RSNA Abdominal Trauma Detection: 4,000+ CT Scans from 23 Global Sites

More than 4,000 computed tomography (CT) exams for abdominal trauma detection, sourced from 23 sites across 14 countries on six continents. The Radiological Society of North America (RSNA) collected this dataset to support the development of AI tools for expediting diagnosis in emergency settings. The dataset includes cases with various abdominal injuries and a roughly equal number of cases without injury.

Use Cases

• Train image classification models to detect the presence of abdominal trauma based on CT scans.
• Develop segmentation models to identify and localize injuries to specific organs like the liver, spleen, or kidneys.
• Benchmark AI-assisted diagnostic tools for expediting trauma assessment in clinical settings.

Strengths

• Includes more than 4,000 CT exams, providing a substantial image collection.
• Data is sourced from 23 sites in 14 countries across six continents, suggesting geographic diversity.
• Contains a balanced set of cases with various abdominal injuries and cases without injury.

Limitations

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

Provenance

Source

Radiological Society of North America (RSNA)

Collection Method

Imaging data collected from 23 medical sites globally to create a ground truth dataset.

Time Range

null

Freshness

Last updated is unknown.

Geography

Data sourced from 23 sites in 14 countries across six continents.