使用大型语言模型新建结直肠患者教育材料:快速工程是提高可读性的关键
India E Ellison1, Wendelyn M Oslock2, Abiha Abdullah3
1Department of Surgery, University of Alabama at Birmingham, AL.
Surgery
|January 5, 2025
概括
大型语言模型 (LLM) 可以改善患者教育材料. 与其他经过测试的LLMs不同,ChatGPT具有基于指标的提示,显著提高了可读性,符合六年级水平的指导方针.
科学领域:
- 医学教育 医学教育
- 卫生识字 卫生识字
- 医疗保健中的人工智能
背景情况:
- 患者教育对于临床结果和减少健康差异至关重要.
- 传统的患者教育材料可能需要大量的劳动力来创建和更新.
- 大型语言模型 (LLM) 为可访问的患者教育材料开发提供了潜在的解决方案.
研究的目的:
- 将现有的结直肠外科手术患者教育材料的可读性与LLMs产生的材料进行比较.
- 评估不同提示对LLM生成材料可读性的影响.
主要方法:
- 收集了52个基线结直肠外科手术教育材料.
- 在ChatGPT 3.5和Perplexity.中使用了三种提示类型 (基本类型,代类型,基于指标类型).
- 使用未配对的t测试计算和比较可读性指标 (弗莱什-金凯德分级,阅读方便,修改分级).
主要成果:
- 现有的材料比LLM生成的内容长得多.
- 基线材料未能满足六年级的可读性标准.
- 与基线相比,使用基于指标的提示的ChatGPT显著改善了可读性 (等级级5.2,易读70.5), (等级级8.1,易读60.4).
- 困惑的可读性通常较低,除了基于metric的提示符之外没有显著的改善.
结论:
- 大多数LLM生成的材料不符合可读性建议.
- 当被提示特定可读性目标时 (基于metric的提示),ChatGPT显示可读性得到改善,并且符合六年级级的指导方针.
- 快速工程对于优化LLM性能在生成可访问的患者教育材料方面至关重要.
相关概念视频
01:21Health Literacy
4.0K
Health literacy is an individual's or a community's capacity to comprehend, receive, read, and use relevant healthcare information and services. The World Health Organization (WHO, 2018) defines health literacy as the cognitive and social skills that determine the ability of individuals to gain access to, understand, and use information in ways that promote and maintain good health. As a result, the WHO helps individuals manage long-term health concerns, participate in preventative...
4.0K
01:21Inflammatory Bowel Disease V: Surgical Management
116
Surgical interventions for inflammatory bowel disease (IBD), which includes ulcerative colitis and Crohn's disease, are essential in managing symptoms and addressing complications. The selection of surgical procedures is contingent upon the specific conditions and complications that stem from these illnesses.
Here are some common surgical interventions for IBD:
Here are some common surgical interventions for IBD:
116