Meeting Banner
Abstract #0658

dAUTOMAP: Decomposing AUTOMAP to Achieve Scalability and Enhance Performance

Jo Schlemper1, Ilkay Oksuz2, James Clough2, Jinming Duan3, Andrew P. King2, Julia A. Schnabel2, Joseph V Hajnal2, and Daniel Rueckert1

1Department of Computing, Imperial College London, London, United Kingdom, 2Biomedical Engineering, King's College London, London, United Kingdom, 3Faculty of Medicine, Institute of Clinical Sciences, Imperial College London, London, United Kingdom

AUTOMAP is a promising generalized reconstruction approach, however, it is not scalable and hence the practicality is limited. We present a novel way for decomposing the domain transformation, which makes the model scale linearly with the input size. We show the proposed method, termed dAUTOMAP, outperforms AUTOMAP with significantly fewer parameters.

This abstract and the presentation materials are available to members only; a login is required.

Join Here