Foundation Model Based on Routine Magnetic Resonance Imaging for Brain Tumor Molecular Profiling and Progression Prediction
JCO Precision Oncology · 2026-02 · 10 (2)
DOI: 10.1200/po-25-00930 ↗Kurz gefasst
Fachartikel in der Zeitschrift JCO Precision Oncology, 2026-02. Verfasst von Junxian Li und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Blood Transfusion, Key Laboratory of Cancer Prevention and Therapy, Tianjin, National Clinical Research Center for Cancer, Tianjin's Cl…. Hauptthemen laut MEDLINE: Brain Neoplasms, Magnetic Resonance Imaging, Glioma. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 29 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.
Abstract (Originalsprache)
PURPOSE To build a self-supervised magnetic resonance imaging (MRI) foundation model from routine clinical scans and to test whether it can support key glioma-related applications, including post-therapy imaging outcome characterization and molecular marker inference. MATERIALS AND METHODS We created the Unified Multimodal Brain Imaging Foundation (UMBIF) model and pretrained it in a self-supervised manner using 51,029 routine brain MRI examinations collected across multiple institutions. Pretraining used a hybrid objective that couples masked-image reconstruction with contrastive representation learning to encourage anatomically and clinically informative embeddings. The pretrained UMBIF encoder was then adapted to downstream multicenter data sets to predict (1) post-treatment radiographic outcomes and (2) molecular biomarkers, including IDH mutation, MGMT promoter methylation, and 1p/19q codeletion. Performance was benchmarked against commonly used convolutional networks and traditional machine learning classifiers, using accuracy, sensitivity, specificity, and receiver operating characteristic-AUC as primary metrics. RESULTS Relative to self-supervised initialization derived from natural-image corpora or from approaches emphasizing only large tumor-area crops (self-supervised learning [SSL]-ImageNet and SSL-Cerebral), the UMBIF encoder-decoder design captured richer, more task-relevant features and consistently improved downstream discrimination. The best pretrained model achieved an accuracy of 0.899 (AUC, 0.815) for post-treatment radiographic outcome characterization. For molecular profiling, it reached accuracies/AUCs of 0.898/0.916 for 1p/19q codeletion, 0.829/0.896 for IDH mutation status, and 0.905/0.859 for MGMT promoter methylation, indicating strong potential utility in clinical decision support. CONCLUSION UMBIF showed robust transferability to both post-therapy imaging assessment and molecular status prediction in glioma. By leveraging large-scale self-supervised pretraining to boost performance while reducing dependence on manual annotations, the framework may facilitate more efficient and reliable diagnostic workflows.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Blood Transfusion, Key Laboratory of Cancer Prevention and Therapy, Tianjin, National Clinical Research Center for Cancer, Tianjin's Clinical Research Center for Cancer, Tianjin Medical University Cancer I…
- Department of Diagnostic and Therapeutic Ultrasonography, Key Laboratory of Cancer Prevention and Therapy, Tianjin, National Clinical Research Center for Cancer, Tianjin's Clinical Research Center for Cancer, Tianjin Me…
- Department of Cancer Epidemiology and Biostatistics, Key Laboratory of Molecular Cancer Epidemiology, Tianjin, National Clinical Research Center for Cancer, Tianjin's Clinical Research Center for Cancer, Tianjin Medical…
- Department of Neurosurgery and Neuro-oncology, Key Laboratory of Cancer Prevention and Therapy, Tianjin, National Clinical Research Center for Cancer, Tianjin's Clinical Research Center for Cancer, Tianjin Medical Unive…
- Department of Molecular Imaging and Nuclear Medicine, Key Laboratory of Cancer Prevention and Therapy, Tianjin, National Clinical Research Center for Cancer, Tianjin's Clinical Research Center for Cancer, Tianjin Medica…
ORCID-Profile: Qian Su
Zitieren
APA
Li, J., Liu, R., Xing, Y., Gao, X., Yin, Q., & Su, Q. (2026). Foundation Model Based on Routine Magnetic Resonance Imaging for Brain Tumor Molecular Profiling and Progression Prediction. JCO Precision Oncology, 10(2). https://doi.org/10.1200/po-25-00930
Vancouver
Li J, Liu R, Xing Y, Gao X, Yin Q, Su Q. Foundation Model Based on Routine Magnetic Resonance Imaging for Brain Tumor Molecular Profiling and Progression Prediction. JCO Precision Oncology. 2026;10(2). doi:10.1200/po-25-00930
BibTeX
@article{Li2026,
author = {Li, Junxian and Liu, Renhe and Xing, Yuchen and Gao, Ximin and Yin, Qiang and Su, Qian},
title = {Foundation Model Based on Routine Magnetic Resonance Imaging for Brain Tumor Molecular Profiling and Progression Prediction},
journal = {JCO Precision Oncology},
year = {2026},
volume = {10},
number = {2},
doi = {10.1200/po-25-00930}
}
EUVIMED