MedMosaic: Medical Audio Recordings with Clinical Questions and Answers

An audio question-answering dataset for clinical and healthcare contexts created by CentificAIResearch. It contains audio recordings paired with clinical questions and answers across multiple QA types, designed to benchmark audio-language models on medical reasoning tasks. The dataset is organized into 7 subfolders.

Use Cases

• Benchmarking audio-language models on medical reasoning tasks based on the described clinical question-answering structure.
• Training models for audio-based clinical decision support based on the described pairing of audio recordings with clinical questions.
• Evaluating model performance on diverse medical QA types based on the description of multiple QA types present.

Strengths

• Designed specifically for evaluating audio understanding in clinical contexts.
• Contains multiple QA types, suggesting diversity in reasoning tasks.
• Organized into 7 subfolders, suggesting a structured data layout.

Limitations

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

Provenance

Source

CentificAIResearch

Freshness

Last updated 2026-06-11 19:23:35; freshness should be verified.