Escâner tridimensional para medição de volume de feridas

Abstract

Treatment and follow-up of Diabetic Foot ulcers is a major public health issue. Current techniques used for following up the healing process of open ulcers are mostly based on approximate area measurements of the wound and involve direct contact, which poses risks of contamination for the patient. This work proposes the use of computer vision techniques to measure not just the area, but the volume of open ulcers, providing additional objective information about the healing process to physicians while avoiding direct contact with the wound during the measurement procedure. The proposed technique involves acquiring a sequence of images of the wound with structured illumination using a low-power laser beam line, followed by depth reconstruction from the geometry of the acquisition device. The wound region is identified in a semi-automatic way and a fictitious surface of what would be the normal skin surface is then used to compute the volume, which corresponds to the missing amount of tissue. Controlled ex-vivo experiments using a rubber model of a Diabetic Foot and a knuckle of pork with an artificial wound representing the human skin demonstrate the viability of the proposed technique.