BrainDINO foundation model achieves generalizable clinical representation learning from brain MRIs

Researchers have developed BrainDINO, a self-supervised foundation model trained on approximately 6.6 million unlabeled brain MRI slices. This model demonstrates strong generalization capabilities across a wide array of clinical tasks, including tumor segmentation, disease classification, and survival prediction. BrainDINO consistently matched or surpassed existing self-supervised baselines, particularly excelling in scenarios with limited labeled data. The findings suggest that large-scale, slice-wise self-supervised learning can create a unified brain MRI representation beneficial for diverse neuroimaging analyses. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Establishes a scalable foundation for data-efficient brain imaging analysis, potentially accelerating research and clinical applications.

RANK_REASON This is a research paper describing a new foundation model for brain MRI analysis.

Read on arXiv cs.AI →

• BrainDINO
• arXiv
• Hugging Face

• paper
• model release