The fusion of statistical and AI Methodologies for interpreting uncertainty in healthcare

  • Aleksandar Novakovic

Student thesis: Doctoral ThesisDoctor of Philosophy by Independent Research


Uncertainty is an ever-present phenomenon in healthcare that affects all stakeholders and can be seen as one of the main triggers for confusion, anxiety and challenge in decision making. This thesis presents a synthesis of themes from the works published in the period between 2018-2022 in collaboration with colleagues from Queen’s University Belfast (UK) and Ontario Tech University (Canada) with the goal to present novel methodologies resulting from a fusion of statistical and AI techniques, the purpose of which is in aiding healthcare stakeholders in their decision making by minimising the negative effect of uncertainty. There are 7 publications in total that are forming the body of work of this thesis, of which 4 are journal articles, 2 are book chapters and 1 is in conference proceedings. Rather than present the papers in chronological order, they are included in the thesis where they are naturally best placed to suit the flow of the narrative. The main body of work, provides the detailed contributions that the publications have made to the statistics, computer science, health informatics and public health domains. In particular, the contributions are presented across the following three themes: (1) novel metrics for assessing the accuracy of streaming clinical decision support systems, (2) their deployment and accelerated benchmark testing, and (3) the development of a public health reporting model for mitigating the spread of infectious diseases in the population. The thesis concludes with a critical account of the significance of the published works, that is, the impact that these have made in the academic, industrial and public domains.
Date of AwardJul 2023
Original languageEnglish
Awarding Institution
  • Queen's University Belfast
SupervisorAdele Marshall (Supervisor) & Carolyn McGregor (Supervisor)


  • Healthcare
  • AI
  • statistical

Cite this