Clinical AI reasoning for multimorbid patients: feasibility of calibrated differential diagnosis and treatment-plan evaluation

This project investigates how AI can build calibrated differential diagnoses for multimorbid patients and how its treatment plans can be reliably evaluated.

Key data

Description

Multimorbidity is common, but information about a patient is often spread across different consultations, medications, test results and follow-up information. This project investigates how AI can bring these pieces together to build a more complete picture of a patient’s health over time.

The main aim is to establish the foundations for AI that can consider several possible conditions at the same time, update its assessment as new patient information becomes available, and indicate how confident it is. In parallel, we will explore how personalised treatment plans proposed by AI can be assessed reliably.