雑誌文献を検索します。書籍を検索する際には「書籍検索」を選択してください。

検索

書誌情報 詳細検索 by 医中誌

Japanese

Clinical research using machine learning in the ICU Tadahiro GOTO 1,2 1TXP Medical Co., Ltd. 2Department of Clinical Epidemiology & Health Economics School of Public Health University of Tokyo pp.798-804
Published Date 2020/10/1
DOI https://doi.org/10.11477/mf.3102200817
  • Abstract
  • Look Inside
  • Reference

Machine learning can accurately perform complex calculations, classification, and predictions compared to conventional biostatistics, with techniques such as non-penalized logistic regression models. Machine learning is classified into unsupervised learning (classification), supervised learning (prediction), and reinforcement learning (decision-support). Unsupervised learning has been used to identify latent subgroups in those with a specific disease (e.g., sepsis), which has traditionally thought to be a homogeneous condition. Clinical research using unsupervised learning is expected to increase in the field of intensive care medicine. By contrast, while supervised learning is widely used to predict patient prognosis, implementation of developed algorithms becomes an important bottleneck in real clinical situations. Machine learning will be one of the most important techniques in clinical research. However, it is more important for clinicians to know the strengths and weaknesses of machine learning and to think about how machine learning can change the future of intensive care and how it can benefit patients, rather than learning specific techniques for machine learning.


Copyright © 2020, MEDICAL SCIENCES INTERNATIONAL, LTD. All rights reserved.

基本情報

電子版ISSN 2186-7852 印刷版ISSN 1883-4833 メディカル・サイエンス・インターナショナル

関連文献

もっと見る

文献を共有