Supervised machine learning scheme for electromyography-based pre-fall detection system

作者：

Highlights：

• A surface electromyography based system for the fall risk assessment is proposed.

• A machine learning scheme was adopted to enhance the degree of generalization.

• The measured sensitivity and specificity values are about 90%.

• The fall is recognized with a lead-time before the impact of about 770 ms.

摘要

•A surface electromyography based system for the fall risk assessment is proposed.•A machine learning scheme was adopted to enhance the degree of generalization.•The measured sensitivity and specificity values are about 90%.•The fall is recognized with a lead-time before the impact of about 770 ms.

论文关键词：Surface Electromyography sensors,Machine Learning,Linear Discriminant Analysis,Fall risk assessment,Features extraction,Wearable devices

论文评审过程：Received 6 June 2017, Revised 1 December 2017, Accepted 27 January 2018, Available online 31 January 2018, Version of Record 8 February 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.01.047