Computer-Aided Diagnosis of Systemic Lupus Erythematosus Using Immunological Profile

Abstract

Systemic Lupus Erythematosus (SLE) or Lupus is a prototype multisystem autoimmune connective tissue disorder with variable prognosis and remitting and relapsing course. As the cause of lupus is still unknown, diagnosing the lupus is very difficult for the physicians. Machine learning plays a very important role for assisting the doctors in diagnosing the diseases very accurately. The dataset used for this work is collected from various hospitals in Tamilnadu. Immunological profiles of 400 patients (200 SLE and 200 Non-SLE) are taken into this experimental study. Various classification techniques are used for classification of data set into SLE or Non-SLE. Experimental results proved that the J48 classifier model gives higher accuracy than other models for the identification of SLE.