UCSF-BMSR: 560 Brain MRIs with 5,136 Annotated Metastases

560 brain MRIs from 412 patients with expert annotations of 5,136 brain metastases. The dataset consists of registered and skull-stripped T1 post-contrast, T1 pre-contrast, FLAIR, and subtraction images, along with voxelwise segmentations in NifTI format. It is a public, clinical, multimodal dataset from the UCSF Center for Intelligent Imaging.

Use Cases

• Train segmentation models for brain metastases based on expert voxelwise annotations.
• Develop multimodal image analysis algorithms based on T1, FLAIR, and subtraction MRI sequences.
• Research treatment response and tumor characteristics for stereotactic radiosurgery.
• Benchmark registration and skull-stripping preprocessing pipelines for clinical brain MRI.

Strengths

• Contains 5,136 expertly annotated brain metastases, providing a substantial ground truth for model training.
• Includes 560 multimodal MRI scans (T1 pre/post, FLAIR, subtraction) from 412 patients.
• Data is preprocessed (registered and skull-stripped), reducing initial processing burden.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.
• Row count is unknown, which may limit suitability assessment for specific tasks.

Provenance

Source

UCSF Center for Intelligent Imaging

Collection Method

Clinical brain MRIs collected from patients with brain metastases.

Time Range

null

Freshness

Last updated: unknown

Geography

null