Diabetes Foot Ulcer Detection Using Inception V3 Deep Learning Technique

Diabetes problems like Diabetic Foot Ulcers (DFU) can lead to frequent hospitalizations and costly treatments. Ulcers in the feet caused by diabetes are extremely prevalent. Mellitus adverse effect, and if left untreated, they can lead to amputations. Diabetic feet impact 15-25% of diabetic individuals worldwide. People who suffer from diabetic foot ulcers (DFU) have a more than twice higher death risk than non-ulcerated diabetic individuals. The present investigation offers a new particle swarm optimization approach that combines deep learning to identify and classify diabetic foot ulcers. The described technique's purpose is to recognize and categorize DFU. To do this, the suggested approach generates feature vectors using the Inception V3, Squeeze net, VGG-16, and VGG-19 model. Among these four techniques Inception v3 technique attains highest accuracy. The particle swarm optimization approach optimizes hyper parameter selection for the Inception-v3 model, improving DFU classification results by eliminating the need for human trial-and-error tweaking. The Inception v3 model works well, and our approach outperforms it with an accuracy of 93.3%. DFU is classified utilizing the Deep stacked sparse auto encoder method. This method outperforms other deep learning algorithms, according to thorough experimental results.