Evaluation of Machine Learning Algorithms for Classification of Psoriasis and Lichen Planus Cutaneous Diseases

The tendency of skin diseases to manifest in a unique and yet similar appearance, absence of enough competent dermatologists, and urgency of diagnosis and classification on time and accurately, makes the need of machine aided diagnosis blatant. This study is conducted with the purpose of broadening the research in skin disease diagnosis with computer by traversing the capabilities of Machine Learning algorithms to classify two skin diseases noticeably close in appearance, Psoriasis and Lichen Planus. The total of 8 state of the art classification Machine Learning algorithms have been trained on the dataset provided by clinical research and development center, Semnan University of Medical Sciences and data provided by Dermnet. The results varied between the minimum of 48% for Kernel Approximation and maximum of 99% accuracy for Random Forest algorithm; other metrics such as precision, recall and f-1 score are also included in the study for each algorithm and each class. With their promising results, these algorithms make the potential of machine aided diagnosis clear. Machine Learning algorithms could provide assistance to physicians and dermatologists by classification of skin diseases in real-time.