Fibrosis and Inflammatory Activity Analysis of Chronic Hepatitis C Based on Extreme Learning Machine

The diagnosis of fibrosis stage and inflammatory activity level in patients with chronic hepatitis C is important in clinical practices. To provide a non-invasive diagnosis for patients with chronic hepatitis C, this study proposed an automatic diagnosis system of chronic hepatitis C using serum indices data of patients to predict the fibrosis stage and inflammatory activity grade of chronic hepatitis C by training the extreme learning machine. Due to the superiority of extreme learning machine such as simple structure and fast calculation speed, the presented automatic diagnosis system can achieve good diagnosis performance. The proposed automatic diagnosis system is test on real clinical cases of chronic hepatitis C based on serum indices. Experimental results demonstrate that the performance of the proposed method exceeds that of the state-of- the-art baselines regarding the diagnosis of fibrosis stage and inflammatory activity grade of chronic hepatitis C.