Purpose: To evaluate the impact of a structured tutorial on the use of a large language model (LLM)-based search engine on radiology residents’ performance in brain MRI differential diagnosis. Methods: In this study, nine radiology residents determined the three most likely differential diagnoses for three sets of ten brain MRI cases with a challenging yet definite diagnosis. Each set was assessed (1) with the support of conventional internet search, (2) using an LLM-based search engine (© Perplexity AI) without prior tutorial, or (3) using the LLM-based search engine after a structured 10-minute tutorial. Reader responses were rated using a binary and numeric scoring system. Reading times and confidence levels (measured on a 5-point Likert scale) were recorded for each case. Search engine logs were examined to quantify user interaction metrics, and to identify hallucinations and misinterpretations in LLM responses. Results: Radiology residents achieved the highest accuracy when employing the LLM-based search engine following the tutorial, indicating the correct diagnosis among the top three differential diagnoses in 62.5% of cases (55/88). This was followed by the LLM-assisted workflow before the tutorial (44.8%; 39/87) and the conventional internet search workflow (32.2%; 28/87). The LLM tutorial led to significantly higher performance (binary scores: p = 0.042, numeric scores: p = 0.016) and confidence ( p = 0.006) but resulted in no relevant differences in reading times. Hallucinations were found in 5.1% of LLM queries. Conclusion: Our findings demonstrate the considerable benefits that even low-effort educational interventions on LLMs can provide, highlighting their potential role in radiology training programs.

Background/Objectives : Large-scale (epi)genomic studies have substantially advanced our understanding of the molecular landscape of meningiomas, most recently embedded in the cIMPACT-NOW update 8. As a result, molecular data are increasingly integrated into risk-adapted treatment algorithms. However, it remains uncertain to what extent non-invasive MRI can capture underlying molecular variation and risk. Methods: We assembled a large, single-institution cohort of 225 newly diagnosed meningiomas (WHO grades 1–3) with available preoperative MRI, as well as comprehensive epigenome-wide methylation and copy-number profiling. Tumors were segmented into core and edema regions using a state-of-the-art automated pipeline from the BraTS challenge. Radiomic features were extracted and used to train Random Forest classifiers to predict WHO grade, molecular risk, and specific alterations such as 1p loss in a hold-out test set. Results: Our models achieved accuracy above 91% for integrated molecular risk classification, 87.5% for 1p chromosomal status, and 76.8% for WHO grade prediction, with corresponding AUCs of 0.91, 0.90, and 0.89, underscoring the robustness of radiomic features in capturing histopathological and, especially, molecular characteristics. Conclusions: Preoperative MRI effectively captures the underlying molecular biology of meningiomas and may enable rapid molecular assessment to inform decision-making and prioritization of confirmatory testing. However, it is not yet ready for clinical use, showing lower accuracy for current WHO grade classification.

Background: Precise glioma segmentation in MRI is essential for accurate diagnosis, optimal treatment planning, and advancing clinical research. However, most deep learning approaches require complete, standardized MRI protocols that are frequently unavailable in routine clinical practice. This study presents and evaluates GlioMODA, a robust deep learning framework designed for automated glioma segmentation that delivers consistent high performance across varied and incomplete MRI protocols. Methods: GlioMODA was trained and validated on the BraTS 2021 dataset (1,251 training, 219 testing cases), systematically assessing performance across eleven clinically relevant MRI protocol combinations. Segmentation accuracy was evaluated using Dice similarity coefficients (DSC) and panoptic quality metrics. Volumetric accuracy was benchmarked against manual ground truth, and statistical significance was established via Wilcoxon signed-rank tests with Benjamini–Yekutieli correction. Results: GlioMODA demonstrated state-of-the-art segmentation accuracy across tumor subregions, maintaining robust performance with incomplete or heterogeneous MRI protocols. Protocols including both T1-weighted contrast-enhanced and T2-FLAIR sequences yielded volumetric differences versus manual ground truth that were not statistically significant for enhancing tumor (ET: median difference 55 mm³, p = 0.157) and whole tumor (WT: median difference –7 mm³, p = 1.0), and exhibited median DSC differences close to zero relative to the four-sequence reference protocol. Omitting either sequence led to substantial and significant volumetric errors. Conclusions: GlioMODA facilitates reliable, automated glioma segmentation using a streamlined two-sequence protocol (T1-contrast + T2-FLAIR), supporting clinical workflow optimization and broader implementation of quantitative volumetry compatible with RANO 2.0 criteria. GlioMODA is published as an open-source, easy-to-use Python package at https://github.com/BrainLesion/GlioMODA/ . Key Points: T1-CE + T2-FLAIR maintains enhancing and whole tumor segmentation comparable to four-sequence MRI. Consistent volumes with T1-CE + T2-FLAIR support reliable RANO 2.0 assessment. Open-source GlioMODA (models + code) supports rapid integration. Importance of the Study: Automated glioma segmentation is limited in practice by incomplete or heterogeneous MRI protocols. GlioMODA directly addresses this barrier by delivering consistent accuracy across 11 clinically relevant sequence combinations and identifying a streamlined protocol (T1-contrast and T2-FLAIR) whose enhancing- and whole-tumor volumes are not statistically different from expert reference. This enables shorter scans and reproducible volumetry compatible with RANO 2.0, facilitating reliable response assessment in trials and routine care. By releasing trained models and code as an easy-to-use open-source package, this work enables external validation and integration into neuro-oncology workflows.

Precise glioma segmentation in magnetic resonance imaging (MRI) is essential for accurate diagnosis, optimal treatment planning, and advancing clinical research. However, most deep learning approaches require complete, standardized MRI protocols that are frequently unavailable in routine clinical practice. This study presents and evaluates GlioMODA, a robust deep learning framework designed for automated glioma segmentation that delivers consistent high performance across varied and incomplete MRI protocols. GlioMODA was trained and validated on the BraTS 2021 dataset (1251 training, 219 testing cases), systematically assessing performance across 11 clinically relevant MRI protocol combinations. Segmentation accuracy was evaluated using Dice similarity coefficients (DSC) and panoptic quality metrics. Volumetric accuracy was benchmarked against manual ground truth, and statistical significance was established via Wilcoxon signed‑rank tests with Benjamini–Yekutieli correction. GlioMODA demonstrated state-of-the-art segmentation accuracy across tumor subregions, maintaining robust performance with incomplete or heterogeneous MRI protocols. Protocols including both T1-weighted contrast-enhanced and T2-FLAIR sequences yielded volumetric differences vs manual ground truth that were not statistically significant for enhancing tumor (median difference 55 mm³, P = .157) and whole tumor (median difference −7 mm³, P = 1.0), and exhibited median DSC differences close to zero relative to the 4‑sequence reference protocol. Omitting either sequence led to substantial and significant volumetric errors. GlioMODA facilitates reliable, automated glioma segmentation using a streamlined 2‑sequence protocol (T1‑contrast + T2‑FLAIR), supporting clinical workflow optimization and broader implementation of quantitative volumetry compatible with RANO 2.0 criteria. GlioMODA is published as an open-source, easy-to-use Python package at https://github.com/BrainLesion/GlioMODA/.

Precise molecular characterization of glioblastoma (GB) is fundamental for accurate risk stratification and therapeutic planning. DNA methylation profiling reliably identifies key molecular features, including O(6)-methylguanine-DNA methyltransferase (MGMT) promoter methylation status and specific molecular subtypes, such as receptor tyrosine kinase (RTK) I and II, and the mesenchymal (MES) subtype. In this study, we investigated the hypothesized correlation between these molecular profiles and preferential tumor locations, which could reveal a link to underlying tumor biology. We analyzed 227 GB patients characterized by DNA methylation profiling. To map significant clusters of tumor occurrence across subtypes and subcomponents, we performed voxel-wise analysis of differential involvement, utilizing 500 permutations to correct for multiple comparisons. While uncorrected frequency differential maps suggested localization tendencies for the RTK I, RTK II, and MES subtypes, stringent statistical correction revealed only one robust association: the non-enhancing component of MES tumors showed significant clustering in the left frontal lobe, the insula, and the temporal lobe. Contrary to prior literature, we observed no significant hemispheric preference regarding MGMT promoter methylation status. Our findings challenge prior assumptions regarding the spatial distinctiveness of GB subtypes and highlight the need to further elucidate the mechanisms governing tumorigenesis and spatial growth patterns.