Classifying Visually Defined Varus Thrust in Knee Osteoarthritis Using Wearable Inertial Sensors and Markerless Motion Analysis

Abstract

Varus thrust (VT) is a gait phenomenon seen in people with knee osteoarthritis (OA) that involves a sudden lateral movement of the knee joint, occurring within the first portion of the stance phase of the gait cycle. It is associated with improper joint loading and disease progression. Currently, visual assessment is the standard method for identifying VT but is subjective and prone to variability. The aim of this study was to explore and evaluate technological methods for assessing VT presence, using objective measurement tools. Visual VT assessment served as the reference standard, while markerless optical motion capture and wearable inertial sensor data were collected concurrently. Visual VT presence was initially assessed using a discrete scale that was based on the number of times it was observed, across multiple walking passes made by each participant. Motion capture data collected from 10 synchronized cameras were used to calculate frontal plane joint excursion (degrees), which was the variable of interest from the optical motion capture system. Participants also wore an inertial sensor on their upper tibia during their walking trials, and from these devices, information on their lateral tibial acceleration (m/s^2) and their peak frontal plane tibial angular velocity (degrees/s) were obtained. The results showed that peak lateral acceleration, measured by wearable sensors, had good discriminatory power in identifying visual VT presence, particularly in more visually apparent cases. These findings represent an important first step toward establishing objective, sensor-based methods for VT detection in clinical and research settings. Further research is needed to validate the outcomes, improve measurement accuracy in moderate presentations, and assess reliability across diverse clinical populations.