Affiliation:
1. Department of Computer Graphics, Vision and Digital Systems, Faculty of Automatic Control, Electronics and Computer Science Silesian University of Technology Gliwice Poland
2. Institute for Artificial Intelligence and Cybersecurity University of Klagenfurt Klagenfurt Carinthia Austria
3. Department of Physical Education and Sport, Faculty of Physical Education and Physiotherapy Opole University of Technology Opole Poland
Abstract
AbstractFrailty is a significant issue in geriatric health, may cause adverse effects such as falls, delirium, weight loss, or physical decline. Over time, various methods have been developed for measuring frailty, including clinical judgment, the frailty index, the clinical frailty scale, and the comprehensive geriatric assessment. These traditional frailty assessment approaches rely on healthcare professionals, which can lead to inaccuracy and require frequent clinic visits, making it burdensome for elderly patients. This review paper explores the latest trends in frailty assessment by measuring gait parameters using wearable sensors, specifically the inertial measurement unit (IMU). The aim of this study is to provide a comprehensive overview of objective methods for evaluating and quantifying frailty. We focus on the application of machine learning (ML) and deep learning (DL) techniques to IMU gait data, highlighting key aspects of recent publications such as algorithms, sensor types, sample sizes, and performance evaluations. By examining the strengths and challenges of each technique, this review aims to guide future studies on utilizing cost‐effective and portable devices integrated with clinical data. This integration can help to propose optimized IMU gait parameters or ML models to detect early‐stage frailty. This advances the emerging trend of intelligent, individualized, and efficient healthcare systems for older adults.This article is categorized under:
Application Areas > Health Care
Technologies > Machine Learning
Technologies > Artificial Intelligence