Predicting dementia with routine care EMR data

作者：

Highlights：

• Dementia can be predicted from EMR routine care data with a level of accuracy suitable for automated pre-screening.

• Unstructured medical notes improve the accuracy of machine learning models for dementia prediction.

• Patients’ health institution affiliations need to be considered in a multi-institution model for dementia.

摘要

•Dementia can be predicted from EMR routine care data with a level of accuracy suitable for automated pre-screening.•Unstructured medical notes improve the accuracy of machine learning models for dementia prediction.•Patients’ health institution affiliations need to be considered in a multi-institution model for dementia.

论文关键词：Dementia,Prediction,Random forest,EMR,Machine learning

论文评审过程：Received 26 October 2018, Revised 24 November 2019, Accepted 25 November 2019, Available online 5 December 2019, Version of Record 12 December 2019.

论文官网地址：https://doi.org/10.1016/j.artmed.2019.101771