People-centred Medical AI

Artificial intelligence (AI) models have been proposed to automate routine medical image analysis tasks for helping doctors achieve better decisions quicker. The design of these AI models tends to be centred on data rather than people, e.g., patients and doctors, where model optimisation focuses on maximising the accuracy of a generic model for the majority of patients instead of improving the performance of all sub-groups of radiologists for all patient cohorts. Such design strategy has produced accurate AI models that can improve the performance of most doctors for the majority of patients. However, doctors can be divided into sub-groups depending on their experience and track record, while patients can be grouped into cohorts subject to their age, family history, previous cancer diagnosis, ethnicity, and scanner characteristics. Therefore, unless we address this research gap and start building AI models that are helpful for all groups of doctors and patients, we will not see a widespread adoption of AI models in clinical practice.

This PhD program aims to address this research gap by proposing solutions to two important issues, namely: 1) the integration of doctors in the design and optimisation of the models; and 2) an unbiased classification for patients, independently to which cohort the patient belongs. The solutions to be proposed in this PhD program target a more usable AI model that will increase the sensitivity and specificity for all sub-groups of doctors, improving the efficiency of the medical image analysis process. Also, given that patients form cohorts in unpredictable ways (depending on dataset population and imaging technology), patients will be more likely to accept AI models that produce a fair outcome for all cohorts.

Open to any UK or international candidates.

Candidates will need to meet the minimum entry requirements for our PhD programme.