Measurement of microvascular perfusion by remote photoplethysmography applied to wound monitoring – Wound-rPPG

Chronic wounds are a major health problem, affecting millions of patients in Europe and resulting in substantial costs to healthcare systems. Traditional methods of wound monitoring, based on subjective visual observations and manual measurements, are often imprecise and painful for patients. To address these issues, this project proposes the use of remote photoplethysmography (rPPG) for spatially resolved assessment of microvascular perfusion in wounds. This non-invasive optical technique captures variations in light intensity resulting from interactions with the human body, offering an innovative approach to wound assessment with the potential for widespread adoption.

The project focuses on developing a novel deep-learning-based methodology for spatially resolved rPPG imaging from video recordings of tissues. It aims to overcome two key challenges: refining our understanding of light-tissue interactions in rPPG and extending deep learning's success to spatially resolved rPPG contexts.

The project comprises three scientific work packages. The first investigates light-tissue interactions to improve our understanding of temporal modulations of reflected light observed in rPPG. This will allow us to untangle the contributions of blood absorption and elastic deformations of the capillary bed on the observed modulations. The second uses deep learning, synthetic video data, and weakly supervised paradigms to estimate rPPG maps and indicators linked to microvascular perfusion. The third validates the technology for wound monitoring in real-world conditions, ensuring practicality and usefulness.

This project resides at the intersection of cutting-edge artificial intelligence, video analysis, and healthcare, offering the potential to revolutionize wound care and management.