Accuracy of Large Language Models to Identify Stroke Subtypes Within Unstructured Electronic Health Record Data.

医学 冲程(发动机) 电子健康档案 非结构化数据 健康档案 自然语言处理 大数据 数据挖掘 医疗保健 机械工程 计算机科学 工程类 经济 经济增长
作者
Dylan Owens,Danh Q. Nguyen,Michael Dohopolski,Justin F. Rousseau,Eric D. Peterson,Ann Marie Návar
出处
期刊:PubMed
标识
DOI:10.1161/strokeaha.125.051993
摘要

While International Classification of Diseases, Tenth Revision codes suffice for identifying stroke events in surveillance, accurately classifying stroke types and subtypes using electronic health records remains challenging due to limitations in structured data. This often necessitates manual review of clinical documentation. This study evaluated whether a large language model, GPT-4o, can accurately identify stroke types and subtypes from unstructured clinical notes. We implemented a retrieval-augmented generation framework with GPT-4o to classify stroke types (ischemic versus hemorrhagic) and ischemic stroke subtypes using electronic health records data. The American Heart Association Get With The Guidelines-Stroke registry served as the gold standard. Model development used a 20% subset of Get With The Guidelines-Stroke-linked data from UT Southwestern Medical Center, with the remaining 80% reserved for testing. External validation used data from the Parkland Health and Hospital System. A total of 4123 stroke hospitalizations from January 2019 to August 2023 were included (UT Southwestern Medical Center: n=2047; Parkland Health and Hospital System: n=2076). Three prompting strategies-zero-shot Chain-of-Thought, expert-guided, and instruction-based-were evaluated. Predictions of GPT-4os were compared with classifications made by trained abstractors contributing to the Get With The Guidelines-Stroke registry. In the external validation set, 79.6% of patients had ischemic stroke and 20.4% hemorrhagic. GPT-4o achieved 98% accuracy (95% CI, 0.97-0.99) in classifying stroke type, where accuracy reflects the overall proportion of correctly classified patients. Sensitivity was 0.98 (95% CI, 0.97-0.99), and specificity was 0.97 (95% CI, 0.96-0.98). For ischemic stroke subtypes, sensitivity ranged from 0.40 (95% CI, 0.31-0.49) for cryptogenic to 0.95 (95% CI, 0.93-0.97) for small-vessel occlusion. Specificity ranged from 0.94 (95% CI, 0.92-0.96) for large-artery atherosclerosis to 0.98 (95% CI, 0.97-0.99) for cardioembolism. Zero-shot Chain-of-Thought prompting-requiring minimal human input-performed comparably to more labor-intensive strategies. Consistency analysis revealed >99% agreement across repeated queries. GPT-4o demonstrated strong accuracy in classifying stroke types but faced challenges with ischemic subtypes.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
落水发布了新的文献求助20
2秒前
an发布了新的文献求助10
3秒前
allezallez完成签到,获得积分10
5秒前
狮子卷卷完成签到,获得积分0
5秒前
儒雅的蜜粉完成签到,获得积分10
7秒前
yoyoya完成签到 ,获得积分10
7秒前
CipherSage应助冷艳的火龙果采纳,获得10
8秒前
masterchen完成签到,获得积分10
9秒前
苏苏完成签到,获得积分10
10秒前
10秒前
3333完成签到,获得积分20
11秒前
星辰完成签到,获得积分10
12秒前
静静小可爱完成签到,获得积分10
12秒前
沉默的友安完成签到 ,获得积分10
15秒前
自然的哈密瓜完成签到,获得积分10
18秒前
柴桑青木应助天天向上采纳,获得30
18秒前
共享精神应助宁霸采纳,获得10
18秒前
喜悦的绮露完成签到,获得积分10
19秒前
有魅力的沧海完成签到 ,获得积分10
20秒前
20秒前
21秒前
熙梓日记完成签到,获得积分10
23秒前
百甲完成签到,获得积分10
24秒前
26秒前
XU2025完成签到 ,获得积分10
27秒前
Owen应助宁霸采纳,获得10
27秒前
123完成签到,获得积分10
27秒前
35秒前
0008完成签到,获得积分10
37秒前
38秒前
不配.应助an采纳,获得10
40秒前
可别熬夜了完成签到 ,获得积分20
41秒前
酷波er应助淡淡猕猴桃采纳,获得10
42秒前
外向海蓝发布了新的文献求助10
43秒前
Dream发布了新的文献求助10
44秒前
zhouleiwang完成签到,获得积分10
45秒前
夏傥完成签到,获得积分10
45秒前
msl2023完成签到,获得积分10
47秒前
zero完成签到,获得积分10
48秒前
49秒前
高分求助中
(应助此贴封号)【重要!!请各位详细阅读】【科研通的精品贴汇总】 10000
Voyage au bout de la révolution: de Pékin à Sochaux 700
First Farmers: The Origins of Agricultural Societies, 2nd Edition 500
Simulation of High-NA EUV Lithography 400
Assessment of adverse effects of Alzheimer's disease medications: Analysis of notifications to Regional Pharmacovigilance Centers in Northwest France 400
The Rise & Fall of Classical Legal Thought 260
Tonal intuitions in "Tristan und Isolde" / by Brian Hyer 200
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 4333409
求助须知:如何正确求助?哪些是违规求助? 3845102
关于积分的说明 12010796
捐赠科研通 3485690
什么是DOI,文献DOI怎么找? 1913365
邀请新用户注册赠送积分活动 956534
科研通“疑难数据库(出版商)”最低求助积分说明 857259