PatSnap/clinical-trial

全球临床试验结构化情报数据集提供了900个深度解析的临床试验记录，这些记录按比例覆盖了所有开发阶段（I期到IV期，以及观察性研究），并特别丰富了早期阶段（I/II期）试验的代表性。与原始注册文本不同，每个记录提取了关键参数，包括研究目的、干预组、试验持续时间和主要终点。该数据集是医学NLP模型、临床试验结果预测和全球研发趋势分析的重要资源。关键特征包括：阶段分层覆盖（记录涵盖所有临床阶段，特别丰富了早期阶段试验）、结构化试验设计（包括干预组数量、持续时间和国家组数量等细节）以及结果与验证跟踪（跟踪试验是否有结果及其验证状态）。

The Global Clinical Trials Structured Intelligence dataset provides 900 deeply parsed clinical trial records drawn proportionally across all development phases (Phase I through IV, plus observational studies), with enriched representation of early-phase trials. Moving beyond raw registry text, each record extracts critical parameters including study purpose, intervention arms, trial duration, and primary endpoints. An essential resource for Medical NLP models, clinical trial outcome prediction, and global R&D trend analysis. Key Features: Phase-Stratified Coverage: Records span all clinical phases with deliberate enrichment of early-phase (Phase I/II) trials. Structural Trial Design: Details on arm_intervention_count, duration, and group_country_count. Outcome & Validation Tracking: Tracks whether the trial has results (if_ct_result) and its validation status (if_validation).