Analysis of Cerebellar Motor Disorders by Visually-Guided Elbow Tracking Movement

In visually-guided slow ramp elbow tracking, patients with cerebellar ataxia show irregular undulations of pursuit velocity which result in a position tracking pattern unlike the smooth constant velocity or rate tracking pattern of normal controls (Beppu et al., 1984). This task provides ample opportunity to use both visual and proprioceptive feedback information for correcting errors. The present study investigated the role of visual information for generation of this saccadic pursuit pattern in the patients. A television screen was divided into upper and lower halves in each of which a vertical strip was displayed. The upper strip (T, target) was moved horizontally from the centre of the screen to the left or right by ramp voltage. The lower strip (D, displacement of the handle) was moved in proportion to angular displacement of the handle by a potentiometer coupled to the handle axis. The subject, while sitting in front of the screen, had to make D match the movement of T by controlling the handle with his arm. The range of T movement was 30 deg in terms of the angular movement of the handle. T velocity was 6.0 or 7.5 deg/s. After a training session, the test was performed in which D or T was suddenly erased from the screen during pursuit, depriving the subject of visual information about the moving limb and/or performance. The procedure gave only minor effects on the performance of the control subjects, but it reduced significantly the velocity undulation of the patients with cerebellar ataxia, producing a smoother continuous pursuit. There were no significant differences in performance between D or T erase tasks. The result supports the hypothesis that the marked undulation pattern during pursuit movement in cerebellar ataxia is due to repeated visually-guided error correction responses. The relative importance of the visual pathway for conveying position information and the proprioceptive pathway for movement velocity information in this visual slow ramp tracking task is also discussed.