Search research articles

关于 JoVE

概览领导团队博客 JoVE 帮助中心

作者

出版流程编辑委员会范围与政策同行评审常见问题投稿

图书馆员

用户评价订阅访问资源图书馆顾问委员会常见问题

研究

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments 存档

教育

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual 教师资源中心教师网站

使用条款与条件

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序

Same author

Early Nephrology Consultation and Acute Kidney Injury in Hospitalized Patients: A Randomized Clinical Trial.

JAMA network open·2026

Same author

Fast Molecular Crystal Structure Prediction Using Sampling by Analogy to Previously Predicted Landscapes.

Journal of chemical theory and computation·2026

Same author

Predicting Intensive Care Readmission Among Hospitalized Children.

medRxiv : the preprint server for health sciences·2026

Same author

New proteomic biomarkers identified in plasma extracellular vesicles in sarcoidosis: a case-control matched study.

Frontiers in immunology·2026

Same author

Balancing Model Performance With Operational Realities in Early Warning Systems-Complexity Where It Matters.

JAMA network open·2026

Same author

<i>Letter:</i> Provider Intuition Predicts Future Retention in Care Among People with HIV.

AIDS patient care and STDs·2026

Search research articles

相关实验视频

Updated: Jun 27, 2025

An R-Based Landscape Validation of a Competing Risk Model

An R-Based Landscape Validation of a Competing Risk Model

Published on: September 16, 2022

使用可变长度时间序列数据的深度学习临床预测模型的开发和外部验证.

Fereshteh S Bashiri¹, Kyle A Carey², Jennie Martin¹

¹Department of Medicine, University of Wisconsin-Madison, Madison, WI 53792, United States.

Journal of the American Medical Informatics Association : JAMIA

|April 29, 2024

概括

长期短期内存/关闭循环单元 (LSTM/GRU) 模型架构与决策树 (PLE-DT) 数据转换的碎片式线性编码相结合,实现了预测临床结果的最高性能. 这种方法在多个临床任务的外部验证中显示出卓越的结果.

关键词:

人工智能在医学中的应用深度学习是一种深度学习.可变长度的时间序列.

更多相关视频

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack

Published on: May 15, 2020

相关实验视频

Last Updated: Jun 27, 2025

An R-Based Landscape Validation of a Competing Risk Model

An R-Based Landscape Validation of a Competing Risk Model

Published on: September 16, 2022

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack

Published on: May 15, 2020

科学领域:

临床信息学是一种临床信息学.
机器学习在医疗保健中的应用
时间序列分析时间序列分析.

背景情况:

从电子健康记录中准确预测临床结果对于患者护理至关重要.
可变长度时间序列数据为预测建模提出了独特的挑战.
深度学习架构为分析复杂的临床数据提供了潜力.

研究的目的:

为了对变长时间序列数据进行深度学习模型的比较和外部验证.
评估与深度学习架构结合的不同数据转换方法.
评估模型在三个不同的临床任务中的性能:临床恶化,严重急性损伤 (AKI) 和疑似感染.

主要方法:

一个多中心的回顾性研究,使用两个医疗中心的2007-2022年数据.
为每项临床任务创建不同的数据集,其中一个站点用于培训,另一个站点用于测试.
三种特征工程方法 (规范化,标准化,PLE-DT) 和三种架构 (LSTM/GRU,TCN,TDW-CNN) 的比较.
使用精度回忆曲线下的面积 (AUPRC) 和接收器操作特征曲线下的面积 (AUROC) 评估模型歧视.

主要成果:

该研究包括超过373,000名培训录取和256,000名测试录取.
长期短期记忆/门循环单元 (LSTM/GRU) 和时间分布式包装与卷积神经网络 (TDW-CNN) 模型在两个任务中实现了最高的AUPRC平均值.
在所有任务中,LSTM/GRU模型显示了最高的平均AUROC (恶化:0.81,AKI:0.92,感染:0.87).
结合PLE-DT与LSTM/GRU,在所有临床任务中产生了最高的AUPRC.

结论:

使用PLE-DT数据转换的LSTM/GRU架构显示出优异的外部验证性能,用于预测临床结果.
虽然使用AUROC的多个模型具有类似的性能,但LSTM/GRU-PLE-DT方法在AUPRC中表现出色.
LSTM架构仍然与较新的模型竞争,PLE-DT可以增强AUPRC用于可变长度时间序列的临床预测.