King's College London

Research portal

PADDIT: Probabilistic Augmentation of Data using Diffeomorphic Image Transformation

Research output: Chapter in Book/Report/Conference proceedingConference paper

Mauricio Orbes-Arteaga, Lauge Sørensen, Jorge Cardoso, Marc Modat, Sebastien Ourselin, Stefan Sommer, Mads Nielsen, Christian Igel, Akshay Pai

Original languageEnglish
Title of host publicationMedical Imaging 2019
Subtitle of host publicationImage Processing
EditorsElsa D. Angelini, Elsa D. Angelini, Elsa D. Angelini, Bennett A. Landman
PublisherSPIE
Volume10949
ISBN (Electronic)9781510625457
DOIs
Publication statusPublished - 15 Mar 2019
EventMedical Imaging 2019: Image Processing - San Diego, United States
Duration: 19 Feb 201921 Feb 2019

Conference

ConferenceMedical Imaging 2019: Image Processing
CountryUnited States
CitySan Diego
Period19/02/201921/02/2019

King's Authors

Abstract

For proper generalization performance of convolutional neural networks (CNNs) in medical image segmentation, the learnt features should be invariant under particular non-linear shape variations of the input. To induce invariance in CNNs to such transformations, we propose Probabilistic Augmentation of Data using Diffeomorphic Image Transformation (PADDIT) - a systematic framework for generating realistic transformations that can be used to augment data for training CNNs. The main advantage of PADDIT is the ability to produce transformations that capture the morphological variability in the training data. To this end, PADDIT constructs a mean template which represents the main shape tendency of the training data. A Hamiltonian Monte Carlo(HMC) scheme is used to sample transformations which warp the training images to the generated mean template. Augmented images are created by warping the training images using the sampled transformations. We show that CNNs trained with PADDIT outperforms CNNs trained without augmentation and with generic augmentation (0.2 and 0.15 higher dice accuracy respectively) in segmenting white matter hyperintensities from T1 and FLAIR brain MRI scans.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454