Showing 1–1 of 1 results for author: de Gelder, A

Autoencoders for Multi-Label Prostate MR Segmentation

Authors: Ard de Gelder, Henkjan Huisman

Abstract: Organ image segmentation can be improved by implementing prior knowledge about the anatomy. One way of doing this is by training an autoencoder to learn a lowdimensional representation of the segmentation. In this paper, this is applied in multi-label prostate MR segmentation, with some positive results. Organ image segmentation can be improved by implementing prior knowledge about the anatomy. One way of doing this is by training an autoencoder to learn a lowdimensional representation of the segmentation. In this paper, this is applied in multi-label prostate MR segmentation, with some positive results. △ Less

Submitted 26 June, 2018; v1 submitted 9 June, 2018; originally announced June 2018.