Design and Development of the Wearable-based Fall Characteristics Monitoring System

碩士 === 國立陽明大學 === 生物醫學工程學系 === 106 === Fall and consequent injury are public health issues and threats for elderly. To improve the quality of life and ability to live independently, fall detection and prevention mechanisms are proposed to cope with occurrence of fall. However, fall is unexpected and...

Full description

Bibliographic Details
Main Authors: Wan-Ting Shi, 施婉婷
Other Authors: Chia-Tai Chan
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/64s3c5
Description
Summary:碩士 === 國立陽明大學 === 生物醫學工程學系 === 106 === Fall and consequent injury are public health issues and threats for elderly. To improve the quality of life and ability to live independently, fall detection and prevention mechanisms are proposed to cope with occurrence of fall. However, fall is unexpected and inevitable event, so clinical professionals have to identify fall characteristics and provide fall prevention strategies. The fall characteristics include fall events, types of fall and fall directions. The occurrence and related circumstance of fall event are mainly recorded through subjective questioning by clinical professionals and self-reported by elderly. However, these ways are not convincible and incomplete, because a part of elderly may misremember or forget cause of fall event. Therefore, the fall characteristics should be monitored and recorded to analyze relationship between falls and particular injures. The fall characteristics monitoring system consists of high accuracy fall event detection algorithm and fall direction identification. The hierarchical fall event detection algorithm and machine learning-based fall direction identification algorithm are proposed. The performance of the sensitivity, specificity, and accuracy using the hierarchical fall event detection algorithm are 99.83%, 98.44%, 99.19 %, respectively. The performance of the sensitivity, and accuracy using the fall direction identification algorithm are 98.52% and 97.34 %, respectively. The results show that the proposed characteristics monitoring system can accurately detect fall event and provide direction of fall for the strategic plan of falls prevention.