MhNet: Multi-scale spatio-temporal hierarchical network for real-time wearable fall risk assessment of the elderly

• Published in: Computers in biology and medicine Vol. 144; p. 105355
• Main Authors: Wu, Shibin, Ou, Jianlin, Shu, Lin, Hu, Guohua, Song, Zhen, Xu, Xiangmin, Chen, Zhuoming
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.05.2022; Elsevier Limited
• Subjects: Accuracy; Aged; Cross-subject; Data collection; Datasets; Deep Learning; Falls; Feature extraction; Gait; Health risks; Hierarchical network; Humans; Internal Medicine; Older people; Other; Plantar pressure; Pressure; Pressure sensors; Quality of Life; Real time; Real-time fall risk assessment; Rehabilitation; Risk Assessment; Sensors; Smartphones; Temporal variations; Wearable; Wearable computers; Wearable Electronic Devices; Wearable technology; Real-time fall risk assessment; Plantar pressure; Hierarchical network; Wearable; Cross-subject
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2022.105355

Continuous fall risk assessment and real-time high falling risk warning are extremely necessary for the elderly, to protect their lives and ensure their quality of life. Wearable in-shoe pressure sensors have the potential to achieve these targets, due to their adequate wearing comfort. However, it is a great challenge to remove the individual differences of foot pressure data and identify the accurate fall risk from fewer gait cycles to realize real-time warning. We explored a hierarchical deep learning network named MhNet for real-time fall risk assessment, which utilized the advantages of two-layer network, to reach hierarchical tasks to reduce probability of misidentification of high fall risk subjects, by establishing a borderline category using the rehabilitation labels, and extracting multi-scale spatio-temporal features. It was trained by using a wearable plantar pressure dataset collected from 48 elderly subjects. This method could achieve a real time fall risk identification accuracy of 73.27% by using only 9 gaits, which was superior to traditional methods. Moreover, the sensitivity reached 76.72%, proving its strength in identifying high risk samples. MhNet might be a promising way in real-time fall risk assessment for the elderly in their daily activities. •Continuous plantar pressure-fall risk dataset was bulit by smart shoes integrated with flexible sensors.•Multi-scale convolution was proposed to capture spatio and temporal features which represented physiological significance.•A 2-layer multi-scale spatio-temporal hierarchical network was constructed for real-time fall risk assessment.•Domain adaptation was improved to minimize the individual variation and achieve better generilization ability.