Liver Fibrosis: Intelligent Analysis of Risk Factors
Bouharati Khaoula *
Department of Epidemiology, Faculty of Medicine, Constantine University, Algeria and Laboratory of Health and Environment, UFAS Setif1 University, Setif, Algeria.
Bouharati Imene
Faculty of Medicine, Paris Sorbonne University, France and Laboratory of Intelligent Systems, UFAS Setif1 University, Setif, Algeria.
Guenifi Wahiba
Faculty of Medicine, Setif University Hospital, UFAS Setif1 University, Setif, Algeria.
Gasmi Abdelkader
Faculty of Medicine, Setif University Hospital, UFAS Setif1 University, Setif, Algeria.
Boucenna Nassim
Faculty of Medicine, Setif University Hospital, UFAS Setif1 University, Setif, Algeria.
Laouamri Slimane
Faculty of Medicine, Setif University Hospital, UFAS Setif1 University, Setif, Algeria.
*Author to whom correspondence should be addressed.
Abstract
Background: Hepatic fibrosis is excessive scarring resulting from the buildup of connective tissue in the liver. The extracellular matrix is produced in excess and / or insufficiently degraded. Fibrosis by itself is asymptomatic but can lead to portal hypertension (fibrosis deflects intrahepatic blood flow) or cirrhosis (fibrosis destroys normal hepatic architecture and induces hepatic dysfunction). Fibrosis can sometimes develop without being linked to a known risk factor. The most common causes of liver fibrosis are hepatitis B and C and alcohol abuse.
Objectives: The objective of this study is the identification of risk factors and their effects. As the set of factors are often unknown and even those that are, their exact weights are ignored. The proposed fuzzy analysis makes it possible to deal with these inaccuracies.
Methods: As the system is very complex to analyze its factors using classical mathematical tools, this study proposes an intelligent analysis in the data processing. A fuzzy inference system is proposed. Risk factors are considered fuzzy variables and therefore uncertain. The proposed analysis system has fuzzy inputs representing risk factors, an output that expresses the degree of certainty of fibrosis affection. A database is created from the real values of cases diagnosed at the level of our hospital service. Inputs are linked to output via inference rules of the form [IF… THEN].
Conclusion: Once the system is established, this will allow variables to be introduced randomly at the inputs to automatically read the output result calculated by the aggregation of all the rules combined and therefore can predict the attack by this disease.
Keywords: Sansevieria suffruticossa, Hepatic fibrosis, aerial,, risk factors, terrestrial, intelligent modeling, dimorphic, fuzzy logic, acropetal,, exodermis, unistratose, scattered, collateral,, cencentric