Identifying Individuals Who Currently Report Feelings of Anxiety Using Walking Gait and Quiet Balance: An Exploratory Study Using Machine Learning

Sensors

Dylan Bradley¹  Ryan McCarthy²  Ali Boolani²  Emily Locke³  Lap-Fai Yu⁴  Haikun Huang⁴  Matthew Lee Smith⁵  Maggie Stark⁶  Chelsea Yager⁷  Mostafa Ali Elwan⁸  Ahmed Ali Torad⁸  Ahmed Mahmoud Kadry⁸  Rebecca Martin⁹ 

[1] Canino School of Engineering Technology, State University of New York, Canton, NY 13617, USA;Department of Biology, Clarkson University, Potsdam, NY 13699, USA;Department of Chemistry and Biomolecular Science, Clarkson University, Potsdam, NY 13699, USA;Department of Computer Science, George Mason University, Fairfax, VA 22030, USA;Department of Environmental and Occupational Health, School of Public Health, Texas A&M University, College Station, TX 77843, USA;Department of Medicine, Lake Erie Osteopathic College of Medicine, Elmira, NY 14901, USA;Department of Neurology, St. Joseph’s Hospital Health Center, Syracuse, NY 13203, USA;Department of Physical Therapy, Clarkson University, Potsdam, NY 13699, USA;Department of Physical Therapy, Hanover College, Hanover, IN 47243, USA;

关键词: anxiety;    gait;    mCTSIB;    balance;    sensors;    APDM monitors;   

DOI  :  10.3390/s22093163

来源: DOAJ

【 摘 要 】

Literature suggests that anxiety affects gait and balance among young adults. However, previous studies using machine learning (ML) have only used gait to identify individuals who report feeling anxious. Therefore, the purpose of this study was to identify individuals who report feeling anxious at that time using a combination of gait and quiet balance ML. Using a cross-sectional design, participants (n = 88) completed the Profile of Mood Survey-Short Form (POMS-SF) to measure current feelings of anxiety and were then asked to complete a modified Clinical Test for Sensory Interaction in Balance (mCTSIB) and a two-minute walk around a 6 m track while wearing nine APDM mobility sensors. Results from our study finds that Random Forest classifiers had the highest median accuracy rate (75%) and the five top features for identifying anxious individuals were all gait parameters (turn angles, variance in neck, lumbar rotation, lumbar movement in the sagittal plane, and arm movement). Post-hoc analyses suggest that individuals who reported feeling anxious also walked using gait patterns most similar to older individuals who are fearful of falling. Additionally, we find that individuals who are anxious also had less postural stability when they had visual input; however, these individuals had less movement during postural sway when visual input was removed.

【 授权许可】

Unknown