Diagnosis of Ear Conditions Using Deep Learning Approach

The lack of medical expertise and facilities makes it difficult to diagnose the diseases and the misdiagnosis can severely affect the health of the patients. This problem can be overcome by automating the process of diagnosis using artificial intelligence algorithms like deep learning. It has been used a lot for developing methods for the diagnosis of diseases with help of medical images. The CNN performs quite well for the diagnosis of diseases using deep learning and has been used for the diagnosis of a lot of diseases. Ear diseases are one of the most common problems that are encountered. They can be easily treated with proper diagnosis and medication using deep learning. This work proposes a deep learning-based method employing CNN for the detection of ear infections. It is capable of classifying ear images belonging to four categories (Chronic otitis media, Earwax plug, Myringosclerosis and Normal) and obtained training and validation accuracies of 99.98% and 97.37% respectively.