Implementation of the SAM algorithm for highlighting lesions in diabetic foot

Abstract

Diabetic foot is a chronic complication of diabetes mellitus that can occur when blood glucose levels are not properly controlled. This condition damages the nerves and/or blood vessels in the feet, and if not properly cared for, it can lead to injuries, infections, and, in severe cases, limb amputation. It is estimated that 71% to 85% of diabetic patients with foot ulcers require amputations. Therefore, it is crucial to detect the development of foot ulcers to improve the quality of life for patients with diabetic foot.

In this work, we propose the implementation of the SAM segmentation algorithm to remove the background from photographs and highlight the lesions present in the foot. The aim is to facilitate the identification of these lesions in healthcare institutions where specialists may not be available.

It is important to note that, although the use of segmentation algorithms can be helpful as a support tool in detecting foot lesions in diabetic patients, it is essential to validate the results with specialized healthcare professionals. Clinical evaluation and accurate diagnosis should be conducted by physicians or specialists trained in the care of diabetic feet.