Measurement of wound healing by image analysis

Chronic wound healing is a major health problem consuming more than $10 Billion/year. Noninvasive measurements of wounds are critical for evaluation of different therapies. In this paper, we present two image processing methods which can be used to compare the effects of different dressings on pig skin wounds. In the first part, the healing process is monitored by area measurements. Using thresholding the wound areas are calculated by pixel counting and corrected with perspective scaling. This technique showed good correlation with manual area measurements. In the second part, 3 D coordinates of the selected wound points are calculated using triangulation from a single colored image coordinates. Such a 3 dimensional image reconstruction from a single camera is made possible by projecting a structured light pattern onto the skin. In this study, a unique pattern consisting of colored dots is used. This color-coded pattern provides many advantages over widely used grid patterns: it is not necessary to see the whole image to recognize the points, since every dot can be identified correctly from its closest eight neighbors, eliminating occlusion problems. Our approach to structured light also mitigates the problem of motion in a subject, since all the coding information is contained in a single image.