Abstract:
Hepatitis is basically an inflammation of the liver. There many forms of hepatitis virus. Among
them, hepatitis B, as well as hepatitis C viruses, lead to liver cancer. Hepatitis C virus (HCV)
is the notorious toxicant of the liver. HCV harms the liver upon a time. There reside mainly four
stages of liver fibrosis by which the liver gets damaged. Those stages are portal fibrosis,
portal fibrosis with few septa, numerous septa without cirrhosis, cirrhosis. At the first stage,
the liver gets inflamed. At the time of fibrosis, defected tissue starts replacing cured tissue
in the inflamed liver. When fibrosis of the liver begins various scarring starts building up and
make it tough for the liver to work thoroughly. HCV includes some symptoms like fever,
nausea or vomiting, headache, diarrhea, fatigue, jaundice, epigastric pain. Nowadays, the
number of HCV infectious individuals is growing globally which has become a matter of
concern. In order to save lives, researchers have over the years worked to discover an
alternative diagnostic means for hepatitis disease using computing intelligence. An early
diagnose as well as prediction of liver disease like hepatitis is quite beneficial. Computing
intelligence is playing a significant role in the realm of healthcare nowadays. Our
objective is to compute the ratio of stages of liver fibrosis and the symptoms of the hepatitis C
virus. In this paper, with the help of data mining, the stages of fibrosis have been classified.
Three significant algorithms Artificial Neural Network (ANN) Naive Bayes (NB) as well
as J48 were used in this paper for the stage prediction of fibrosis of the liver. All of those three
efficient algorithms performed quite well but among them, the J48 decision tree algorithm gave
a paramount performance with an accuracy of 96.03%. By Utilizing the J48 algorithm it is
possible to classify the different stages of liver fibrosis sophisticatedly as well as to develop
an expert system in future work. So J48 is considered the most convenient algorithm for
the classification of the stages of liver fibrosis caused by the hepatitis C virus as well as their early
diagnosis. This data mining process will also help to avoid the hazard of invasive procedures
for the diagnosis of liver fibrosis.
