Attention-Based Dual-Branch Deep Learning for Overall Survival Prediction in Patients with Brain Metastases Using MRI: A Multicenter Retrospective Study
Cancers · 18.11.2026 · 18 (22) · 3693
DOI: 10.3390/cancers18223693 ↗Kurz gefasst
Fachartikel in der Zeitschrift Cancers, online erschienen am 18.11.2026. Verfasst von Ji Zhang und 10 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation, Zhejiang Natural Science Foundation, Key project of Zhejiang Provincial Health Science and Technology Program.
Abstract (Originalsprache)
Objectives
To develop and validate an attention-based deep learning (DL) model to predict overall survival (OS) in patients with brain metastases (BM) using pre-treatment magnetic resonance image (MRI).
Methods
A total of 216 BM patients treated with radiotherapy or neurosurgical resection were retrospectively enrolled from two hospitals and divided into training (n = 128), internal validation (n = 56), and external validation (n = 32) cohorts. Three feature aggregation methods and seven radiomics models were developed to predict OS. A DL model named DeepBMSurv, incorporating a cross-attention mechanism, was built using pre-treatment MR images. Its prognostic performance was compared with radiomics models and graded prognostic assessment (GPA). A nomogram integrating GPA, radiomics score (Rad-score) and DL score (DL_RiskScore) was also developed. Model performance was assessed by the concordance index (C-index).
Results
Among the radiomics models, the RadScoremean model showed the most stable discrimination across cohorts, with C-index values of 0.685, 0.672, and 0.671 in the training, internal validation, and external validation cohorts. DeepBMSurv achieved numerically higher C-indices than both the radiomics model and GPA model, with C-indices of 0.806 vs. 0.685 vs. 0.620 in the training cohort, 0.755 vs. 0.672 vs. 0.628 in the internal validation cohort, and 0.733 vs. 0.671 vs. 0.579 in the external validation cohort. The nomogram integrating DL_RiskScore, Rad-score, and GPA achieved C-index values of 0.809, 0.764 and 0.733 in training, internal validation, and external validation cohorts, respectively.
Conclusions
A DL model based on pre-treatment MRI can effectively predict survival outcomes in patients with BM, offering a potential tool for individualized prognostic risk stratification.
Förderung
- National Natural Science Foundation
- Zhejiang Natural Science Foundation
- Key project of Zhejiang Provincial Health Science and Technology Program
- Major project of the Wenzhou Science and Technology Bureau
- project of the Wenzhou Science and Technology Bureau
- Zhejiang Engineering Research Center for Innovation and Application of Intelligent Radiotherapy Technology
- Zhejiang-Hong Kong Precision Theranostics of Thoracic Tumors Joint Laboratory
- Wenzhou Key Laboratory of Basic Science and Translational Research of Radiation Oncology
Zitieren
APA
Zhang, J., He, J., Guo, M., Zhang, Z., Mu, Y., Shi, B., Wu, J., Lv, X., Yu, W., Xie, C., & Jin, X. (2026). Attention-Based Dual-Branch Deep Learning for Overall Survival Prediction in Patients with Brain Metastases Using MRI: A Multicenter Retrospective Study. Cancers, 18(22), 3693. https://doi.org/10.3390/cancers18223693
Vancouver
Zhang J, He J, Guo M, Zhang Z, Mu Y, Shi B, et al. Attention-Based Dual-Branch Deep Learning for Overall Survival Prediction in Patients with Brain Metastases Using MRI: A Multicenter Retrospective Study. Cancers. 2026;18(22):3693. doi:10.3390/cancers18223693
BibTeX
@article{Zhang2026,
author = {Zhang, Ji and He, Jiaqian and Guo, Mingrou and Zhang, Zhenhua and Mu, Yiran and Shi, Bowen and Wu, Jianping and Lv, Xiuling and Yu, Wenliang and Xie, Congying and Jin, Xiance},
title = {Attention-Based Dual-Branch Deep Learning for Overall Survival Prediction in Patients with Brain Metastases Using MRI: A Multicenter Retrospective Study},
journal = {Cancers},
year = {2026},
volume = {18},
number = {22},
pages = {3693},
doi = {10.3390/cancers18223693}
}
EUVIMED