利用预先训练的语言模型从评估报告中提取发作频率
Rashmie Abeysinghe1,2, Shiqiang Tao1,2, Samden D Lhatoo1,2
1Department of Neurology, McGovern Medical School, The University of Texas Health Science Center at Houston, Houston, TX, USA.
NPJ digital medicine
|April 14, 2025
概括
从临床笔记中提取发作频率对于治疗至关重要. 生成型大语言模型,特别是GPT-4,在结构化这些重要患者数据方面表现出强的表现.
科学领域:
- 医疗信息学 医疗信息学
- 自然语言处理自然语言处理.
- 的研究研究.
背景情况:
- 发作频率对于评估治疗疗效和患者安全至关重要.
- 临床叙述通常包含非结构化的发作频率细节,这对数据提取构成了挑战.
研究的目的:
- 开发和评估一种方法,从非结构化的临床文本中提取结构化的发作频率信息.
- 在这个提取任务中比较基于BERT和生成的大型语言模型的性能.
主要方法:
- 基于BERT的微调模型 (bert-large-cased,biobert-large-cased,Bio_ClinicalBERT) 和生成模型 (GPT-4,GPT-3.5 Turbo,Llama-2-70b-hf) 进行了微调.
- 研究了两个任务:提取发作频率短语和提取发作频率属性.
- 整合两个任务的结果,以创建最终的结构化输出.
主要成果:
- 在所有评估任务中,GPT-4表现出卓越的性能.
- GPT-4在频率短语提取方面获得了85.79%的F1分数,在属性提取方面获得了91.84%.
- 使用GPT-4的最终集成结构化输出获得了F1得分85.82%.
结论:
- 精心调整的生成大型语言模型显示出从有限的临床文本中提取结构化数据的巨大潜力.
- 这种方法可以改善临床叙述的利用,用于患者的治疗和研究.
相关概念视频
01:24Arteries of the Lower Limbs
167
Epilepsy is a chronic neurological disease marked by recurrent, unpredictable seizures. These seizures are caused by abnormal electrical discharges in the brain, leading to behavior, sensation, or consciousness alterations. They can also cause transient impairment of awareness, interfering with daily activities.
Various factors can trigger epilepsy, including genetic factors, brain damage, metabolic causes, and unknown etiology. Diagnosis of epilepsy involves electroencephalography (EEG), which...
Various factors can trigger epilepsy, including genetic factors, brain damage, metabolic causes, and unknown etiology. Diagnosis of epilepsy involves electroencephalography (EEG), which...
167
01:13Seizures: Classification
285
Epilepsy is primarily characterized by unpredictable seizures, either provoked by an identifiable factor, such as injury or illness, or unprovoked, occurring spontaneously without apparent cause.
Seizures are typically classified into two main categories: focal and generalized seizures.
Focal Seizures
Focal seizures originate from specific regions of the brain. These seizures are further sub-classified into two types:
Seizures are typically classified into two main categories: focal and generalized seizures.
Focal Seizures
Focal seizures originate from specific regions of the brain. These seizures are further sub-classified into two types:
285