Modification of an Ontology to Incorporate Clinical Process of Cardio Vascular Disease Detection with a Case Study of Automated Prediction of Cardiomegaly Using Machine Learning Models

Healthcare is providing us with an ocean of data and therefore an ocean of opportunities. Electronic health data generation is growing with a maximum growth rate in comparison to other domains of electronic data. Death and physical disability due to Cardio Vascular Diseases are common nowadays irrespective of the obvious medical causes and medical parameters. Many research works have been done in this domain but the data set does not contain the needed medical parameters or the health data signals in most of these works. In this paper, we study general machine learning and deep learning works in this domain and also the research works that deal with actual medical parameters and signal. An existing ontology is enhanced to accommodate clinical process needed including types of investigations for detection of cardiovascular diseases. Finally, Cardiomegaly has been detected from a data set obtained from Chest X-Ray PA View using Neural Network, SVM, XGBoost, Logistic Regression. The objective of this research is to find out ways of improving the existing healthcare services using the power of Artificial Intelligence considering the scarcity of needed healthcare services, infrastructure and experts in many places and at many times.