Patient identity and clinical features predict brain tumor segmentation accuracy more strongly than model choice, with localized spatial biases consistent across models and no formal fairness guarantees in any.
hub
arXiv preprint arXiv:1811.02629 (2018)
20 Pith papers cite this work. Polarity classification is still indexing.
20
Pith papers citing it
abstract
Gliomas are the most common primary brain malignancies, with different degrees of aggressiveness, variable prognosis and various heterogeneous histologic sub-regions, i.e., peritumoral edematous/invaded tissue, necrotic core, active and non-enhancing core. This intrinsic heterogeneity is also portrayed in their radio-phenotype, as their sub-regions are depicted by varying intensity profiles disseminated across multi-parametric magnetic resonance imaging (mpMRI) scans, reflecting varying biological properties. Their heterogeneous shape, extent, and location are some of the factors that make these tumors difficult to resect, and in some cases inoperable. The amount of resected tumor is a factor also considered in longitudinal scans, when evaluating the apparent tumor for potential diagnosis of progression. Furthermore, there is mounting evidence that accurate segmentation of the various tumor sub-regions can offer the basis for quantitative image analysis towards prediction of patient overall survival. This study assesses the state-of-the-art machine learning (ML) methods used for brain tumor image analysis in mpMRI scans, during the last seven instances of the International Brain Tumor Segmentation (BraTS) challenge, i.e., 2012-2018. Specifically, we focus on i) evaluating segmentations of the various glioma sub-regions in pre-operative mpMRI scans, ii) assessing potential tumor progression by virtue of longitudinal growth of tumor sub-regions, beyond use of the RECIST/RANO criteria, and iii) predicting the overall survival from pre-operative mpMRI scans of patients that underwent gross total resection. Finally, we investigate the challenge of identifying the best ML algorithms for each of these tasks, considering that apart from being diverse on each instance of the challenge, the multi-institutional mpMRI BraTS dataset has also been a continuously evolving/growing dataset.
hub tools
citation-role summary
dataset 3
background 1
citation-polarity summary
representative citing papers
SAMRI fine-tunes only the mask decoder of SAM on 1.1 million MRI slices from 30 datasets to reach mean DSC 0.87 on 47 targets and strong zero-shot performance.
MedFL-Stress shows that FedBN reduces the best-to-worst hospital Dice gap by 41% versus FedAvg in federated brain tumor segmentation under simulated cross-hospital MRI shifts, improving the weakest site by 3.5 points with only a 0.5-point mean drop.
HiPerfGNN uses VQ-VAE on DSC perfusion curves to form hierarchical tumor habitat graphs that predict IDH mutation (AUC 0.96 internal, 0.89 external), 1p/19q codeletion, and WHO grade.
MedRCube is a new fine-grained evaluation framework that benchmarks 33 MLLMs on medical imaging, ranks Lingshu-32B highest, and finds a significant positive link between shortcut behaviors and diagnostic performance.
TaCo contrastively embeds semantic, generative, and transformation tasks from medical imaging into a joint space to reveal which tasks cluster, blend, or remain distinct.
Multimodal glioma survival models achieve better performance through additive integration of image and RNA features rather than cross-modal synergies, as quantified by lower measured interactions in stronger architectures.
SegGuidedNet achieves Dice scores of 0.905 on BraTS2021 and 0.897 on BraTS2023 with sub-region attention supervision that adds under 0.2% parameters and provides free spatial interpretability.
A multi-center whole-body FDG PET/CT foundation model with early fusion and masked autoencoding pretraining achieves label-efficient tumor segmentation on downstream tasks.
QSurv uses Gauss-Legendre numerical quadrature and time-conditioned low-rank adaptation to enable scalable nonparametric continuous-time survival modeling with theoretical error bounds.
A self-supervised approach uses consistent spatial relationships of anatomical structures across patients to improve 3D multi-modal medical image representations, yielding modest gains on segmentation and classification tasks.
CC-DiceCE boosts recall for small lesion segmentation in MRI with minimal degradation in other metrics and generally outperforms blob loss across datasets.
A novel weakly supervised anomaly detection method for brain MRI that uses discriminative dual prompt tuning for pseudo masks and region-aware spatial attention with location-based random embeddings to achieve SOTA results with under 8 million parameters on BraTS and MSD datasets.
MK-ResRecon predicts missing slices with a multi-kernel texture-aware loss while IdentityRefineNet3D refines the combined 3D volume to enable accurate reconstruction from highly sparse 2D inputs.
InfiltrNet uses a CNN-Transformer hybrid with distance-transform labels to generate infiltration risk maps from MRI and outperforms baselines on BraTS 2020 and 2025 datasets.
UniME combines a pretrained unified ViT encoder with modality-specific CNN encoders to improve brain tumor segmentation performance when some MRI modalities are missing.
Post-hoc normalizing flows for OOD detection in medical imaging achieve 84.61% AUROC on MedOOD and 93.8% on MedMNIST, outperforming ViM, MDS, and ReAct.
A multi-dataset cross-domain knowledge distillation approach improves unified performance on medical image segmentation, classification, and detection by transferring domain-invariant features from a joint teacher model to task-specific students.
BraTS 2021 provides a 2,040-patient mpMRI benchmark for brain tumor sub-region segmentation and MGMT methylation classification, hosted on Synapse and Kaggle with $60,000 in awards.
ADRUwAMS reports Dice scores of 0.9229 (whole tumor), 0.8432 (tumor core), and 0.8004 (enhancing tumor) on BraTS 2020 after training on BraTS 2019/2020 datasets.
citing papers explorer
-
Fairboard: a quantitative framework for equity assessment of healthcare models
Patient identity and clinical features predict brain tumor segmentation accuracy more strongly than model choice, with localized spatial biases consistent across models and no formal fairness guarantees in any.
-
SAMRI: Segment Any MRI
SAMRI fine-tunes only the mask decoder of SAM on 1.1 million MRI slices from 30 datasets to reach mean DSC 0.87 on 47 targets and strong zero-shot performance.
-
MedFL-Stress: A Systematic Robustness Evaluation of Federated Brain Tumor Segmentation under Cross-Hospital MRI Appearance Shift
MedFL-Stress shows that FedBN reduces the best-to-worst hospital Dice gap by 41% versus FedAvg in federated brain tumor segmentation under simulated cross-hospital MRI shifts, improving the weakest site by 3.5 points with only a 0.5-point mean drop.
-
Hierarchical Perfusion Graphs for Tumor Heterogeneity Modeling in Glioma Molecular Subtyping
HiPerfGNN uses VQ-VAE on DSC perfusion curves to form hierarchical tumor habitat graphs that predict IDH mutation (AUC 0.96 internal, 0.89 external), 1p/19q codeletion, and WHO grade.
-
MedRCube: A Multidimensional Framework for Fine-Grained and In-Depth Evaluation of MLLMs in Medical Imaging
MedRCube is a new fine-grained evaluation framework that benchmarks 33 MLLMs on medical imaging, ranks Lingshu-32B highest, and finds a significant positive link between shortcut behaviors and diagnostic performance.
-
Probing Intrinsic Medical Task Relationships: A Contrastive Learning Perspective
TaCo contrastively embeds semantic, generative, and transformation tasks from medical imaging into a joint space to reveal which tasks cluster, blend, or remain distinct.
-
Quantifying Cross-Modal Interactions in Multimodal Glioma Survival Prediction via InterSHAP: Evidence for Additive Signal Integration
Multimodal glioma survival models achieve better performance through additive integration of image and RNA features rather than cross-modal synergies, as quantified by lower measured interactions in stronger architectures.
-
SegGuidedNet: Sub-Region-Aware Attention Supervision for Interpretable Brain Tumor Segmentation
SegGuidedNet achieves Dice scores of 0.905 on BraTS2021 and 0.897 on BraTS2023 with sub-region attention supervision that adds under 0.2% parameters and provides free spatial interpretability.
-
An Open Multi-Center Whole-Body FDG PET/CT Foundation Model for Tumor Segmentation
A multi-center whole-body FDG PET/CT foundation model with early fusion and masked autoencoding pretraining achieves label-efficient tumor segmentation on downstream tasks.
-
A Scalable Nonparametric Continuous-Time Survival Model through Numerical Quadrature
QSurv uses Gauss-Legendre numerical quadrature and time-conditioned low-rank adaptation to enable scalable nonparametric continuous-time survival modeling with theoretical error bounds.
-
Beyond Instance-Level Self-Supervision in 3D Multi-Modal Medical Imaging
A self-supervised approach uses consistent spatial relationships of anatomical structures across patients to improve 3D multi-modal medical image representations, yielding modest gains on segmentation and classification tasks.
-
Learning to Look Closer: A New Instance-Wise Loss for Small Cerebral Lesion Segmentation
CC-DiceCE boosts recall for small lesion segmentation in MRI with minimal degradation in other metrics and generally outperforms blob loss across datasets.
-
RASALoRE: Region Aware Spatial Attention with Location-based Random Embeddings for Weakly Supervised Anomaly Detection in Brain MRI Scans
A novel weakly supervised anomaly detection method for brain MRI that uses discriminative dual prompt tuning for pseudo masks and region-aware spatial attention with location-based random embeddings to achieve SOTA results with under 8 million parameters on BraTS and MSD datasets.
-
MK-ResRecon: Multi-Kernel Residual Framework for Texture-Aware 3D MRI Refinement from Sparse 2D Slices
MK-ResRecon predicts missing slices with a multi-kernel texture-aware loss while IdentityRefineNet3D refines the combined 3D volume to enable accurate reconstruction from highly sparse 2D inputs.
-
InfiltrNet: Dual-Branch CNN-Transformer Architecture for Brain Tumor Infiltration Risk Prediction
InfiltrNet uses a CNN-Transformer hybrid with distance-transform labels to generate infiltration risk maps from MRI and outperforms baselines on BraTS 2020 and 2025 datasets.
-
Uni-Encoder Meets Multi-Encoders: Representation Before Fusion for Brain Tumor Segmentation with Missing Modalities
UniME combines a pretrained unified ViT encoder with modality-specific CNN encoders to improve brain tumor segmentation performance when some MRI modalities are missing.
-
Safeguarding AI in Medical Imaging: Post-Hoc Out-of-Distribution Detection with Normalizing Flows
Post-hoc normalizing flows for OOD detection in medical imaging achieve 84.61% AUROC on MedOOD and 93.8% on MedMNIST, outperforming ViM, MDS, and ReAct.
-
Multi-Dataset Cross-Domain Knowledge Distillation for Unified Medical Image Segmentation, Classification, and Detection
A multi-dataset cross-domain knowledge distillation approach improves unified performance on medical image segmentation, classification, and detection by transferring domain-invariant features from a joint teacher model to task-specific students.
-
The RSNA-ASNR-MICCAI BraTS 2021 Benchmark on Brain Tumor Segmentation and Radiogenomic Classification
BraTS 2021 provides a 2,040-patient mpMRI benchmark for brain tumor sub-region segmentation and MGMT methylation classification, hosted on Synapse and Kaggle with $60,000 in awards.
-
Adaptive Dual Residual U-Net with Attention Gate and Multiscale Spatial Attention Mechanisms (ADRUwAMS)
ADRUwAMS reports Dice scores of 0.9229 (whole tumor), 0.8432 (tumor core), and 0.8004 (enhancing tumor) on BraTS 2020 after training on BraTS 2019/2020 datasets.