UniMed-5M: Unified Medical Multimodal Dataset for 8 Imaging Modalities

General-Medical-AI released UniMed-5M, a dataset for training and evaluating the UniMedVL model. The dataset is organized into parts containing Parquet metadata files for different medical imaging modalities. The dataset page was last updated on February 5, 2026.

Use Cases

• Train medical image generation models based on multimodal medical data
• Evaluate model performance across 8 medical imaging modalities
• Develop unified medical vision-language models based on the dataset's structure

Strengths

• Designed for a unified model covering 8 medical imaging modalities
• Associated with a published paper (arXiv:2510.15710) and a project page
• Last updated on 2026-02-05, indicating recent maintenance

Limitations

• Row count, file formats, and column-level documentation are unknown
• Description metadata is limited; actual data quality requires manual inspection after download
• License information is unknown, which may restrict usage

Provenance

Source

General-Medical-AI

Collection Method

Likely aggregated for training a unified medical multimodal model.

Time Range

null

Freshness

Last updated 2026-02-05 13:29:39

Geography

null