Detection of Sore-risk Regions on the Foot Sole with Digital Image Processing and Passive Thermography in Diabetic Patients

According to the International Diabetes Federation, approximately four million people die from diabetes each year, and ten million more suffer disabling or potentially terminal complications, such as diabetic foot disease. Therefore, the proposal is the early detection by detecting non-homogeneous temperature regions with radiometric data and image processing by IR thermography. The samples were taken from 12 patients diagnosed with diabetes mellitus without a history of diabetic foot or visible anatomical alterations, by a cooled IR sensor programmed exclusively to measure human skin temperature. The radiometric data was processed to eliminate interferences and segment the lower limb since the acquisition was carried out in a non-controlled environment; where lies the relevance of this paper. The results were presented as binary images with a defined segmented region against a contrasting background, reviling areas corresponded to higher temperatures regard the rest of the sole fo the foot as a consequence of non-homogeneous temperature pattern. A total of 6 cases showed a normal homogeneous temperature, 2 cases had fault origins and 4 cases presented spotted regions that are prospective to prone lesions. This procedure set a precedent for the analysis of early detection of the diabetic foot with radiometric data processing instead of using IR images as input; which could be used to retrieve the coordinates and carry out an exhaustive numerical analysis or pattern recognition (i.e. thermal fingerprint) to support the anticipated diagnosis based on support systems such as medical thermography as perspective.