ORIGINAL RESEARCH article
Front. Behav. Neurosci.
Sec. Pathological Conditions
This article is part of the Research TopicArtificial Intelligence for Behavioral Neuroscience: Unlocking mechanisms, modeling behavior, and advancing predictionView all articles
Customized SAM-Med3D with Multi-view Adapter and T2-FLAIR Mismatch Features for Glioma IDH Genotyping and Grading
Provisionally accepted
- 1Central South University, Changsha, China
- 2Xiamen University, Xiamen, China
- 3Renmin University of China, Beijing, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Objective: Gliomas, the most aggressive type of brain tumor, are infamous for their low survival rates. Tumor grading and isocitrate dehydrogenase (IDH) status are key prognostic biomarkers for gliomas, yet obtaining these markers usually relies on invasive methods such as biopsy. 6 As an effective and noninvasive alternative, multimodal MRIs can potentially reveal the spatial information and microenvironment of tumors. Low-grade gliomas as well as IDH-mutant gliomas often exhibit T2-FLAIR mismatch signals. Medical image foundational models are capable of exploring complex representations in medical images, and fine-tuning these models may further enhance the diagnosis of gliomas. Methods: We propose a multi-task network named MTSAM for glioma IDH genotyping and Grading simultaneously. MTSAM first uses dilated convolutions to simulate large-field convolutions, conducting an overview of the T2 and FLAIR images. Then, we employ convolutions for detailed exploration to weight T2 and FLAIR images, and subtract the weighted T2 and FLAIR images to obtain T2-FLAIR mismatch features. T2-FLAIR mismatch features are concatenated 16 with multimodal MRIs and input into the customized SAM-Med3D. Customized SAM-Med3D is fine-tuned by exploring the complementary information among multi-view information, including MRIs, handcrafted radiomics (HCR), and clinical features, and then extracts deep features for accurate IDH genotyping and grading. Results: MTSAM achieves AUCs of 92.38% and 94.31% for glioma IDH typing and grading on the UCSF-PDGM dataset, respectively, and AUCs of 91.56% and 93.37% on the BraTS2020 dataset, outperforming other methods. Additionally, we utilize Grad-CAM to perform a visualization analysis of the attention maps of MTSAM, demonstrating its potential in non-invasive glioma diagnosis. Conclusion: The proposed method demonstrates that we can effectively fuse multi-view non-invasive information and fully explore the knowledge learned by medical image foundational
Keywords: deep learning, Glioma, Grading, IDH Genotyping, Medical Foundational Model, Multi-modal MRIs
Received: 15 Sep 2025; Accepted: 15 Dec 2025.
Copyright: © 2025 Li, Li, Hu, Zhang, Wang and Yang. 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) or licensor 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.
* Correspondence: Yunyi Hu
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.