159,549 new question-answer pairs form the Kvasir-VQA-x1 dataset, a large-scale benchmark for medical visual question answering in gastrointestinal endoscopy. SimulaMet created this multimodal dataset to advance robust MedVQA systems. The dataset was featured in the MediaEval Medico 2025 Challenge and was last updated on Hugging Face in August 2025.
Use Cases
- Benchmarking medical visual question answering models based on the 159,549 QA pairs.
- Training multimodal AI for medical reasoning in gastrointestinal endoscopy based on the described MedVQA focus.
- Developing robust clinical decision support systems based on the dataset's design for gastrointestinal endoscopy analysis.
Strengths
- Contains 159,549 new question-answer pairs, indicating a substantial scale for training and evaluation.
- Designed specifically for benchmarking medical visual question answering, providing a clear purpose.
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
- SimulaMet
- Freshness
- Last updated 2025-08-27 12:50:11; freshness should be verified.