Few-shot learning enables medical image segmentation models to adapt to new tasks using only a small number of labelled examples. However, adaptation performance depends strongly on which examples are selected for the support set. Effective support sets should capture relevant va…
Deep learning models for prostate MRI-based cancer grading may encode clinical covariates that either reflect useful disease-related signal or non-generalising shortcut information, but their role is usually assumed. We propose a causal-reasoning framework for probing covariate d…