Data Privacy

Medical images are inherently personal in nature, and thus developed methods must protect individual privacy.

Home  /  Harmonisation  /  Segmentation  /  Translation  /  Privacy  /  Explainable AI

Papers

We, therefore, propose UniFed, a unified federated harmoni- sation framework, which enables three key processes to be completed: 1) the training of a federated partially labelled harmonisation network, 2) the selection of the most appropriate pretrained model for a new unseen site, and 3) the incorporation of a new site into the harmonised federation.

We propose an Unsupervised Source-Free Domain Adaptation (SFDA) method, SFHarmony, and demonstrate the approach for classification, regression and segmentation tasks.

Review article, explores: data availability, interpretability, model bias and data privacy.

Adapted my harmonisation framework to work in a distributed setting, through modelling site features as Gaussian distributions.