机器学习早期预警模型在医疗和外科住院患者中的可靠性
Pedro J Caraballo1,2, Anne M Meehan1, Karen M Fischer2
1Department of Medicine, Mayo Clinic, Rochester, MN 55905, United States.
JAMIA open
|January 7, 2025
概括
机器学习预警系统 (EWS) 在一般医院病房中表现不同. 该模型对医疗患者的表现明显好于手术患者,突出显示了亚种群差异.
科学领域:
- 临床信息学 临床信息学
- 医疗保健中的人工智能
- 患者监测 患者监测
背景情况:
- 基于机器学习 (ML) 的早期预警系统 (EWS) 在一般医院病房中用于识别有恶化风险的患者.
- 这些系统旨在促进危险患者及时的救援干预.
研究的目的:
- 评估基于ML的EWS在成人医疗和手术患者的亚群表现,这些患者被录入普通医院病房.
- 评估患者异质性对基于ML的EWS的有效性和临床鉴别的影响.
主要方法:
- 集成到电子健康记录中的EWS每15分钟计算得分,以预测复合不良事件 (AE).
- 副作用包括全因死亡率,转移到重症监护室,心脏骤停或快速反应小组评估.
- 基于上一次EWS分数,使用接收器操作特征曲线 (ROC-AUC) 和精度回忆曲线 (PRC-AUC) 下的面积来评估性能.
主要成果:
- 总共有35,937例医疗住院病例 (6.05%的AE) 和25,214例外科住院病例 (19.77%的AE) 被分析.
- 在医疗和手术患者之间的得分分布中观察到显著的差异 (P < .001).
- 基于ML的EWS在医疗患者 (ROC-AUC 0.869,PRC-AUC 0.988) 与手术患者 (ROC-AUC 0.677,PRC-AUC 0.878) 相比显示出更高的性能.
结论:
- 医疗和外科患者群体之间的异质性显著影响了基于ML的EWS性能.
- 模型有效性和临床鉴别受到这些亚种群差异的影响.
- 描述目标患者亚群对于为普通医院病房开发有效的基于ML的EWS至关重要.
更多相关视频
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
8.2K
07:31Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
7.0K
相关概念视频
01:25Errors occurring during blood pressure monitoring
591
Blood pressure monitoring is a crucial clinical procedure in diagnosing and managing various cardiovascular conditions. Despite its significance, the accuracy of blood pressure measurements can be compromised by multiple factors, potentially leading to either falsely high or low readings. These inaccuracies are critical as they can significantly impact patient care. So, it is vital to understand these challenges deeply and adopt strategic approaches to minimize errors.
Several factors...
Several factors...
591
01:15Receiver Operating Characteristic Plot
74
A ROC (Receiver Operating Characteristic) plot is a graphical tool used to assess the performance of a binary classification model by illustrating the trade-off between sensitivity (true positive rate) and specificity (false positive rate). By plotting sensitivity against 1 - specificity across various threshold settings, the ROC curve shows how well the model distinguishes between classes, with a curve closer to the top-left corner indicating a more accurate model. The area under the ROC curve...
74