+1 Recommend
0 collections
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Divide and Conquer: Stratifying Training Data by Tumor Grade Improves Deep Learning-Based Brain Tumor Segmentation

      Read this article at

          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.


          It is a general assumption in deep learning that more training data leads to better performance, and that models will learn to generalize well across heterogeneous input data as long as that variety is represented in the training set. Segmentation of brain tumors is a well-investigated topic in medical image computing, owing primarily to the availability of a large publicly-available dataset arising from the long-running yearly Multimodal Brain Tumor Segmentation (BraTS) challenge. Research efforts and publications addressing this dataset focus predominantly on technical improvements of model architectures and less on properties of the underlying data. Using the dataset and the method ranked third in the BraTS 2018 challenge, we performed experiments to examine the impact of tumor type on segmentation performance. We propose to stratify the training dataset into high-grade glioma (HGG) and low-grade glioma (LGG) subjects and train two separate models. Although we observed only minor gains in overall mean dice scores by this stratification, examining case-wise rankings of individual subjects revealed statistically significant improvements. Compared to a baseline model trained on both HGG and LGG cases, two separately trained models led to better performance in 64.9% of cases ( p < 0.0001) for the tumor core. An analysis of subjects which did not profit from stratified training revealed that cases were missegmented which had poor image quality, or which presented clinically particularly challenging cases (e.g., underrepresented subtypes such as IDH1-mutant tumors), underlining the importance of such latent variables in the context of tumor segmentation. In summary, we found that segmentation models trained on the BraTS 2018 dataset, stratified according to tumor type, lead to a significant increase in segmentation performance. Furthermore, we demonstrated that this gain in segmentation performance is evident in the case-wise ranking of individual subjects but not in summary statistics. We conclude that it may be useful to consider the segmentation of brain tumors of different types or grades as separate tasks, rather than developing one tool to segment them all. Consequently, making this information available for the test data should be considered, potentially leading to a more clinically relevant BraTS competition.

          Related collections

          Most cited references 41

          • Record: found
          • Abstract: found
          • Article: not found

          The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS).

          In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences. Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients-manually annotated by up to four raters-and to 65 comparable scans generated using tumor image simulation software. Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%-85%), illustrating the difficulty of this task. We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously. Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements. The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource.
            • Record: found
            • Abstract: found
            • Article: found
            Is Open Access

            Advancing The Cancer Genome Atlas glioma MRI collections with expert segmentation labels and radiomic features

            Gliomas belong to a group of central nervous system tumors, and consist of various sub-regions. Gold standard labeling of these sub-regions in radiographic imaging is essential for both clinical and computational studies, including radiomic and radiogenomic analyses. Towards this end, we release segmentation labels and radiomic features for all pre-operative multimodal magnetic resonance imaging (MRI) (n=243) of the multi-institutional glioma collections of The Cancer Genome Atlas (TCGA), publicly available in The Cancer Imaging Archive (TCIA). Pre-operative scans were identified in both glioblastoma (TCGA-GBM, n=135) and low-grade-glioma (TCGA-LGG, n=108) collections via radiological assessment. The glioma sub-region labels were produced by an automated state-of-the-art method and manually revised by an expert board-certified neuroradiologist. An extensive panel of radiomic features was extracted based on the manually-revised labels. This set of labels and features should enable i) direct utilization of the TCGA/TCIA glioma collections towards repeatable, reproducible and comparative quantitative studies leading to new predictive, prognostic, and diagnostic assessments, as well as ii) performance evaluation of computer-aided segmentation methods, and comparison to our state-of-the-art method.
              • Record: found
              • Abstract: found
              • Article: not found

              Glioblastoma multiforme: exploratory radiogenomic analysis by using quantitative image features.

              To derive quantitative image features from magnetic resonance (MR) images that characterize the radiographic phenotype of glioblastoma multiforme (GBM) lesions and to create radiogenomic maps associating these features with various molecular data.

                Author and article information

                Front Neurosci
                Front Neurosci
                Front. Neurosci.
                Frontiers in Neuroscience
                Frontiers Media S.A.
                05 November 2019
                : 13
                1Support Center for Advanced Neuroimaging (SCAN), University Institute of Diagnostic and Interventional Neuroradiology, Inselspital, Bern University Hospital, University of Bern , Bern, Switzerland
                2Graduate School for Cellular and Biomedical Sciences, University of Bern , Bern, Switzerland
                3Institute for Surgical Technology and Biomechanics, University of Bern , Bern, Switzerland
                4Healthcare Imaging A.I. Lab, Insel Data Science Center, Inselspital, Bern University Hospital , Bern, Switzerland
                Author notes

                Edited by: Bjoern Menze, Technical University of Munich, Germany

                Reviewed by: Roberto Viviani, University of Innsbruck, Austria; Benedikt Wiestler, Technical University of Munich, Germany

                *Correspondence: Michael Rebsamen michael.rebsamen@

                This article was submitted to Brain Imaging Methods, a section of the journal Frontiers in Neuroscience

                †These authors share senior authorship

                Copyright © 2019 Rebsamen, Knecht, Reyes, Wiest, Meier and McKinley.

                This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

                Page count
                Figures: 10, Tables: 2, Equations: 0, References: 52, Pages: 13, Words: 7258
                Funded by: Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung 10.13039/501100001711
                Funded by: Krebsliga Schweiz 10.13039/501100004361
                Original Research


                Comment on this article