Functional Identification of Shoulder Joint Centers
University of Delaware
Analysis of upper extremity kinematics is difficult due to the lack of bony landmarks on the upper arm and shoulder. The medial and lateral epicondyles provide two potential bony landmarks for the upper arm, and the shoulder joint center provides a potential third landmark. Two methods commonly used to estimate the location of the shoulder joint center are the constant offset method (Rab, Petuskey, and Bagley, 2002) and the spherical fit method (Hicks and Richards, 2005). The constant offset method is susceptible to error due to variability of human geometry. The spherical fit method has been shown to be more accurate than constant offset methods for the hip; however, the methods have not been compared for the shoulder joint center. The objective was to compare the accuracy of the shoulder joint centers found using the functional spherical fit method as well as the constant offset method proposed by Rab, Petuskey, and Bagley (2002) relative to a physical measure of the shoulder joint center found using ultrasound. The physical measure was compared to the centers found with the constant offset method and functional identification for four positions: adduction-ER, 90º abduction-ER, 180º abduction-ER, and 90º flexion, IR. The centers were found for each method using the elevation of the humerus as a constraint. Specifically, motion trial data was analyzed at maximum elevation limits ranging from -30º below horizontal to 80º above horizontal at 10º intervals. Results indicated that the angle of elevation when using the spherical fit should be limited to zero degrees relative to the horizontal in order to minimize error. Using this minimized difference, the functional identification is most accurate in the adduction-ER position. Since both estimation methods are relative to the acromion marker, both are susceptible to errors in the other positions. In positions that require abduction, the scapula displays posterior tilt, resulting in the acromion rotating behind the glenoid cavity and causing the estimated centers to become posterior to the actual shoulder joint center. Conversely, in the 90º flexion-IR position, the scapula tilts anteriorly, causing the acromion to rotate in front of the glenoid cavity and produce estimated centers anterior to the actual shoulder joint center. Therefore, although the functional identification is more accurate in the adduction-ER position, both methods are characterized by substantial errors when implemented through a full range of motion.