一种自然语言处理算法用于对阿尔茨海默病和相关痴呆症患者的自杀行为进行分类:使用电子健康记录数据的开发和验证
medRxiv : the preprint server for health sciences
|August 7, 2023
概括
一个新的自然语言处理 (NLP) 算法准确地识别了阿尔茨海默病和相关痴呆症 (ADRD) 患者的自杀念头. 这种机器学习模型实现了高精度和回忆,有助于临床文档分析.
科学领域:
- 计算语言学计算语言学
- 临床信息学是一种临床信息学.
- 神经科学是一个神经科学.
背景情况:
- 阿尔茨海默病和相关痴呆症 (ADRD) 与自杀风险增加有关.
- 在临床文档中准确识别自杀念头对于患者护理至关重要.
- 在电子健康记录中识别自杀行为现有的方法有限.
研究的目的:
- 开发和验证一种自然语言处理 (NLP) 算法,用于识别和分类ADRD患者的自杀行为.
- 利用机器学习 (ML) 和深度学习 (DL) 技术来增强临床概念提取.
- 提高在大型医疗保健数据集中检测自杀念头的准确性和效率.
主要方法:
- 利用MIMIC-III和MIMIC-IV数据集,通过ICD代码识别患有ADRD和自杀念头的患者.
- 使用共弦相似度与Scan数据集计算临床笔记的语义相似度得分.
- 应用常规ML和DL模型,根据手册注释进行验证,用于分类.
- 将文档分类为八个自杀行为类别.
主要成果:
- 该NLP算法实现了高性能,准确度和回忆率高达98%用于识别ADRD患者的自杀念头.
- 该模型展示了在MIMIC数据集中有效地复制人类关于自杀想法的注释的能力.
- 分类器在识别和分类自杀行为时的表现与人类注释器相当.
结论:
- 开发的NLP算法为在ADRD人群中识别和分类自杀念头文档提供了坚实的基础.
- 这一进步增强了NLP在医疗保健中的应用,用于临床概念提取,特别是自杀想法.
- 这项研究强调了人工智能驱动工具的潜力,以支持脆弱患者群体的心理健康监测和临床决策.
相关概念视频
Alzheimer's Disease: Overview
521
Alzheimer's Disease (AD) is a continually advancing neurodegenerative disorder, distinguished by escalating memory loss, cognitive dysfunction, and dementia. The disease unfolds in three stages: preclinical, mild cognitive impairment (MCI), and dementia. Its onset is insidious, and the progression gradual, with the cause not well explained by other disorders.
The clinical diagnosis of AD hinges on the presence of memory and other cognitive impairments. Biomarkers, such as changes in Aβ...
The clinical diagnosis of AD hinges on the presence of memory and other cognitive impairments. Biomarkers, such as changes in Aβ...
521
Alzheimer's Disease: Treatment
215
Alzheimer's Disease (AD), a neurodegenerative disorder, is pathologically identified by amyloid plaques and neurofibrillary tangles composed of tau protein. AD pharmacotherapy aims to manage cognitive symptoms, delay disease progression, and treat behavioral symptoms. The treatment is primarily symptomatic and palliative, with no definitive disease-modifying therapy available. Cholinesterase inhibitors, including donepezil (Aricept), rivastigmine (Exelon), and galantamine (Razadyne), are...
215


