Perpetual Modelling and Decision Making for Biomedical Systems

Computational power and maturity of mathematical techniques has allowed us to be able to analyse vast amount of data in the biomedical sector which has led to the development of various successful modelling and decision making techniques. For example: in cancer prediction, automatic decision making in patient care, physiological-compartmental modelling etc. Despite such advances, the developed models and techniques are often criticised by clinicians for lack of simplicity and transparency, and increased complexity, thus making such methodologies impractical to use and adopt by non-experts (medical staff, nurses etc.). This project will focus on the creation of modelling and decision making methodologies that are transparent, and easily interpretable by focusing on Computational Intelligence (CI) techniques, specifically those techniques trying to mimic human cognition and intuition (Fuzzy Logic, Neural-Fuzzy Logic, multi-modal learning, adaptive/evolving systems etc.). Using such techniques computational structures will be created that mimic the human cognitive behaviour aiming to develop transparent- interpretable models, and decision making paradigms for use in the Biomedical sector. Paramount system behaviour will be the one of life-long learning (perpetual systems), e.g. the system adapts its behaviour during the patient care/therapy to match the patient’s changing characteristics. Examples include the prediction and treatment schedule of cancer, mechanical ventilation support systems, monitoring of the lung function as well as ‘patient-tailored’ systems for the critically ill (Intensive Care Units).