A New Approach using Deep Learning and Reinforcement Learning in HealthCare

Nowadays, skin cancer is one of the most important problems faced by world, due especially to rapid development cells and excessive exposure UV rays. Therefore, early detection at an stage employing advanced automated systems based on AI algorithms plays a major job in order effectively identifying detecting disease, reducing patient health financial burdens, stopping its spread skin. In this context, several approaches models have been presented throughout last few decades improve rate using dermoscopic images. This work proposed model that can help dermatologists know detect just seconds. combined merits two artificial intelligence algorithms: Deep Learning Reinforcement following great success we achieved classification recognition images medical sector. research included four main steps. Firstly, pre-processing techniques were applied accuracy, quality, consistency dataset. The input obtained from HAM10000 database. Then, watershed algorithm was used for segmentation process performed extract affected area. After that, deep convolutional neural network (CNN) utilized classify into seven types: actinic keratosis, basal cell carcinoma, benign dermatofibroma melanocytic nevi, melanoma vascular lesions. Finally, regards reinforcement learning part, Q_Learning train retrain our until found best result. accuracy metric evaluate efficacy performance method, which high 80%. Furthermore, experimental results demonstrate how be with tasks.