Upper Extremity X-Ray Images with 13 Pathology Labels

Over 1,200,000 digital X-ray studies focus on chest, shoulder, elbow, and wrist joints. The dataset, created by UniDataPro, includes 13 labeled pathologies for medical imaging and machine learning research.

Use Cases

• Train a computer vision model to classify the 13 pathology labels from X-ray images of upper extremity joints.
• Analyze the distribution of pathologies across different joint types (chest, shoulder, elbow, wrist).
• Develop segmentation models for joint structures using the large collection of over 1.2 million X-ray studies.
• Compare imaging features between the 200,000+ studies with protocol and the 1,000,000+ studies without protocol.

Strengths

• Large scale with over 1,200,000 X-ray studies.
• Includes 13 specific pathology labels for supervised learning tasks.
• Focuses on a defined anatomical scope: chest, shoulder, elbow, and wrist joints.

Limitations

• Unknown sample size per pathology, which may indicate class imbalance.
• Dataset composition includes a majority (over 1 million) of studies without protocol, which may limit certain analyses.
• No information on data provenance, acquisition standards, or patient demographics.

Provenance

Source

UniDataPro via Hugging Face

Freshness

Last updated on April 4, 2025.