Medical Records Parsing Validation Set with 288 Indian Lab Reports and Prescriptions

288 images of laboratory reports and prescriptions from Indian healthcare settings, curated by ekacare for validating AI systems that parse medical documents. The dataset was last updated on Hugging Face on July 24, 2025. It is designed to enable the digitization of healthcare data while maintaining clinical accuracy.

Use Cases

• Validate optical character recognition (OCR) and information extraction models based on diverse medical document formats.
• Benchmark the performance of clinical natural language processing (NLP) systems based on real-world laboratory reports and prescriptions.
• Train models to handle the structural variability of medical forms based on the described diverse templates.
• Assess the clinical accuracy of automated data digitization pipelines based on the dataset's stated purpose.

Strengths

• Contains 288 carefully selected medical document images.
• Represents diverse formats and templates encountered in Indian healthcare settings.
• Explicitly designed for evaluating clinical accuracy in information extraction.

Limitations

• Row count, column definitions, and file formats are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset's geographic focus on India may limit generalizability to other healthcare systems.

Provenance

Source

ekacare via Hugging Face

Collection Method

Likely collected and curated from real-world medical documents.

Freshness

Last updated 2025-07-24 16:57:02

Geography

India