Development a Hybrid Model based Deep Learning to Diagnosis of Autoimmune Diseases

Abstract

Systemic Lupus Erythematosus (SLE) is a chronic autoimmune disease that can cause pimples and butterfly-shaped hives on the skin of the nose and cheeks, and if left untreated, can spread to the entire body. Apart from damaging the skin, lupus can cause inflammation or damage to the joints, muscles, inner membrane or around the lung, heart, kidney and brain. Considering the importance of this disease and the early diagnosis of ANF, in this study, an approach based on deep learning and gray wolf meta-heuristic algorithm was discussed in order to identify this disease among patients. In this study is from Taiwan Precision Medicine Initiative (TPMI), which was used by 946 patients with their own diseases in the period of June 2019 to June 2020. The results showed that the use of the proposed method can increase the detection accuracy to 0.8936. Meanwhile, for LR, RF, SVM, LGBM, GBT, and XGB models, the detection accuracy is 0.7887, 0.8345, 0.7729, 0.7993, 0.7975, and 0.8345, respectively.