Recent advances in artificial intelligence have produced highly accurate diagnostic models across a wide range of medical applications. However, these systems often provide little insight into how decisions are made, limiting clinician confidence and adoption in healthcare settings.
Large Language Models (LLMs) provide an opportunity to transform explainable AI by translating complex model outputs into human-readable explanations supported by current biomedical knowledge. When combined with retrieval-augmented generation and medical knowledge bases, LLMs have the potential to generate evidence-based explanations that are both faithful to the underlying AI model and understandable to healthcare professionals.
This project aims to investigate how large language models can enhance transparency, interpretability, and clinician interaction in next-generation medical AI systems.
Required knowledge
- Python programming
- Machine learning and deep learning
- Large Language Models (LLMs) and Natural Language Processing (NLP)
- Retrieval-Augmented Generation (RAG) (desirable)
- Explainable AI (XAI) and Human-AI interaction
- Basic knowledge of biomedical AI or medical imaging (desirable)
