相关实验视频

Updated: Feb 24, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

大型语言模型的比较性能在眼科医生推选中

Pedro Cardoso-Teixeira¹, João Alves Ambrósio¹, Mariana Garcia¹

¹Ophthalmology Department, Unidade Local de Saúde Entre o Douro e Vouga, Santa Maria da Feira, PRT.

|February 23, 2026

概括

先进的语言模型 (LLM) 在分类眼科推方面表现有前途,通过上下文学习提高准确性. 虽然对常见病例有效,但对于罕见或模两可的投诉,性能下降,需要人类监督.

科学领域:

眼科医生眼科眼科
人工智能的人工智能
自然语言处理自然语言处理.

背景情况:

医疗推的自动分类可以提高效率.
大型语言模型 (LLM) 为分析临床文本提供了潜力.

研究的目的:

评估五个先进的LLM的分类准确性和一致性,以分类葡萄牙眼科推小组.
评估监督的语境学习对这些推的LLM绩效的影响.

主要方法:

3,831个葡萄牙眼科推小贴士被五个LLM分类 (ChatGPT 4o,ChatGPT 5.1,困惑Pro,克劳德·索内特4.5,克劳德·奥普斯4.1).
最初使用了零射击提示策略,随后进行了上下文学习阶段 (957个标记的示例).
计算了分类的准确性,一致性和Fleiss的kappa协议.

主要成果:

基线准确率平均为68.7%,在上下文学习后提高到73.4%.
聊天GPT 5.1实现了峰值准确度 (79.5%);聊天GPT 4o显示了最大的一致性增长 (66.8%至93.8%).
对于常见类别 (如糖尿病查) 的表现超过了90%,但对于罕见/模两可的投诉,表现较低. 中期协议的范围从中度到实质 (κ = 0.462-0.801).

结论:

先进的LLM在解释眼科推方面显示出显著的潜力,通过上下文学习显著提高了准确性和一致性.
对于罕见或模两可的转介类型,LLM的表现较低,这凸显了需要仔细实施的需要.
在眼科中,LLM可以作为可扩展的,低成本的诊断辅助工具,而这取决于人类监督和进一步的临床验证.

关键词:

人工智能的人工智能是人工智能.克劳德: 克劳德是我们的朋友.临床决策支持临床决策支持 gpt gpt是什么意思在上下文学习学习.大型语言模型.眼科的分类对象.复杂性是为了复杂性.转介分类是指转介分类的分类.

更多相关视频

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

相关概念视频

Improving Translational Accuracy

Improving Translational Accuracy

Improving Translational Accuracy

Improving Translational Accuracy

Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...