Early Disease Detection Through Nail Image Processing Based on Ensemble of Classifier Models

Abstract

Medical science has progressed in many ways and different methods have been developed for the diagnosis of diseases in the human body and one of the ways to identify the diseases is through the close examination of nails of the human palm. The main aim of this study is to compare the performance of various classifier models that are used for the prediction of various diseases. The Performance analysis is done by applying image processing, different data mining and machine learning techniques to the extracted nail image through our proposed system which does nail analysis using a combination of 13 features (Nail Color, Shape and Texture) extracted from the nail image. In this paper we have compared different machine learning classifiers like Support Vector Machine, Multiclass SVM and K-Nearest Neighbor through ensemble of these classifiers with different features so as to classify patients with different diseases like Psoriasis, Red Lunula, Beau�s Lines, Clubbing, etc. These approaches were tested with data images from Hospitals and workplaces. The performance of the different classifiers have been measured in terms of Accuracy, Sensitivity and Specificity.