Biomechanics of upper extremity movements in piano keystroke Shinichi Furuya1, Tomoko Aoki2, Hidehiro Nakahara1 and Hiroshi Kinoshita1 Graduate School of Medicine, Osaka University, Japan, email: [email protected] 2 Department of Environmental and Symbiotic Sciences, Prefectural University of Kumamoto

1

INTRODUCTION Artistic piano performance requires skills of manipulating many acoustic variables of music. The level of sound pressure (SPL), and keystroke tempo (ST) are of particular importance among these. Although trained pianists can control these variables effortlessly, little is know about motor control of their keystroke motion. The present study investigated the effects of SPL and ST on control of the upper extremity movements and the associated muscular activity during striking the keys at various combinations of SPL and ST. METHODS Eight highly-trained pianists struck two keys (G3 and G4; octave width) by the thumb and little finger at combinations of SPLs (pp=89.0, p=92.4, mf=95.8, f=99.2, and ff=102.6 dB: range = ± 5%) and STs (1, 2.25, 3.5, 4.75, and 6 Hz: range = ± 5%). The striking touch was limited to “staccato”. At each of SPL and ST combinations, 30 repetitive strikes were performed and the data were recorded. The upper limb kinematics were recorded using a 2D position sensor camera. LEDs were mounted on the centers of the MCP, wrist, elbow, shoulder and hip joints. The vertical movement of the G3 key was also recorded by another position sensor camera. Using a telemetric surface EMG system, muscular activities from the right side of 6 upper limb muscles were recorded. These were the pectoralis major, latissimus dorsi, triceps brachii, biceps brachii, flexor digitorum superficialis, and extensor digitorum communis. RESULTS Kinematics: With an increase in SPL, the range of vertical movements of all limb segments, and the peaks of the hand descending velocity and angular velocities for all joints were increased (Fig. 1). The rate of these increases was greater at higher SPLs (f and ff). The effect of ST on the limb kinematics was more complicated compared to the SPL effect. With increased ST, the movement ranges of all segments were decreased (Fig. 1). On the other hand, the peak descending velocity of the hand remained constant across all STs. The peak angular velocity for the wrist joint decreased with ST while those for the elbow and shoulder joints remained nearly constant at all STs (Fig. 1). ANOVA revealed significant SPL x ST interaction effects for the angular movement range of the shoulder, and peak angular velocity at the elbow joint. EMGs: A burst of activity was revealed in all muscles of the upper limb prior to the finger-key contact moment. The magnitude of the peaks of all muscular activity was clearly SPL- and ST-dependent. The greater the SPL, and the higher the ST, the greater the amplitude of the muscular activity in all muscles examined (Fig. 2). ANOVAs revealed a significant SPL x ST interaction effect on the mean EMG value for all muscles. That is, these values increased with SPL, and this increase was much larger at faster STs. DISCUSSION SPL-related increases in joint angular velocity and co-contraction (CC) of the agonist and antagonist pairs of all upper limb muscles before keystroke were observed. Because the CC relates to joint stiffness, it is likely that by increasing

joint stiffness and movement speed, the pianists increase the amount of the momentum transfer from the hand into the key. The effect of SPL x ST interaction was found for the range of shoulder joint rotation, and peak elbow joint velocity. That is, with an increase in SPL, the range of shoulder joint motion was increased but this increase was much smaller at faster STs. Likewise, velocity of the elbow joint was increased with SPL, and this increase was much larger at faster STs. When striking at a slower ST, quickly switching of the movement direction from downward to upward, or vice versa is less required. In this case, the pianist can benefit from the use of proximal joint motion that increases not only the endpoint velocity but also the effective mass involved in the finger-key contacting dynamics. However, when striking at a faster ST, the upper-arm motion is commonly constrained by a strong CC of the shoulder agonist and antagonist muscles. This is because the motion at the proximal segment having a greater inertia costs energetically higher to accelerate and decelerate during repetitive motion than the distal segment with smaller inertia. The movement at the elbow joint is then the best way to compensate for the decrease in the upper-arm motion. This is because less muscular effort is required to make a switching movement at the elbow joint than at the shoulder joint because of the involvement of segments with less moment of inertia. It is also attributed to the fact that the muscles for the elbow rotation can definitely generate larger force than those for the wrist rotation.

Fig. 1

Fig. 2