15 - Machine learning techniques in healthcare informatics: Showcasing prediction of type 2 diabetes mellitus disease using lifestyle data

Abstract

Machine learning (ML) is an emerging area of research in the healthcare industry. The healthcare data are being generated from multiple sources along with different formats. Integrating the health data and then bringing it into the common platform for further analysis require advanced tools and techniques to generate valuable information. The healthcare professions are not able to achieve valuable knowledge for actionable clinical intelligence because of the heterogeneity, inconsistency, incompleteness, etc. of health data. Technological advancements like data mining and machine learning can be used for the same. This study discusses the role of the machine learning paradigm in healthcare analytics. The chapter also presents and implements the framework for developing machine learning models for type 2 diabetes mellitus (T2DM) disease. In this chapter, lifestyle indicators rather than clinical/pathological parameters have been used for the prediction of type 2 diabetes mellitus. The current study has involved different experts like diabetologists, endocrinologists, dieticians, nutritionists, etc. for selecting the contributing lifestyle parameters to promote health and manage diabetes. As such the study aims to develop an intelligent knowledge-based system for prediction of T2DM without the conduct of clinical tests. It can save the patient suffering from undue delays caused by unnecessary readmissions and pathological tests in hospitals. The proposed work emphasizes the use of machine learning techniques namely K-Nearest Neighbor (KNN), Logistic Regression (LR), Naïve Bayes (NB), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), and Artificial Neural Network (ANN) for prediction of T2DM disease. The RF attained highest accuracy of 93.56% followed by DT, LR, SVM, NB, ANN, and KNN as 92.70%, 91.41%, 90.98%, 89.27%, 87.98%, and 84.54%, respectively. The other statistical performance measures were calculated in terms of Precision, Recall, Specificity, F1-score, Misclassification Rate, Receiver Operating Curve, etc.