A Bayesian Belief Network approach for predicting kernicterus
A lot of research have been conducted using expert systems in the diagnosis of neonatal jaundice but none has been conducted on kernicterus. Kernicterus is a complication of neonatal jaundice in which bilirubin accumulates in the grey matter of the brain, causing an irreversible neurological damage. In this paper, a Bayesian belief network was designed for predicting neonatal jaundice. The BBN model has 15 nodes and had 97% and 94% accuracy in classifying neonatal jaundice and kernicterus respectively.
Keywords: Jaundice, Kernicterus, Bayesian Belief Network