X-ray Clinical Reports Without Findings for Medical NLP, 171k Images

74,154 patient records and 171,040 medical images form this collection of X-ray reports without clinical findings. Authored by InfoBayAI, the dataset is designed to support the development of diagnostic AI systems. It captures authentic imaging characteristics such as scanner variability and acquisition protocols, along with clinical narratives.

Use Cases

• Train models to identify normal X-ray characteristics based on the large-scale collection of reports without findings.
• Develop NLP systems for parsing and structuring clinical narratives mentioned in the dataset description.
• Benchmark AI models for medical image classification using the described authentic imaging characteristics like scanner variability.
• Support research in healthcare AI by providing a dataset focused on the absence of pathology in radiology reports.

Strengths

• Contains data from 74,154 patients, providing a substantial sample size.
• Includes 171,040 medical images, offering a large-scale resource for model training.
• Designed to capture authentic imaging characteristics like scanner variability and acquisition protocols.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the text report component is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

InfoBayAI

Freshness

Last updated 2026-06-02 05:45:45; freshness should be verified.