To develop a multimodal AI model trained on mIF imaging and patient-specific clinical data to predict immunotherapy response

Immunotherapy has benefited many melanoma patients, but not everyone responds to treatment, and some experience severe side effects. This project aims to develop a multimodal AI model trained on both advanced tissue imaging and clinical data to better guide treatment decisions. Using multiplex immunofluorescence imaging, which detects multiple biomarkers in a single tissue sample, the model will capture a detailed picture of the tumour microenvironment, a key factor in treatment response. Adding clinical data will further enhance the model’s ability to maximise therapeutic benefits while minimising risks.