Learning an MR acquisition-invariant representation using Siamese neural networks
W.M. Kouw (TU Delft - Pattern Recognition and Bioinformatics, University of Copenhagen)
M. Loog (TU Delft - Pattern Recognition and Bioinformatics, University of Copenhagen)
Wilbert Bartels (University Medical Center Utrecht)
A.M. Mendrik (University Medical Center Utrecht, Netherlands eScience Center)
More Info
expand_more
Abstract
Generalization of voxelwise classifiers is hampered by differences between MRI-scanners, e.g. different acquisition protocols and field strengths. To address this limitation, we propose a Siamese neural network (MRAI-net) that extracts acquisition-invariant feature vectors. These can consequently be used by task-specific methods, such as voxelwise classifiers for tissue segmentation. MRAI-net is tested on both simulated and real patient data. Experiments show that MRAI-net outperforms voxelwise classifiers trained on the source or target scanner data when a small number of labeled samples is available.
No files available
Metadata only record. There are no files for this record.