Efficient Clinical Decision Making by Learning from Missing Clinical Data

Clinical decision making frequently involves making decisions under uncertainty because of missing key patient data (e.g. demographics, episodic and clinical diagnosis details) - this information is essential for modern clinical decision support systems to perform learning, inference and prediction operations. Machine learning and clinical informatics experts aim to reduce this clinical uncertainty by learning from the missing clinical attributes with a view to improve the overall decision making. These high-dimensional clinical datasets are often complex and carry multifaceted patterns of key missing clinical attributes. In this paper we highlight the problem of learning from incomplete real patient data acquired from Raigmore Hospital in Scotland, UK) from a statistical perspective-the likelihood-based approach to deal with this challenging issue. There are multiple benefits of our approach: to complement existing SVM (Support Vector Machine) techniques to deal with missing data within a statistical framework, and to illustrate a set of challenging statistical machine learning algorithms, derived from the likelihood-based framework that handles clustering, classification, and function approximation from missing/incomplete data in an intelligent and resourceful manner. Our work concentrates on the implementation of mixture modelling algorithms as well as utilising Expectation-Maximization techniques for the estimation of mixture components and for dealing with the missing clinical data of chest pain patients.