Artificial Intelligence and Wearable Sensor-Based Gait Analysis in Chronic Ankle Instability: A Scoping Review

Abstract

Background: Chronic ankle instability (CAI) frequently arises from recurrent ankle sprains, resulting in persistent pain, weakness, and altered gait patterns that impair mobility and quality of life. Wearable sensors such as inertial measurement units (IMUs, accelerometers, gyroscopes, and smart insoles enable real-time gait monitoring outside clinical settings. Artificial intelligence (AI) techniques enhance the detection of these abnormalities with high accuracy and enable objective, continuous, and real-world gait analysis, offering new opportunities for early detection and rehabilitation monitoring.

Objective: This scoping review maps the extent, nature, and evidence on Al-driven analysis of gait data from wearable sensors in CAl assessment, emphasising clinical and physiotherapy applications, key trends, gaps, and future directions Io map and summarise current evidence on the use of Al methods combined with wearable sensor- based gait analysis in individuals with chronic ankle instability.

Results: Emerging studies demonstrate Al models achieving over 90% accuracy in classifying CAl gait deviations using IMU and shoe-integrated sensors, capturing spatiotemporal asymmetries and joint kinematics. ML excels in joint impairment differentiation, while deep learning aids rehabilitation prediction. Physiotherapy relevance includes objective progress tracking; however, studies are limited to small cohorts, with gaps in longitudinal validation, diverse populations, and unsupervised learning.

Conclusion: Wearable integrations offer promising, portable tools for CAL gait assessment and physiotherapy, but larger trials and standardised protocols are needed to bridge gaps and guide clinical adoption.