A Deep Learning Model that Diagnosis Skin Diseases and Recommends Medication

Abstract

Skin diseases are very common and the diagnosis is tricky and challenging. Latest research in the field of medicine along with the help of advanced technology has proved to be quite useful not only in diagnosis but also for treatment. Application of deep learning methods for diagnosis of skin diseases has given remarkable results. Computer aided results are quick and provide a quick overview of the disease.This paper aims to diagnose the skin disease from the infected skin image captured and provide details of the disease and recommend medication. To achieve this we have used stateof-the-art convolutional neural network(CNN) architecture MobileNetV2 for model building and training.This is particularly helpful for both doctors and individuals to analyze the disease. For doctors, they can validate their opinion with this prediction and individuals can have an idea of the disease at the beginning itself and can be helpful to prevent the disease further since prevention is always better than cure.