Two new research papers explore the use of AI in clinical trial matching, addressing challenges in patient enrollment and decision accountability. The first paper introduces VERDICT, a neurosymbolic agent that uses Satisfiability Modulo Theories (SMT) to ensure consistent policy application and accountable decision-making, outperforming existing LLM-only and neurosymbolic baselines. The second paper presents TrialGPT 2.0, an AI-assisted system evaluated across multiple oncology settings, which improved clinician efficiency by 55% and identified additional trial opportunities for 90.9% of patients, while also introducing a new dataset, NIH-TrialBench, for reproducibility. AI
IMPACT These systems demonstrate AI's potential to improve efficiency and accuracy in high-stakes medical decision-making, potentially accelerating drug development and patient access to trials.
RANK_REASON Two academic papers published on arXiv detailing new AI systems and methodologies for clinical trial matching.
- NIH Institutes and Centers
- NIH-TrialBench
- TrialGPT 2.0
- United States National Institutes of Health
- arXiv
- Hugging Face
- MaxSMT
- Satisfiability Modulo Theories
- SIGIR 2016 Workshop WebQA II
- TREC 2021
- VERDICT
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →