17 / 33
Volume 8
Journal of Gastrointestinal & Digestive System
ISSN: 2161-069X
Page 45
JOINT EVENT
Pediatric Gastro 2018
Digestive Diseases 2018
October 22-23, 2018
October 22-23, 2018 Berlin, Germany
3
rd
International Conference on
Digestive and Metabolic Diseases
Pediatric Gastroenterology Hepatology & Nutrition
13
th
International Conference on
&
Machine learning modeling applied to identification of liver fibrosis’ degree based on combination of
non-invasive methods
Leandro Augusto Ferreira
Brazil
A
bout 150 million people are carriers of the hepatitis C virus (HCV) in the world. About 25% are at risk of developing
cirrhosis; however, with HCV they are detected with some degree of liver fibrosis. To obtain the degree of liver fibrosis
with has been used the biopsy as gold standard. Although there are non-invasive methods such as the transient elastography-
FibroScan®, acoustic radiation force impulse (ARFI), enhanced liver fibrosis (ELF), the aspartate aminotransferase-to-platelet
ratio index (APRI), and the FIB-4 index, they can be influenced by some factors such as body weight and therefore disrupt the
results. Here, we intend to present some models of machine learning that combine results through these methods (FibroScan®,
ARFI, ELF, APRI and FIB-4) and other anthropometric variables to improve the accuracy of their particles in relation to a liver
biopsy. The data is from patients with hepatitis C from the clinical division of the Department of Gastroenterology, Hospital
das Clínicas, Faculty of Medicine, University of Sao Paulo, in Sao Paulo, Brazil.
ferreira.laf@gmail.comJ Gastrointest Dig Syst 2018, Volume 8
DOI: 10.4172/2161-069X-C7-083