Prediction of weaning from mechanical ventilation using Convolutional Neural Networks

作者：

Highlights：

• Present a decision support system (DSS) that uses available patient information to predict time to weaning readiness.

• The DSS predict the most appropriate treatment action in the next hour for a given patient state.

• Use of two complementary forms of explanation: feature importance and counterfactual explanations.

• Step towards making ML-based DSS both personalised and actionable.

摘要

•Present a decision support system (DSS) that uses available patient information to predict time to weaning readiness.•The DSS predict the most appropriate treatment action in the next hour for a given patient state.•Use of two complementary forms of explanation: feature importance and counterfactual explanations.•Step towards making ML-based DSS both personalised and actionable.

论文关键词：Ventilator weaning,Mechanical ventilation,Deep learning,Feature importance

论文评审过程：Received 18 December 2020, Revised 3 May 2021, Accepted 3 May 2021, Available online 5 May 2021, Version of Record 13 May 2021.

论文官网地址：https://doi.org/10.1016/j.artmed.2021.102087