Journal of Medical Systems

, Volume 35, Issue 1, pp 121–126

A Novel Method for Diagnosing Cirrhosis in Patients with Chronic Hepatitis B: Artificial Neural Network Approach

Authors

  • Mohammad Reza Raoufy
    • Department of Physiology, School of Medical SciencesTarbiat Modares University
  • Parviz Vahdani
    • Department of Infectious DiseasesLogman Hospital
  • Seyed Moayed Alavian
    • Baqiyatallah Research Center for Gastroenterology and Liver DiseaseBaqiyatallah University of Medical Sciences
  • Sahba Fekri
    • Faculty of MedicineShahid Beheshti University of Medical Sciences
  • Parivash Eftekhari
    • Department of Physiology, School of Medical SciencesTarbiat Modares University
    • Neuromuscular Systems Laboratory, Faculty of Biomedical EngineeringAmirkabir University of Technology
Original Paper

DOI: 10.1007/s10916-009-9348-8

Cite this article as:
Raoufy, M.R., Vahdani, P., Alavian, S.M. et al. J Med Syst (2011) 35: 121. doi:10.1007/s10916-009-9348-8

Abstract

We designed an artificial neural network (ANN) to diagnose cirrhosis in patients with chronic HBV infection. Routine laboratory data (PT, INR, platelet count, direct bilirubin, AST/ALT, AST/PLT) and age were collected from 144 patients. Cirrhosis in these patients was diagnosed by liver biopsy. The ANN’s ability was assessed using receiver-operating characteristic (ROC) analysis and the results were compared with a logistic regression model. Our results indicate that the neural network analysis is likely to provide a non-invasive, accurate test for diagnosing cirrhosis in chronic HBV-infected patients, only based on routine laboratory data.

Keywords

CirrhosisChronic hepatitis BArtificial neural networkLogistic regressionLaboratory data

Copyright information

© Springer Science+Business Media, LLC 2009