在RCT中使用GPT? 使用人工智能来确定遵守临床试验报告准则
James G Wrightson1, Paul Blazey2, David Moher3
1Department of Physical Therapy, The University of British Columbia Faculty of Medicine, Vancouver, British Columbia, Canada.
大型语言模型 (AI-LLMs) 在评估临床试验报告准则的遵守方面表现有前途. 在体育医学试验中,GPT-4实现了90%的准确性,而微调的Llama 2在体育医学试验中达到83%的准确性.
科学领域:
- 医学研究方法论医学研究方法论.
- 医疗保健中的人工智能
- 临床试验报告标准
背景情况:
- 遵守报告准则对于临床试验质量至关重要.
- 以前试图改善坚持的尝试产生了不一致的结果.
- AI-LLMs提供了一个潜在的新方法来评估报告准则的合规性.
研究的目的:
- 评估AI-LLMs在确定报告准则合规性的准确性.
- 评估OpenAI GPT-4和Meta Llama 2模型的性能.
- 分析体育医学和运动科学临床试验报告的合规性.
主要方法:
- 对113份临床试验报告的回顾性数据分析.
- 使用了OpenAI GPT-4 Turbo,GPT-4 Vision和Meta Llama 2 70B的模型. 这是一个非常好的选择.
- 模型被要求回答报告指南问题;Llama 2进行了微调.
主要成果:
- 对于文章文本,GPT-4 Turbo实现了90%的准确性 (F1分数=0.89).
- 微调的Llama 2实现了83%的准确性 (F1得分=0.84).
- 对于参与者流程图,GPT-4视觉实现了100%的准确性,但在缺失的细节方面遇到了困难 (57%的准确性).
结论:
- 无论是GPT-4还是精细调整的Llama 2,都显示出评估报告准则遵守的潜力.
- 建议进一步开发高效的,开源的AI-LLM.
- 提高AI-LLM在这个领域的准确性的方法需要进一步探索.
更多相关视频
08:36Author Spotlight: Evaluating the Adjuvant Efficacy and Safety of Angong Niuhuang Pill in Viral Encephalitis Treatment
Published on: April 19, 2024
06:05The Participant-Reported Implementation Update and Score PRIUS: A Novel Method for Capturing Implementation-Related Data Over Time
Published on: February 19, 2021
相关概念视频
Clinical Trials
There are four phases in a clinical trial. A phase one...
Introduction to Documentation and Reporting
Nursing documentation records essential information and details regarding a patient's care and treatment in written or electronic form. It is a critical aspect of nursing practice that involves documenting assessments, interventions, outcomes, and other relevant details about a patient's health status.
Documentation maps the patient's health journey by creating a comprehensive...
Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...
Guidelines for Nursing Documentation I
Factual:
The following points emphasize the significance of upholding accurate and unbiased documentation in healthcare.
Data Reporting and Recording
Blind Procedures
