Abstract
Lesion studies of the cerebellar vermis (Optican and Robinson, 1980) and the fastigial nuclei (Goldberg et al. 1993) showed that the cerebellum is responsible for the saccade adaptation. The adaptive modification of the vestibulo-ocular reflex was explained by a model (Fujita, 1982) which introduced the long-term depression (LTD) and potentiation (LTP) of the transmission efficacy from parallel fibers (pf) to Purkinje cells (P-cells). LTD and LTP were demonstrated to be responsible for changes in the receptive field of the P-cells (Jorntell and Ekerot, 2002). Here we present a cerebellar model with the same LTD and LTP mechanism to explain the saccade adaptation in the amplitude as well as direction by repetitive intrasaccadic target displacement (McLaughlin, 1967, Miller et al. 1981, Deubel, 1987).
We assume the followings:
When the image of a target is outside the fovea at the end of a saccade, it evokes a climbing fiber response retinotopically on the contralateral cerebellar vermis. It reflects a simple perceptual, not cognitive process.
Pf responses appears bilaterally and extensively on the vermal areas.
Projection from the P-cells to target neurons in the fastigial nuclei is anatomically and functionally fixed.
Projection from the fastigial cells to burst neurons in the premotor areas for the vertical and horizontal saccadic components as well as to the superior colliculus is also anatomically and functionally fixed.
Synaptic efficacy from pf to P-cells or inhibitory interneurons is modified by the LTD and LTP.
Above-mentioned physiological data and the five assumptions leads us to successful computer simulation of the saccade adaptations both in the amplitude and direction. The more elaborate mechanisms for the selective adaptation of the human saccades (Fujita et al. 2002) remained unresolved.