PMC-VQA: Over 1.7 Million Visual Question Answering Samples from Clinical Figures

Over 1,700,000 Visual Question Answering samples derived from figures and charts in biomedical articles from PubMed Central. This preliminary release by DermaVLM was last updated on Hugging Face in August 2025. A full dataset card and accompanying research paper are currently in preparation.

Use Cases

• Train visual question answering models based on clinical figures and charts.
• Benchmark multimodal AI performance on biomedical visual reasoning tasks.
• Fine-tune large language models for interpreting medical imagery described in the dataset.
• Develop educational tools for medical training using annotated figure-question pairs.

Strengths

• Contains over 1,700,000 VQA samples, indicating a large-scale resource.
• Derived from a reputable source of biomedical literature, PubMed Central (PMC).

Limitations

• This is a preliminary release; a full dataset card and research paper are not yet available.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count beyond the total sample number is unknown, which may limit suitability assessment.

Provenance

Source

PubMed Central (PMC)

Collection Method

Derived from figures and charts in biomedical articles.

Freshness

Last updated 2025-08-14 14:18:49; freshness should be verified.