This thesis explores the integration of established statistical methods and novel Artificial Intelligence (AI) techniques to enhance healthcare outcomes across diverse clinical domains. The work is structured in two main parts. The first part focuses on validating the utility of modern analytical frameworks in real-world clinical settings, including urinary tract cancer prediction, paediatric oncology, and fertility treatment optimisation. This section establishes the value of a synergistic approach, combining traditional statistical models with advanced machine learning algorithms to improve diagnostic accuracy and treatment personalisation. The second part of the thesis addresses complex predictive modelling challenges in assisted reproduction, particularly in the absence of ideal predictive outcomes for follicular growth. By employing innovative AI methodologies such as agent-based modelling, this research develops advanced predictive tools to optimise fertility treatments. These models provide deeper insights into follicular dynamics, overcoming limitations of existing methods and offering new avenues for clinical decision support. Overall, this thesis demonstrates that a thoughtful combination of established and novel analytical frameworks can bridge the gap between cutting-edge data science and clinical practice. The contributions presented not only advance the state-of-the-art in medical informatics but also provide practical, validated solutions that have the potential to significantly improve patient care and clinical workflows.
Artsiom Hramyka (Thu,) studied this question.