methods:

• Utilized a silicone cast of a human palm to simulate live tissue for iPPG analysis.
• Developed a multi-stage evaluation model to assess factors like ROI, resolution, distance, and normalization.
• Applied image registration algorithms, specifically an intensity-based method with a moving reference frame.

Main Results:

• Image registration improved signal-to-noise ratio (SNR) by up to 75% (from 4.75 to 8.34).
• The silicone cast method provided a reliable baseline for algorithm evaluation.
• Identified key parameters influencing iPPG waveform extraction.

Conclusions:

• Image registration is effective for motion compensation in iPPG.
• Silicone casts offer a viable method for pre-clinical evaluation of iPPG algorithms.
• Optimized iPPG analysis can lead to improved cardiovascular monitoring.