PhD Supervisor: Edward Rolando Núñez Valdez (UNIOVI); Auxiliary supervisors David Camacho (UPM), Jia-Chun Lin (NTNU), Rafał Cupek (SUT); R&D cooperation: BioKeralty , TNP

Objectives: The primary goal of this project is to thoroughly investigate the key techniques, algorithms, and tools required to develop and implement a highly personalised health recommendation system. This system will use Large Language Models (LLMs) and machine learning techniques to interpret medical literature, patient records, and current health trends. The fundamental purpose of this project is to provide personalised health advice and early warning signals. Furthermore, we aim to employ advanced LLMs and explainable artificial intelligence to analyse large volumes of medical research and patient data. To achieve this, innovative algorithms will be applied to ensure transparency and accuracy in the recommendations generated by the system. As an integral part of this objective, comprehensive research will be conducted to develop methods that enable the quantification and visualisation of the reasoning process of the LLM. This approach will significantly contribute to the understanding and continuous improvement of the personalised health recommendation system.

Expected Results: The main expected outcome is to conduct a series of experiments to develop a tool tailored for healthcare providers and patients. This tool aims to support informed decision-making by leveraging a comprehensive understanding of individual health contexts and the latest medical knowledge. The potential tool seeks to enhance medical decision-making and will be fundamentally supported by a recommendation system that is accurate, transparent, scalable, and ensures user privacy and security.

Applied research: The creation of a tool based on these technologies aims to significantly enhance medical decision-making, resulting in tangible benefits for the well-being of patients and healthcare personnel in healthcare institutions. The integration of advanced Large Language Models (LLMs) and explainable artificial intelligence into the tool will provide transparency and accuracy in decision support, fostering the enhancement and comprehension of the personalised health recommendation system. The developed components will be verified by BioKeralty in their healthcare applications. These systems have the potential to expand and improve the technological portfolio used by BioKeralty in their health centres and hospitals. The TNP will support DC3 with laboratory infrastructure.

Planned secondments: UPM (4 months); NTNU(4 months); SUT (4 months)                   

Enrolment in Doctoral degree: UNIOVI