A Stratified Cascaded Approach for Brain Tumor Segmentation with the Aid of Multi-modal Synthetic Data

Yasmina Al Khalil, Aymen Ayaz, Cristian Lorenz, Jürgen Weese, Josien Pluim, Marcel Breeuwer

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

1 Citaat (Scopus)

Samenvatting

Gliomas are one of the most widespread and aggressive forms of brain tumors. Accurate brain tumor segmentation is crucial for evaluation, monitoring and treatment of gliomas. Recent advances in deep learning methods have made a significant step towards a robust and automated brain tumor segmentation. However, due to the variation in shape and location of gliomas, as well as their appearance across different tumor grades, obtaining an accurate and generalizable segmentation model is still a challenge. To alleviate this, we propose a cascaded segmentation pipeline, aimed at introducing more robustness to segmentation performance through data stratification. In other words, we train separate models per tumor grade, aided with synthetic brain tumor images generated through conditional generative adversarial networks. To handle the variety in size, shape and location of tumors, we utilize a localization module, focusing the training and inference in the vicinity of the tumor. Finally, to identify which tumor grade segmentation model to utilize at inference time, we train a dense, attention-based 3D classification model. The obtained results suggest that both stratification and the addition of synthetic data to training significantly improve the segmentation performance, whereby up to 55% of test cases exhibit a performance improvement by more than 5% and up to 40% of test cases exhibit an improvement by more than 10% in Dice score.

Originele taal-2Engels
TitelData Augmentation, Labelling, and Imperfections
SubtitelSecond MICCAI Workshop, DALI 2022, Held in Conjunction with MICCAI 2022, Singapore, September 22, 2022, Proceedings
RedacteurenHien V. Nguyen, Sharon X. Huang, Yuan Xue
UitgeverijSpringer
Hoofdstuk10
Pagina's92-101
Aantal pagina's10
ISBN van elektronische versie978-3-031-17027-0
ISBN van geprinte versie978-3-031-17026-3
DOI's
StatusGepubliceerd - 2022
Evenement25th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2022 - Singapore, Singapore
Duur: 18 sep. 202222 sep. 2022
Congresnummer: 25

Publicatie series

NaamLecture Notes in Computer Science (LNCS)
UitgeverijSpringer
Volume13567
ISSN van geprinte versie0302-9743
ISSN van elektronische versie1611-3349

Congres

Congres25th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2022
Verkorte titelMICCAI 2022
Land/RegioSingapore
StadSingapore
Periode18/09/2222/09/22

Bibliografische nota

Funding Information:
Acknowledgements. This research is part of the openGTN project, supported by the European Union in the Marie Curie Innovative Training Networks (ITN) fellowship program under project No. 764465.

Vingerafdruk

Duik in de onderzoeksthema's van 'A Stratified Cascaded Approach for Brain Tumor Segmentation with the Aid of Multi-modal Synthetic Data'. Samen vormen ze een unieke vingerafdruk.

Citeer dit