Recent work has uncovered alarming disparities in the performance of machine learning models in healthcare. In this study, we explore whether such disparities are present in the UK Biobank fundus retinal images by training and evaluating a disease classification model on these images. We assess possible disparities across various population groups and find substantial differences despite strong overall performance of the model. In particular, we discover unfair performance for certain assessment centres, which is surprising given the rigorous data standardisation protocol. We compare how these differences emerge and apply a range of existing bias mitigation methods to each one. A key insight is that each disparity has unique properties and responds differently to the mitigation methods. We also find that these methods are largely unable to enhance fairness, highlighting the need for better bias mitigation methods tailored to the specific type of bias.

AI is real to many of us in business. Yet much of the debate about machine learning, AI and the use of Big Data remains hyperbolic. Headlines screeching about robot takeovers and mass job losses might do well for click-rates. Data specialists and early adopters may wax lyrical about the technological advances being made, and how these processes far outstrip the capacity, speed and computational power of mere mortals. But both polarities are far behind the more complex, interesting reality – and they're missing out on the real news for us humans.