REAL-TIME HIP RANGE OF MOVEMENT MEASUREMENT IN TELEREHABILITATION; COMPARISON OF HUMAN POSE ESTIMATION TOOL, PHYSIOROM, VS 3D MOTION CAPTURE SYSTEM

Abstract

INTRODUCTION AND AIMS Telerehabilitation is widely acknowledged, but its adoption in musculoskeletal cohorts has been restricted due to challenges in accurately measuring objective parameters such as range of movement (ROM). This research assesses the accuracy of a human pose estimation tool utilising two-dimensional (2D) video (PhysioROM, Coviu) for evaluating active hip flexion and abduction ROM during common rehabilitation exercises, compared to three-dimensional (3D) motion capture. METHODS Fifty participants (18 to 80 years) were recruited, each performing three common hip movements; standing hip flexion, standing hip abduction and a sit-to-stand. The motions were captured using the 12-camera, 3D Vicon motion capture system in conjunction with simultaneous 2D video, simulating a home telehealth consultation scenario. Joint kinematics were computed using either the Vicon Plug-in Gait model (3D Vicon) or the human pose estimation tool integrated into the PhysioROM algorithm (2D PhysioROM). RESULTS No significant differences were observed in peak hip ROM measurements during the tasks undertaken. Mean peak hip flexion was 2.4° (p=1.0) less for 3D Vicon versus 2D PhysioROM in standing hip flexion, though 2.1° (p=0.46) greater during sit-to-stand movement. Hip abduction was 5.3° (p< 0.001) less for 3D Vicon versus 2D PhysioROM in standing hip abduction. Statistical parametric mapping revealed significant differences (p<0.001) in hip flexion between Vicon 3D and 2D PhysioROM for 32% of timepoints for both standing hip flexion and sit-to-stand movements, clustered around full hip extension. During standing hip abduction, significant differences (p<0.001) were found at 58% of timepoints clustered around neutral postures. Overall mean absolute deviation was 10.1°, 9.7° and 9.0° for hip flexion/extension, sit-to-stand, and hip abduction respectively. CONCLUSION Validity of PhysioROM to accurately measure peak hip flexion and peak hip abduction was high. PhysioROM may provide an effective method of accurately and objectively quantifying hip movement during a telerehabilitation session using 2D video.