A previous study found that a high level pitcher might generate greater joint forces and torques in order to produce greater shoulder and elbow angular velocities during skillful pitching performance (Fleisig et al. 1999). On the other hand, in multijoint limb movement, torque at one joint occurs not only from muscles acting on that joint but also from interactions due to the rotations of other joints. Therefore, it is unlikely that the joint motions of the throwing arm will be due exclusively to the muscles at the corresponding joints. There have been many studies analyzing throwing arm dynamics in baseball pitch but none have yet determined the contributions of joint torques and motion-dependent moments to the distal endpoint velocity of upper extremity in multijoint throwing arm motion. Therefore, the purpose of this study was to quantify the contributions of joint torques and motion-dependent moments to produce the hand velocity of the throwing arm during an overhand baseball pitch. A throwing arm dynamical model consisting of 4 segments (trunk, upper arm, fore arm and hand) and 7 degrees of freedom (DOF) of shoulder, elbow and wrist joints was developed. Four baseball pitchers participated in the experiment and their pitching motions were recorded by high-speed video camera. Joint motions and torques of 3 joints were calculated using the DLT method and inverse dynamics. An equation of motion in the dynamical throwing arm model was utilized to calculate the contributions of joint torque-dependent and motion-dependent components to the hand velocity. These dependent velocities were then decomposed into the components of 7 DOF. The results showed that the joint torque-dependent component was the greatest contributor, the angular velocity-dependent component was the second contributor, and the trunk motion-dependent component was the third contributor in producing the hand velocity of throwing arm. Analysis of these contributions in all joint motions implied the following;

a)The joint torque-dependent velocities of the shoulder horizontal adduction-abduction, external-internal rotation, wrist flexion-extension and supination-pronation contributed to the positive hand velocity.

b)The joint torque-dependent velocities of the elbow extension-flexion and wrist ulnar-radial deviation contributed to the negative hand velocity.

c)The shoulder internal rotation torque generates rapid internal rotation motion and plays a major role for producing high hand velocity.

d)The angular velocity-dependent component and the trunk motion-dependent component cause positive hand velocity accompanied with rapid elbow extension.

e)The motion-dependent moments due to the rotations of the throwing arm segments and trunk affected the throwing arm motion and were helpful to produce the positive hand velocity.

The scapular movement could not be monitored directly with the currently available video-based method because the scapular moves widely under subcutaneous tissues. In the present study, a convenient, video-based method was developed to estimate the scapular rotation indirectly for given humeral rotation measured in dynamic activities. Four male university students with no shoulder injury participated in the present study. Each subject was asked to hold the right arm at each of the 25 selected positions. While the subject was holding the position, reflective markers were attached on the skin above the selected scapular anatomical landmark and the three-dimensional coordinates of the markers were measured with optical motion capture system. The orientations of the scapula and humerus relative to thorax were expressed as Euler angles. The scapular rotations were expressed as a function of humeral rotations. The derived functions were then used to estimate the scapular rotations for given set of humeral rotations during humeral elevation. The validity of the present methodology was tested with the use of electromagnetic sensor. The present method well predicts scapular internal rotation for horizontal adduction greater than 30° and scapular upward rotation during humeral elevation. The scapular posterior tilt was not estimated accurately.