Abstract: In Defence of mathematical models for deep learning based registration

Lasse Hansen*, Maximilian Blendowski, Mattias P. Heinrich

*Corresponding author for this work


Deep learning based methods have not reached clinically acceptable results for common medical registration tasks that could be adequately solved using conventional methods. The slower progress compared to image segmentation is due to the lower availability of expert correspondences and the very large learnable parameter space for naive deep learning solutions. We strongly believe that it is necessary and beneficial to integrate conventional optimisation strategies as differentiable modules into deep learning based registration.

Original languageEnglish
Title of host publicationBildverarbeitung für die Medizin 2020
EditorsThomas Tolxdorff, Thomas M. Deserno, Heinz Handels, Andreas Maier, Klaus H. Maier-Hein, Christoph Palm
Number of pages1
PublisherSpringer Vieweg, Wiesbaden
Publication date12.02.2020
ISBN (Print)978-3-658-29266-9
ISBN (Electronic)978-3-658-29267-6
Publication statusPublished - 12.02.2020
EventBildverarbeitung für die Medizin 2020 - International workshop on Algorithmen - Systeme - Anwendungen
- Berlin, Germany
Duration: 15.03.202017.03.2020
Conference number: 237969


Dive into the research topics of 'Abstract: In Defence of mathematical models for deep learning based registration'. Together they form a unique fingerprint.

Cite this