Proposed System for Remote Detection of Skin Diseases Using Artificial Intelligence

Abstract

Skin diseases are prevalent diseases with visible symptoms and affect around 900 million of people in the world at any time. More than a half of the population is affected by it at an indefinite time. Dermatology is uncertain, unfortunate and strenuous to diagnose due to its complications. In the dermatology field, many times thorough testing is carried out to decide or detect the skin condition the patient may be facing. This may vary over time on practitioner to practitioner. This is also based on the person’s experience too. Hence, there is a need for an automated system which can help a patient to diagnose skin diseases without any of these constraints. We propose an image based automated system for recognition of skin diseases using Artificial intelligence. This system will make use of different techniques to analyze and process the image data based on various features of the images. Since skin diseases have visible symptoms, we can use images to identify those diseases. Unwanted noise is filtered and the resulting image is processed for enhancing the image. Complex techniques are used for feature extraction such as Convolutional Neural Network (CNN) followed by classifying the image based on the algorithm of softmax classifier. Diagnosis report is generated as an output. This system will give more accurate results and will generate them faster than the traditional method, making this application more efficient and dependable. This application can also be used as a real time teaching tool for medical students in the dermatology domain.