Nicola K. Dinsdale

This paper develops a deep learning model using domain adaptation to predict high-resolution fibre orientations from diffusion MRI, informed by microscopy data, improving connectivity mapping in human brains.

UniFed addresses harmonisation and privacy in neuroimaging by enabling federated learning with partially labelled distributed data, producing high-quality segmentations.

STAMP trains and prunes a UNet for medical image segmentation with online pruning, achieving higher performance and reducing parameters by over 85% with fewer labels.

This study compares domain adaptation techniques to improve white matter hyperintensity segmentation across different MRI datasets, finding that unsupervised domain adversarial training performs best.

This paper introduces a deep learning framework to reduce scanner bias and other confounds in MRI data, improving harmonization and generalizability across datasets.

This paper uses a 3D CNN to predict brain age from MRI data and finds that prediction errors correlate with clinical measurements and image-derived phenotypes.