Non-Invasive Assessment Model of Liver Disease Severity by Serum Markers Using Cloud Computing and Internet of Things

Abstract

Information on the stage of liver fibrosis is essential for decisions on antiviral treatment for chronic hepatitis B virus (HBV). This paper aims to establish a non-invasive assessment model with serum markers using cloud computing and the Internet of Things for the evaluation of liver disease severity and its prognosis. Based on the Internet of Things, the multiple and key information system of liver fibrosis or cirrhosis are constructed using the serum markers data. In the cloud platform, the probability density functions of indexes are used to select the optimized indicators. The logistic regression is used to establish the non-invasive assessment model. The patients were selected with CHB and underwent liver biopsy in the Second Xiangya Hospital, Central South University. There are two inclusion criteria: first, the patient received a liver biopsy according to “Proclaim Prevention and Cure Guide For Chronic Hepatitis B”of Chinese Medical Association in 2015; second, the patient has a history of hepatitis B or HBV surface antigen (HBsAg) positive more than six months, and HBsAg and (or) HBV DNA is still positive. Results of clinical data applications show that the accuracy of the non-invasive assessment model reaches greater than 70% for the recognition of significant liver fibrosis. In addition, the discriminant accuracy can be improved by increasing the number of indicators. The established non-invasive assessment model can be used for auxiliary clinical diagnosis after the further validation.

Keywords

• Non-invasive assessment
• cloud computing
• serum markers
• liver disease severity