使用电子健康记录嵌入式预测心力衰竭患者30天的入院预测:比较评估
Prabin Shakya1, Ayush Khaneja1, Kavishwar B Wagholikar1,2
1Laboratory of Computer Science, Massachusetts General Hospital, 399 Revolution Drive, 7th Floor, Boston, MA, 02145, United States, 1 8595360114.
JMIR medical informatics
|November 25, 2025
概括
预测心力衰竭的再入院是至关重要的. 在患者数据上训练的Word2vec嵌入式显著改善了预测模型,优于BERT和传统方法,用于更好的患者风险分层.
科学领域:
- 医疗信息学医学信息学
- 机器学习在医疗保健中的应用
- 临床预测建模临床预测建模
背景情况:
- 心力衰竭 (HF) 构成了重大的公共卫生挑战,其特点是高的无计划再入院率.
- 目前用于预测高频回收的模型的性能不足,需要改进功能工程.
- 电子健康记录 (EHR) 数据为开发预测模型提供了丰富的来源.
研究的目的:
- 评估和比较各种特征嵌入方法的有效性,以提高心力衰竭患者意外再入院的预测.
- 确定嵌入技术是否可以提高机器学习模型在识别高风险再入院患者的准确性.
主要方法:
- 对比了三个嵌入方法:word2vec在术语代码/概念唯一标识符 (CUI) 和BERT在概念描述上,与一热编码的基线相比.
- 使用后勤回归,极端梯度增强 (XGBoost) 和人工神经网络 (ANN) 模型.
- 通过使用MIMIC-IV数据集中的心力衰竭队列 (N=21,031) 的AUROC和F1得分来评估模型性能.
主要成果:
- 嵌入方法在所有测试的算法中显著改善了预测模型性能.
- 无论使用的嵌入方法,XGBoost都表现出卓越的性能.
- 在特定数据集上训练的Word2vec嵌入实现了更高的AUROC (0.65),相比于在概念描述上预先训练的BERT嵌入 (0.59).
结论:
- 在EHR数据上训练的Word2vec嵌入式有效地区分心力衰竭再入院病例,优于一次热编码和预训练的BERT嵌入式.
- 这些嵌入方法为预测再录取的自动特征选择提供了一种可行的方法.
- 观察到的AUROC改善支持增加风险分层和针对心力衰竭患者的有针对性的临床干预.
相关概念视频
01:30Methods of Documentation VII: EMR
1.4K
Electronic Medical Records (EMRs) primarily center around electronically documenting patients' health information within a single healthcare organization or practice. They contain essential clinical data related to a patient's medical history, diagnoses, medications, treatment plans, lab results, and other pertinent information relevant to the specific encounter or episode of care. EMRs are designed to streamline documentation and workflow processes within individual healthcare...
1.4K
01:23Holter Monitor: 24-Hour Monitoring
1.9K
Holter monitoring is a continuous electrocardiography (ECG) recording that tracks the heart's electrical activity over an extended period, generally 24 to 48 hours. This noninvasive diagnostic tool detects irregular heart rhythms that may not be captured during a standard ECG performed in a clinical setting.DeviceThe Holter monitor is a portable, small device connected to several electrodes on the patient's chest. These electrodes detect the heart's electrical signals and transmit them to the...
1.9K
01:30Heart Failure IV: Classification and Diagnostic Evaluation
300
Heart failure can be classified in various ways, with the most common classifications based on physical activity limitations, disease progression, severity, and treatment strategies.The Functional Classification of Heart Failure divides patients into four categories based on physical activity limitation due to symptom burden.Class I: Patients in this class have cardiac disease but no physical activity limitations. Ordinary activities like walking, climbing stairs, or routine tasks do not cause...
300
01:30Pulse rhythm
1.3K
Pulse rhythm refers to the pattern of pulsations within specific intervals, offering valuable insights into the regularity or irregularity of the heart's beats as observed through the pattern of pulsation within specific intervals. A regular pulse exhibits a consistent heart rate with uniform waveforms and pulsation force, variations of which can be classified as normal, weak, or bounding.
Conversely, an irregular pulse pattern is termed dysrhythmia, stemming from disruptions in cardiac...
Conversely, an irregular pulse pattern is termed dysrhythmia, stemming from disruptions in cardiac...
1.3K
01:19Purpose of Health Records II
1.4K
Health records serve various essential purposes in the healthcare system. Here are some key purposes:
1.4K
01:30Heart Failure V: Medical Management
204
Medical Management of Acute Decompensated Heart Failure (ADHF)The primary goals of therapy for patients hospitalized with acute decompensated heart failure (ADHF) include:Relieving symptomsOptimizing volume statusSupporting oxygenation and ventilationMaintaining cardiac output (CO) and end-organ perfusionIdentifying and addressing the cause of ADHFPreventing complicationsProviding patient education on factors precipitating HF exacerbationPlanning for dischargeOngoing monitoring and assessment...
204