earticle

영어

Purpose: We aimed to propose and validate a novel method for measuring and analyzing cervical spine range of motion (ROM) using built-in smartphone cameras and artificial intelligence technology. Methods: The participants were 15 healthy men (average age: 22.40±1.30 y, height: 175.60±4.17 cm, weight: 74.07±8.05 kg). Cervical spine ROM during extension, flexion, lateral bending, and rotation was collected using a three-dimensional (3D) motion analysis system. An application with the Pose Detection feature of the Google ML Kit was used to capture and record body segment positions. ROM was calculated using Visual 3D software, and multivariate linear regression analysis was performed on the collected data. Results: Lateral flexion (left) had the highest determination coefficient (Adjusted R2=0.851), showing the most consistent biomechanical pattern, while extension had a lower determination coefficient (Adjusted R2=0.747). ROM estimates that regression models had a root mean square error between 4.43° and 5.85° and a normalized root mean square error between 9.90% and 18.03%. Conclusion: This study presents a data-driven approach to understanding cervical motions using smartphones, providing foundational data for clinical evaluation.