Please note: this event has passed
Title: Investigating medication taking behaviours in breast cancer survivors under complexity
Abstract: Sub-optimal medication-taking behaviours are common in breast cancer survivors prescribed adjuvant endocrine therapy (AET), increasing the risk of recurrence and mortality. The empirical literature has identified many variables underlying this phenomenon such as experiences of side effects, sociodemographic and clinical characteristics, and physiological and psychological components. However, complex associations between these variables have been identified which in turn makes it challenging to study and design behavioural interventions to support patients effectively.
In this seminar, we will discuss the utility of the theory of complex systems in explaining the interplay among various system components underlying medication adherence to AET in breast cancer survivors. Second, we will demonstrate how the Bayesian Network Analysis (BN) was applied to investigate the complexity of component interactions based on the joint probability of the variables. Our approach combined human expert experience, prior knowledge, and Bayesian network inference and learning to find the most effective network based on clinical trial data. Thus, we will complete the talk by discussing the opportunities and challenges of combining machine learning and human expert experience in medication adherence research. Bayesian Network diagrams (directed acyclic graphs) from the presenter’s experiment will be provided to visualise differences and similarities in human, machine learning, and combined approaches.
Biography: Taru Sorsa is an MSc Health Psychology student at the institute of Psychiatry, Psychology, and Neuroscience at King's College London. She is also an active member of the Institute of Mathematics and Its Applications (IMA) and has integrated applied maths and statistics into her research endeavours at King’s. This presentation will regard her MSc thesis investigating the utility of the theory of complex systems and Bayesian Network Analysis in medication adherence research. She is interested in specialising in complex disease modelling utilising mathematical, statistical, machine learning and AI tools and techniques.