Integrated optimisation method for personalised modelling and case studies for medical decision support

Abstract

Personalised modelling aims to create a unique computational diagnostic or prognostic model for an individual. The paper reports a new Integrated Method for Personalised Modelling (IMPM) that applies global optimisation of variables (features) and neighbourhood of appropriate data samples to create an accurate personalised model for an individual. The proposed IMPM allows for adaptation, monitoring and improvement of an individual's model. Three medical decision support problems are used as illustrations: cancer diagnosis and profiling; risk of disease evaluation based on whole genome SNPs data; chronic disease decision support. The method leads to improved accuracy and unique personalised profiling that could be used for personalised treatment and personalised drug design.