Posterior marker set for shoulder kinematic analysis using optoelectronic systems

Comprehensive evaluation of the shoulder function relies on accurate tracking of humerothoracic and scapulothoracic motion. This study presents a posterior marker configuration comprising three custom 3D-printed clusters (trunk, scapula, upper arm) for use with infrared optoelectronic motion capture systems. The protocol includes a static, unilateral estimation of the glenohumeral joint center (GH). Eight healthy adults performed cycles of abduction/adduction, flexion/extension, horizontal abduction, internal/external rotation, and scapular retraction/protraction. Kinematic trajectories obtained with the proposed marker set were compared to those from ISB-recommended frames using RMSE, linear regression (R², slope, intercept), and one-way ANOVA on the range of motion (ROM) of the principal angle across subjects. Humerothoracic kinematics showed mean RMSE values of 3.47° (abduction), 12.53° (flexion), 5.39° (horizontal abduction), and 10.61° (internal/external rotation), with R² > 0.996 and slopes between 0.93 and 1.05. Scapulothoracic retraction/protraction resulted in an RMSE of 4.64°, R² = 0.95, and slope of 0.82. Scapular medial/lateral rotation during flexion yielded an RMSE of 6.64°, R² = 0.97, and slope of 1.21. No significant ROM differences were observed between protocols (p > 0.05). To evaluate suitability for constrained environments, cluster visibility was analyzed using ArUco markers. Under the assumption of comparable visibility between monocular video and low-baseline infrared setups, the mean ArUco fill rate was 99.10 %. The mean Euclidean distance between the new static and functional GH center estimates was 14.33 mm. The spinal configuration demonstrated accuracy comparable to sternal tracking while reducing setup complexity, supporting its use in semi-structured environments such as clinical or field-based assessments.