Detection and Automatic Deletion of Staples in Images of Wound of Abdominal Surgery for m-Health Applications

Abstract

Patients undergoing abdominal operations have to perform a post-operative follow-up during long periods of time after surgery. These face-to-face consultations have an economic cost and affect patients’ quality of life. To simplify the follow-up of these patients, a smartphone application could be used. Based on an image of their surgical wound, post-operative complications may be detected. There are two cues to detect wound infection: Inflammation and reddening. In this work, we aim at detecting and segmenting the wound. The first step of this process consists of locating and deleting the staples that maintain the wound closed because they distort the overall color of the image. To do this, we applied the discrete fuzzy morphological dilation operation to the whole image. Next, we applied an inpainting algorithm to restore the regions detected as staples. Finally, using fuzzy sets we determine if there are areas of the wound that present a deviation with respect to the red color, and study if the wound is reddening.