An embedded hardware implementation for semi-automated irregular wound area measurement

Companies and clinicians in the wound care market currently track wound care progress by hand through image analysis or area measurements with a ruler. This analysis is usually performed manually by physicians or nurses. The wounds under measurement do not commonly take the form of regular geometric shapes and thus are difficult to measure accurately manually. Therefore, the results vary widely depending on the level of measurement accuracy provided between each care provider the patient has while under treatment. The amount of variability is significant using this approach, and the need for a consistent, wound measurement technique is important in the growing wound care marketplace.;This research focuses on the implementation of a semi-automated boundary recognition technique for embedded system applications. Specifically, the study details wound images that are analyzed using common digital image processing techniques combined with a set of active contour boundary detection algorithms known as "snakes." A accurate and consistent area calculation can be performed by placing and locating a known object within the image and determining the wound size measurements from the boundary identified by the snake algorithm. The techniques were implemented with embedded systems that assure portability to foster usage within the hospital environment.