학술논문
Multi-class glioma segmentation on real-world data with missing MRI sequences: comparison of three deep learning algorithms
Document Type
article
Author
Hugh G. Pemberton; Jiaming Wu; Ivar Kommers; Domenique M. J. Müller; Yipeng Hu; Olivia Goodkin; Sjoerd B. Vos; Sotirios Bisdas; Pierre A. Robe; Hilko Ardon; Lorenzo Bello; Marco Rossi; Tommaso Sciortino; Marco Conti Nibali; Mitchel S. Berger; Shawn L. Hervey-Jumper; Wim Bouwknegt; Wimar A. Van den Brink; Julia Furtner; Seunggu J. Han; Albert J. S. Idema; Barbara Kiesel; Georg Widhalm; Alfred Kloet; Michiel Wagemakers; Aeilko H. Zwinderman; Sandro M. Krieg; Emmanuel Mandonnet; Ferran Prados; Philip de Witt Hamer; Frederik Barkhof; Roelant S. Eijgelaar
Source
Scientific Reports, Vol 13, Iss 1, Pp 1-15 (2023)
Subject
Language
English
ISSN
2045-2322
Abstract
Abstract This study tests the generalisability of three Brain Tumor Segmentation (BraTS) challenge models using a multi-center dataset of varying image quality and incomplete MRI datasets. In this retrospective study, DeepMedic, no-new-Unet (nn-Unet), and NVIDIA-net (nv-Net) were trained and tested using manual segmentations from preoperative MRI of glioblastoma (GBM) and low-grade gliomas (LGG) from the BraTS 2021 dataset (1251 in total), in addition to 275 GBM and 205 LGG acquired clinically across 12 hospitals worldwide. Data was split into 80% training, 5% validation, and 15% internal test data. An additional external test-set of 158 GBM and 69 LGG was used to assess generalisability to other hospitals’ data. All models’ median Dice similarity coefficient (DSC) for both test sets were within, or higher than, previously reported human inter-rater agreement (range of 0.74–0.85). For both test sets, nn-Unet achieved the highest DSC (internal = 0.86, external = 0.93) and the lowest Hausdorff distances (10.07, 13.87 mm, respectively) for all tumor classes (p