Development of machine learning models to support medical treatment in the field of resuscitation and emergency medicine

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K09302
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 55060:Emergency medicine-related
• Research Institution: The University of Tokyo
• Principal Investigator: Seki Tomohisa 東京大学, 医学部附属病院, 助教 (30528873)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 救急医療 / 蘇生医療 / 機械学習 / 予測モデル

Outline of Final Research Achievements

The purpose of this study is to develop a machine learning model that supports clinical judgment and management in the flow of medical treatment in the field of emergency resuscitation, improves clinical outcomes, and reduces the burden on clinicians. We developed a machine learning model to predict cases of presumed cardiogenic cardiopulmonary arrest based on information at the time of arrival at the hospital, and a machine learning model to predict the risk of in-hospital mortality based on blood sampling data and patient background information at the time of admission. This study demonstrated the applicability of machine learning technology to the field of emergency resuscitation, and the risk stratification by machine learning is expected to contribute to the improvement of the quality of medical care.

Free Research Field

救急医療

Academic Significance and Societal Importance of the Research Achievements

本研究において、本領域に対する機械学習の適用性を検討した研究は限られていたが、現存するデータベースを用いた予測モデルの開発がより正確なリスク層別化に資する可能性があることを示し、医療の質の向上に資する可能性が期待できると考えられた。診療に伴って蓄積されつつも、人間が扱いやすい粒度まで単純化されたスコアなどで利用しきれていない患者データの特徴を利用することが可能になると考えられ、現状でデータから定量化できていない情報を計算機上で扱い、予測を出力するモデルの重要性が示された。