Making the Improbable Possible: Generalizing Models Designed for a Syndrome-Based, Heterogeneous Patient Landscape

IF 3 3区医学 Q2 CRITICAL CARE MEDICINE

Critical Care Clinics Pub Date : 2023-10-01 DOI:10.1016/j.ccc.2023.02.003

Joshua Pei Le BS , Supreeth Prajwal Shashikumar PhD , Atul Malhotra MD , Shamim Nemati PhD , Gabriel Wardi MD, MPH

引用次数: 0

查看原文本刊更多论文

使不可能成为可能：为基于综合征的异质患者环境设计的通用模型。

综合征，如败血症，通常在重症监护室遇到。尽管临床医生很容易掌握这些条件，但这些条件可能会限制机器学习算法的性能。个别医院的实践模式可能会限制外部的可推广性。数据丢失是优化算法性能的另一个障碍，并且存在各种策略来缓解这种情况。数据科学的最新进展，如迁移学习、保形预测和持续学习，可能会提高机器学习算法在危重患者中的可推广性。这些方法的随机试验表明，在这一点上，以患者为中心的结果有所改善。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Critical Care Clinics 医学-危重病医学

CiteScore

8.00

自引率

0.00%

发文量

审稿时长

>12 weeks

期刊介绍： Critical Care Clinics updates you on the latest trends in patient management, keeps you up to date on the newest advances, and provides a sound basis for choosing treatment options. Published four times a year—in January, April, July, and October—each issue focuses on a single topic in critical care, including cardiac emergencies, sepsis, infectious diseases, shock and trauma, neurologic problems, toxicology, pulmonary disorders, intensive care procedures, and patient observation. Critical Care Clinics is also an ideal resource for those in the fields of anesthesia, emergency medicine, and respiratory medicine.