I-Corps: Non-invasive Camera-based Blood Perfusion Imaging

William Marsh Rice University, Houston TX

Investigators

Abstract

The broader impact/commercial potential of this I-Corps project is to evaluate the commercial potential of a new imaging modality that can reliably measure blood perfusion. The system will enable surgeons to visualize and analyze blood flow at a surgical site in real-time, thereby assisting them in making better clinical decisions in the operating room, reduce the rate of post-operative complications, leading to overall reduction in healthcare costs. The current state-of-the-art for blood perfusion imaging is either based on injecting a contrast agent into the bloodstream and are therefore invasive with potential side-effects, or based on sophisticated Laser-based imaging device which can be difficult to operate during surgery. The system developed here relies only on the video recordings of the observation site to reliably estimate blood perfusion, and hence is both non-invasive and is easy-to-use. Beyond surgery, there is a potential of using the system to monitor critical care patients and also as a point-of-care device in podiatry, especially for the early detection of diabetic neuropathy. This I-Corps project further develops a non-contact, non-invasive, camera-based and clinically accurate blood perfusion measurement system. Using low-cost, mass-produced hardware, combined with sophisticated signal processing algorithms, the system can estimate blood perfusion maps in real time. The key innovation lies in: (i) data fusion from multiple sensors - camera and pulse oximeter, (ii) a new computational imaging and signal processing algorithm to recover blood perfusion below the skin, from videos of the skin surface, and (iii) a novel approach to compensate motion of the imaged tissue surface. The technology can work not only with video recordings of the external tissue, such as skin surface, but also internal tissue and organs, such as during a minimally invasive surgery with a laparoscope. The system has been successfully demonstrated in detecting small changes in blood perfusion during partial and full occlusion of the blood vessels.