Cortical-cerebellar relationship during continuous motor learning measured by PET H.I. van Mier, J.S. Per lmutter , M.E. Raichle and S.E. Petersen Washington University School of Medicine, Box 8111, St. Louis, MO 63110 Introduction. Data from neurophysiological, neuroanatomical, and lesion studies have suggest the existence of a strong relationship between many areas in the cerebral cortex with different parts of the contralateral cerebellum. We report data from a PET study on motor learning which show distinct functional similarities between sets of cortical and cerebellar regions. Such information could prove to be useful adjunct to connectional studies outlining the relationships between the cortical and cerebellar regions. Methods. We tested 32 normal right handed subjects who moved a pen continuously through a cut-out maze and square pattern as quickly and accurately as possible with their eyes closed. A first group of 16 subjects performed the task with their dominant right hand in a clockwise direction; a second group used their non-dominant left hand and traced the designs in a counter-clockwise direction. Activation was measured by comparing a condition where subjects held the pen in their hand without moving it (REST) to several maze tracing conditions: 1) during initial unpracticed performance (NAIVE), 2) after 10 min of practice on the same maze (PRAC) and 3) following practice with the task, but tracing a novel maze (NOVEL). To distinguish practice related from motor performance changes (e.g. changes in movement velocity) subjects were scanned while tracing an overleamed square design at high (SQ. FAST) as well as at low speed (SQ. SLOW). These conditions were also compared to REST. Results. Three types of effects were found 1. Movement related effects (increases in activity across conditions which shifted lateralization with the hand used) in the contralateral primary motor area and ipsilateral anterior, medial cerebellum. Localizations for the other two effects were the same for right and left hand performance. 2. Practice related effects (higher activations during NAIVE and NOVEL conditions than in PRAC, SQ.FAST and SQ. SLOW) in right pailetal and premotor cortex and left cerebellum, and 3. Capacity related effects (higher activation during NAIVE, NOVEL, PRAC, and SQ.FAST than SQ. SLOW) in left premotor and right cerebellar areas. As can be seen in Fig. 1, the cortical and cerebellar localizations of the effects are consistently contralateral. Z = 5 6 Z = -12 ~Movementrel~ed effe~ O Pra~ice rel~ed effe~ ~ ~ , CapaciW related effect Right Hand A performance [] Left Hand performance Z = -32 Figure 1. Primary motor, premotor, parietal and cerebellar activations displayed in Talairach '88 space. Conclusion. The consistency of the activation pattems and effects between specific cortical regions and contralateral cerebellar regions have implications on at least two levels. The results suggest physiological and probably anatomical relationships between specific cortical and cerebellar regions. Furthermore, the specific task related effects implicate separate functional relationships between premotor and cerebellum depending on their lateralization.