CT-Outcome (TrialBench v2)
Updated 2024 trial outcome benchmark with temporal splits.
Composite
73.4
Experimental validation
Clinical
Stages
Phase IIPhase III
Modalities
cross-modality
Task types
trial-outcome
Size
trials: 25,000
License
MIT
First release
2024-07
Last updated
2025-03
Official site
Leaderboard
→ leaderboard
Dataset
Code / GitHub
HuggingFace
→ HF
Paper
TrialBench: Multi-Modal Artificial Intelligence-Ready Clinical Trial Datasets · Chen J, Hu Y, Wang Y, et al. · 2024 · paper · doi:10.48550/arXiv.2407.00631 · 25 citations
Flags
none
Experts
Groups
Hosted by
Related benchmarks
Rubric (7-criterion)
rigor
4
coverage
3
maintenance
4
adoption
2
quality
4
accessibility
5
industry_relevance
4
Notes
Temporal splits are key improvement.