Research map / Label-efficient and robust learning / Semi-supervised learning
Extracting PICO Sentences from Clinical Trial Reports using Supervised Distant Supervision
Byron C. Wallace, Joël Kuiper, Aakash Sharma, Mingxi (Brian) Zhu, Iain J. Marshall
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace direct-only training with SDS (add distant supervision filtered by learned mapping) | Direct only · PICO Population (retrospective) | Mean AUC: 0.904 -> 0.947 | Improved |
| replace direct-only training with SDS | Direct only · PICO Outcomes (retrospective) | Mean AUC: 0.837 -> 0.900 | Improved |
| add SDS filtering of candidate sentences to plain distant supervision | DS · PICO Population (retrospective) | Mean AUC: 0.941 -> 0.947 | Improved |
| add SDS filtering to plain distant supervision | DS · PICO Interventions (retrospective) | Precision@20: 0.172 -> 0.172 | No clear effect |
| add SDS filtering to plain distant supervision | DS · PICO Population (retrospective) | Precision@10: 0.202 -> 0.212 | Improved |
| add SDS filtering to plain distant supervision | DS · PICO Population (retrospective) | Mean NDCG@20: 0.484 -> 0.548 | Improved |
| use SDS instead of Nguyen pooling of direct and distant models | Nguyen · PICO Population (retrospective) | Mean AUC: 0.917 -> 0.947 | Improved |
| use SDS instead of Nguyen pooling | Nguyen · PICO Interventions (retrospective) | Precision@3: 0.419 -> 0.389 | Hurt |
| use SDS instead of Nguyen pooling | Nguyen · PICO Outcomes (retrospective) | Mean NDCG@20: 0.339 -> 0.333 | Hurt |
| use SDS instead of Nguyen pooling | Nguyen · PICO Interventions (prospective) | Precision@3: 0.854 -> 0.903 | Improved |
| use SDS instead of Nguyen pooling | Nguyen · PICO Population (prospective) | Precision@3: 0.907 -> 0.927 | Improved |
| use SDS instead of Nguyen pooling | Nguyen · PICO Outcomes (prospective) | Precision@3: 0.880 -> 0.887 | Improved |
Closest papers by meaning
- MedAlign: A Clinician-Generated Dataset for Instruction Following with Electronic Medical Records (AAAI 2024 · similarity 0.85)
- EHRNote-ChatQA: A Benchmark for Evidence-Grounded Multi-Turn Clinical Question Answering over Longitudinal Discharge Summaries (NeurIPS 2026 · similarity 0.83)
- DoAtlas-1: A Causal Compilation Paradigm for Clinical AI (NeurIPS 2026 · similarity 0.83)
- SLIDERS: Systematic Reviews via Automated Evidence Synthesis and Reconciliation (NeurIPS 2026 · similarity 0.83)
- EHRNoteQA: An LLM Benchmark for Real-World Clinical Practice Using Discharge Summaries (NeurIPS 2024 · similarity 0.83)
- Context Clues: Evaluating Long Context Models for Clinical Prediction Tasks on EHR Data (ICLR 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
