Fig. 3 Mean and standard deviation of the manual tracking per- formance scores for all subjects (three subjects, two repetitions). The upper panel indicates the total movement of the cursor in each 4.4-s period (pathlength), while the lower panel indicates the “fixation error” – the average distance between the cursor and the fixation point. The data was not recoverable in every other hand- tracking condition
consistent with the active use of eye and right hand, as well as areas in striate and pre-striate cortex consistent with visual processing of the target and cursor motion. Because of the slice locations, we cannot report cerebral activation sites above the level of the basal ganglia (ap- proximately above z=16 mm, Talairach and Tournoux 1988); however, below this axial level, we did not ob- serve any areas of significant activation outside the cere- bellar, posterior parietal and occipital cortical areas.
Contrasting all tracking movement versus fixation
Comparing all the three tracking tasks (tasks A–C) ver- sus the baseline condition (task D), there was a broad ar-
ea of significant activation across the anterior lobe and anterior vermis of the cerebellum (Fig. 4B–D), with peak activation in the intermediate cortex of the ansiform lob- ule (Duvernoy 1995) ipsilateral to the tracking hand (Fig. 4C). There was also a significant activation site in the contralateral anterior lobe (Fig. 4C) and a strong acti- vation site in the ipsilateral posterior lobe of the cerebel- lum, in what we have identified as the ipsilateral para- median lobe, but which may also include parts of the dorsal paraflocculus (biventer lobule, Fig. 4F, G). There was a very small corresponding area of contralateral paramedian activation (Fig. 4F). There was also strong activation of visual cortical areas (Fig. 4A, B). There were no activated loci that could be unambiguously lo- cated in the cerebellar nuclei; this is not unexpected with functional imaging at 1.5T.
Activation during hand movement
Testing the main effect of hand movement (tasks A and C, both with hand movement, against tasks B and D, both without hand movement) demonstrated significant activation of the same cerebellar areas as shown in Fig. 4 (data not shown, see Table 2). Peak significance levels were slightly higher, and the extent of the activat- ed areas slightly larger. Thus, the major contributing factor to the contrast shown in Fig. 4 was that due to hand tracking movement. Contrasting just the hand-only condition with the fixation condition (task A vs D) con- firmed this; the same set of areas shown in Fig. 4 were activated, with comparable levels of significance (Table 2).
Activation during eye movement
Testing the main effect of eye movement (tasks B and C, both with eye movement, against tasks A and D, both with ocular fixation) resulted in no significant activation sites except in occipital areas. Primary visual cortex and an area corresponding in MT/V5 were activated, as they had been in other contrasts (e.g. Fig. 4). Contrasting the eyes-only condition against the fixation condition (task B vs. D) showed a small area of vermal activation just below the standard statistical thresholds used for all oth- er analysis (reduced from P=0.001, k=2 to P=0.01, k=1) in the same location as activated in Fig. 4D.
Activation during co-ordinated eye and hand movement
To examine activation due specifically to co-ordinated eye movement, we contrasted the eye-hand condition with hand-only condition (task C vs. A). This showed weak activation of the visual areas seen in Fig. 4, but no significant activation in the cerebellum. Thus, as before, the contribution of eye movement to the cerebellar acti- vation was negligible in this experiment.