King's College London

Research portal

Generalized Wasserstein Dice Loss, Test-Time Augmentation, and Transformers for the BraTS 2021 Challenge

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review

Lucas Fidon, Suprosanna Shit, Ivan Ezhov, Johannes C. Paetzold, Sébastien Ourselin, Tom Vercauteren

Original languageEnglish
Title of host publicationBrainlesion
Subtitle of host publicationGlioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 7th International Workshop, BrainLes 2021, Held in Conjunction with MICCAI 2021, Revised Selected Papers
EditorsAlessandro Crimi, Spyridon Bakas
PublisherSpringer Science and Business Media Deutschland GmbH
Pages187-196
Number of pages10
ISBN (Print)9783031090011
DOIs
Published2022
Event7th International Brain Lesion Workshop, BrainLes 2021, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2021 - Virtual, Online
Duration: 27 Sep 202127 Sep 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12963 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference7th International Brain Lesion Workshop, BrainLes 2021, held in conjunction with the Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
CityVirtual, Online
Period27/09/202127/09/2021

Bibliographical note

Funding Information: Acknowledgments. This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Sk lodowska-Curie grant agreement TRABIT No 765148; Wellcome [203148/Z/16/Z; WT101957], EPSRC [NS/A000049/1; NS/A000027/1]. Tom Vercauteren is supported by a Medtronic/RAEng Research Chair [RCSRF1819\7\34]. Data used in this publication were obtained as part of the RSNA-ASNR-MICCAI Brain Tumor Segmentation (BraTS) Challenge project through Synapse ID (syn25829067). Funding Information: This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Sk̷lodowska-Curie grant agreement TRABIT No 765148; Wellcome [203148/Z/16/Z; WT101957], EPSRC [NS/A000049/1; NS/A000027/1]. Tom Vercauteren is supported by a Medtronic/RAEng Research Chair [RCSRF1819\7\34]. Data used in this publication were obtained as part of the RSNA-ASNR-MICCAI Brain Tumor Segmentation (BraTS) Challenge project through Synapse ID (syn25829067). Publisher Copyright: © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

King's Authors

Abstract

Brain tumor segmentation from multiple Magnetic Resonance Imaging (MRI) modalities is a challenging task in medical image computation. The main challenges lie in the generalizability to a variety of scanners and imaging protocols. In this paper, we explore strategies to increase model robustness without increasing inference time. Towards this aim, we explore finding a robust ensemble from models trained using different losses, optimizers, and train-validation data split. Importantly, we explore the inclusion of a transformer in the bottleneck of the U-Net architecture. While we find transformer in the bottleneck performs slightly worse than the baseline U-Net in average, the generalized Wasserstein Dice loss consistently produces superior results. Further, we adopt an efficient test time augmentation strategy for faster and robust inference. Our final ensemble of seven 3D U-Nets with test-time augmentation produces an average dice score of 89.4% and an average Hausdorff 95% distance of 10.0 mm when evaluated on the BraTS 2021 testing dataset. Our code and trained models are publicly available at https://github.com/LucasFidon/TRABIT_BraTS2021.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454