Development, validation, and feature extraction of a deep learning model predicting in-hospital mortality using Japan's largest national ICU database: a validation framework for transparent clinical Artificial Intelligence (cAI) development.

Journal: Anaesthesia, critical care & pain medicine
Published Date: Oct 24, 2022

Abstract

OBJECTIVE: While clinical Artificial Intelligence (cAI) mortality prediction models and relevant studies have increased, limitations including the lack of external validation studies and inadequate model calibration leading to decreased overall accuracy have been observed. To combat this, we developed and evaluated a novel deep neural network (DNN) and a validation framework to promote transparent cAI development.

Authors

  • Euma Ishii
    Department of Global Health Promotion, Tokyo Medical and Dental University, Tokyo, Japan.
  • Nobutoshi Nawa
    Department of Medical Education Research and Development, Tokyo Medical and Dental University, Tokyo, Japan.
  • Satoru Hashimoto
    Division of Applied Chemistry, Faculty of Science and Technology, Oita University Graduate School of Engineering, 700, Dan-noharu, Oita 870-1124, Japan. tinoue@oita-u.ac.jp.
  • Hidenobu Shigemitsu
    Institute of Global Affairs, Tokyo Medical and Dental University, Tokyo, Japan.
  • Takeo Fujiwara
    Department of Global Health Promotion, Tokyo Medical and Dental University, Tokyo, Japan. Electronic address: fujiwara.hlth@tmd.ac.jp.

Keywords

External Resources

Popular Topics
Recent Journals