Glioblastoma Overall Survival Prediction With Vision Transformers

2025-08-04

AI Summary

Accurate overall survival (OS) prediction for glioblastoma (GBM) patients remains challenging, particularly when relying on labor-intensive tumor segmentation or annotated lesion masks. Method: We propose a segmentation-free, end-to-end vision transformer (ViT) framework that directly learns prognostic latent features from raw multi-modal MRI volumes—bypassing preprocessing steps such as tumor delineation and significantly reducing computational overhead. Contribution/Results: To the best of our knowledge, this is the first study applying ViT to segmentation-free GBM survival prediction. Evaluated on the BraTS dataset, our model achieves 62.5% OS classification accuracy. While not surpassing state-of-the-art (SOTA) accuracy, it attains superior balance across precision, recall, and F1-score—demonstrating enhanced robustness and clinical applicability. This work establishes a novel paradigm for prognostic modeling in low-resolution medical imaging, offering a streamlined, annotation-efficient alternative to conventional segmentation-dependent approaches.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multimodal Learning
• Natural Language Processing: Safety and Robustness

Application Category

• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Glioblastoma is one of the most aggressive and common brain tumors, with a median survival of 10-15 months. Predicting Overall Survival (OS) is critical for personalizing treatment strategies and aligning clinical decisions with patient outcomes. In this study, we propose a novel Artificial Intelligence (AI) approach for OS prediction using Magnetic Resonance Imaging (MRI) images, exploiting Vision Transformers (ViTs) to extract hidden features directly from MRI images, eliminating the need of tumor segmentation. Unlike traditional approaches, our method simplifies the workflow and reduces computational resource requirements. The proposed model was evaluated on the BRATS dataset, reaching an accuracy of 62.5% on the test set, comparable to the top-performing methods. Additionally, it demonstrated balanced performance across precision, recall, and F1 score, overcoming the best model in these metrics. The dataset size limits the generalization of the ViT which typically requires larger datasets compared to convolutional neural networks. This limitation in generalization is observed across all the cited studies. This work highlights the applicability of ViTs for downsampled medical imaging tasks and establishes a foundation for OS prediction models that are computationally efficient and do not rely on segmentation.

Problem

• Predicting glioblastoma survival using MRI images without tumor segmentation
• Simplifying workflow and reducing computational resource requirements
• Evaluating Vision Transformers for downsampled medical imaging tasks

Innovation

• Vision Transformers for MRI feature extraction
• Eliminates need for tumor segmentation
• Computationally efficient OS prediction model