THE PREDICTION OF HEPATITIS B VIRUS (HBV) USING ARTIFICIAL NEURAL NETWORK (ANN) AND GENETIC ALGORITHM (GA)

Authors

  • Douglas Ibrahim Department of Computer Science, Adamawa State College of Agriculture, P.M.B. 2088 Ganye, Adamawa State, Nigeria
  • Asabe Sandra Ahmadu Department of Computer Science, Modibbo Adama University, Yola, P.M.B. 2076 Yola, Adamawa State, Nigeria.
  • Yusuf Musa Malgwi Department of Computer Science, Modibbo Adama University, Yola, P.M.B. 2076 Yola, Adamawa State, Nigeria.
  • Bamanga Mahmud Ahmad Department of Computer Science, Federal University of Lafiya, Nasarawa State, Nigeria

DOI:

https://doi.org/10.51594/csitrj.v2i1.275

Abstract

The hepatitis B virus causes a liver infection called hepatitis B (HBV). It might be severe and go away on its own. Some kinds, however, can be persistent, leading to cirrhosis and liver cancer. HBV can be transmitted to others without the individual being aware of it; some persons have no symptoms, while others only have the first infection, which later resolves. Others develop a chronic illness as a result of their condition. In chronic cases, the virus attacks the liver for an extended period of time without being detected, causing irreparable liver damage. The manual approach has a high number of errors due to human decision-making, and visual screening is time-consuming, tiresome, and costly in terms of manpower. To predict the occurrence of Hepatitis virus (HBV), this research project thesis suggested an algorithm; Artificial Neural Network (ANN), and genetic algorithm (GA). To develop, evaluate and validate the performance of the model developed using ANN. Medical records of nine hundred patients were collected in the Northern Senatorial District (Mubi South), Central Senatorial District (Hong), and Southern Senatorial District (Ganye) regions of Adamawa state, Nigeria. Three hundred (300) patient records were collected from each general hospital, for a total of 900 patient records. The success of the proposed technique is demonstrated when ANN is paired with GA, Accuracy (66.30%), Specificity (66.33%), and Sensitivity (77.53%) were discovered. In this study, hepatitis B virus (HBV) was predicted using Artificial Neural Network (ANN) classifier and Genetic algorithm optimization tool were used to select the features that are responsible for hepatitis B virus (Sex, Loss of Appetite, Nausea and vomiting, Yellowish skin and eye, Stomach pain, Pain in muscles and joint). The prediction was found to have acceptable performance measures which will reduce future incidence of the outbreak and aid timely response of medical experts.

Keywords: Hepatitis B Virus (HBV), Prediction, Features, Classification.

Published

2021-12-01

Issue

Section

Articles