Evaluation of approaches to estimate scapular kinematics during baseball pitching

Biomechanical analyses of pitching possess limitations in accurately measuring dynamic scapular orientation and are thus unable to distinguish between glenohumeral and scapulothoracic contributions to global shoulder motion. In lieu of direct measurement, several methods to estimate scapular kinematics have been developed. This study evaluated the ability of the linear model and the double calibration acromion marker cluster (D-AMC) approaches to estimate scapular kinematics throughout a full-speed pitching motion. Each approach's estimates were compared against scapulothoracic range of motion limits established in a non-pitching biplane fluoroscopy study involving various functional arm movements that approximate physiological limits of scapular motion. Fourteen healthy collegiate pitchers participated. Motion capture measured upper extremity joint kinematics during full-speed fastball pitches. Linear model and D-AMC approaches estimated scapulothoracic kinematics during each pitch. Linear model estimates of scapulothoracic kinematics were largely within established physiological limits on each scapular axis of motion while D-AMC estimates exceeded fluoroscopy-established bounds for more subjects and by larger, less physiologically plausible amounts. These findings demonstrate that the linear model outperforms the D-AMC and suggest that it is a viable approach to estimate scapular kinematics during pitching. Finally, these results offer additional evidence to support the accepted pattern of scapular kinematics during pitching.