Journal of Neonatal Biology

Prediction of Cyanotic and Acyanotic Congenital Heart Disease using Machine Learning Methods

International Conference and Expo on Neonatology & Perinatology - October 21, 2022 | Webinar

October 21, 2022 | Webinar

Haris Khurram

Assistant Professor, Pakistan

Scientific Tracks Abstracts: J Neonatal Biol

Abstract:

Congenital heart disease is the most common in neonates and it is a major cause of pediatric illness and childhood morbidity and mortality. The main purpose of this study was to predict the best model for predicting cyanotic and acyanotic congenital heart disease in children during pregnancy and identify their risk factors. The data were collected from the Pediatric Cardiology Department at Chaudhry Pervaiz Elahi Institute of Cardiology Multan, Pakistan. A sample of 3900 mothers whose children were diagnosed with cyanotic or acyanotic congenital heart disease was taken. Models were compared using the area under curve, sensitivity, and specificity. There are 53.6% of males and 46.4% of females having acyanotic congenital heart disease and 54.5% of male’s ads 45.5% of females have cyanotic congenital heart disease. For the artificial neural network model, the area under the curve, sensitivity, and specificity are 90.12, 65.76, and 97.23 respectively. The results of the area under the curve show that the best fit model for our data is the artificial neural network. Children with a positive family history are very high at risk of being cyanotic and acyanotic congenital heart disease. Importance of Research: The research identified the parental related factor which increases the incidence of cyanotic and acynotic congenital heart disease in children. And provide a methodology that can be used to make an expert system which can predict the type of congenital heart disease in children before its birth. Which help medical practitioners and scientist to prevent the neonate and take care at earlier stages. Keywords: Congenital Heart Disease, Cyanotic heart disease, Acyanotic heart disease, Logistic Regression model, Artificial Neural network.

Biography :

Haris Khurram has finished his PhD in Statistics in the area of Bayesian Non-Parametric Modelling from Bahauddin Zakariya University. Earlier he has secured M.Phil. in Statistics in the area of Graphical Estimation of the parameters of Continuous Family of Distributions. Bahauddin Zakariya University, Multan was his alma mater. He is an Assistant Professor of Statistics in National University of Computer and Emerging Sciences. He has more than 25 research papers in reputed journals and various conference papers at national and international conferences. He is the reviewer of various prestigious journals and also serves as an editorial board member of a few journals.