Identification of Diabetic Foot Ulcer in Images using Machine Learning

It is a known fact that foot ulcers are common complications of patients with uncontrolled diabetes. If the ulcer is not detected early and treated aggressively, it might progress to a fulminating one and the limb might have to be amputated. Such complications can be avoided by early diagnosis and prompt treatment. There are several classifications of diabetic foot ulcers that would facilitate the treatment. In the current study deep learning frameworks like inception-v3, VGG-16, and VGG-19 were trained on the datasets images containing details of the ulcerated and non-ulcerated feet of the patient. The features that were extracted were provided as the input to the machine learning algorithms for studying the accuracy of the various classifiers. The inception v3 model and the SVM classifier achieve an accuracy of 99.8% as compared to other models.