Uri Nahum | Searcharxiv

arXiv2026

A Benchmark of (MRI-) Foundation Models to Predict IDH Mutational Status in Glioma

Non-invasive prediction of glioma molecular status from routine magnetic resonance imaging (MRI) has shown promising performance, but model generalization remains challenging given small-scale matched imaging-genomic datasets. Foundation models may address this bottleneck, but a comprehensive benchmark is needed to establish the impact of diverse architectures, pre-training domains, and objectives. Given the use case of isocitrate dehydrogenase (IDH) mutation prediction from FLAIR and post-contrast T1 MRIs, we compared four image-based foundation models, BrainIAC, MRI-CORE, BiomedCLIP, and BrainDINO, against radiomics-based TabPFN and logistic regression baselines. Prediction performance and calibration were assessed across four public adult glioma cohorts and an external post-treatment cohort. Within-cohort, TabPFN matched or outperformed all visual encoders, achieving 0.92 (0.03) AUROC and 0.74 (0.17) AUPRC (mean (SD) across all datasets). Among visual encoders, BiomedCLIP performed best (0.85 (0.08) AUROC), with BrainDINO competitive (0.82 (0.09) AUROC), while MRI-specific encoders (BrainIAC, MRI-CORE) consistently underperformed. Cross-cohort transfer showed moderate AUROC degradation but stronger AUPRC sensitivity to prevalence shifts. On the external cohort, BiomedCLIP achieved the highest AUROC (0.74 (0.07)), whereas TabPFN provided superior calibration (Expected Calibration Error 0.07 (0.01)). These results indicate that representation modality and evaluation context critically influence foundation-model performance in MRI-based molecular prediction. Tabular foundation models on radiomic features provide a strong, well-calibrated baseline, while image foundation models may offer complementary value under clinically distinct distribution shifts. Code available at https://github.com/nathanhollet/idh-status-prediction

eess.IV↗

arXiv2018

Adaptive Eigenspace Segmentation

Image segmentation is an inherently ill-posed problem and thus requires regularization in order to limit the search space to reasonable solutions. A majority of segmentation methods integrates these regularization terms in one way or the other in an energy functional using a balancing term. The tuning of this parameter that either favours more the regularization or the data conformity is critical and, unfortunately, the success of the optimization process strongly depends on it. Often the optimal settings change from image to image. In this paper we propose a novel general framework based on an adaptive eigenspace that was first proposed for solving inverse problems. The resulting method proves accurate and yields robust results, without the need for optimization techniques or being sensitive to the parameter choice. In fact, the method solves a symmetric positive definite sparse system and hence, uses only a fraction of the computational cost. The method is very versatile and does not need parameter-tuning, when segmenting objects from any kind of an image or when segmenting different organs. As the adaptive eigenspace is determined directly from the image to segment, the approach also does not need a tedious training phase.

math.NA↗