ORIENTATION OF HUMAN OPTOKINETIC NYSTAGMUS TO GRAVITY - A MODEL-BASEDAPPROACH

Optokinetic nystagmus (OKN) was induced by having subjects watch a mov ing display in a binocular, head-fixed apparatus. The display was comp osed of 3.3 degrees stripes moving at 35 degrees/s for 45 s. It subten ded 88 degrees horizontally by 72 degrees vertically of the central vi sual field and could be oriented to rotate about axes that were uprigh t or tilted 45 degrees or 90 degrees. The head was held upright or was tilted 45 degrees left or right on the body during stimulation. Head- horizontal (yaw axis) and head-vertical (pitch axis) components of OKN were recorded with electro-oculography (EGG). Slow phase velocity vec tors were determined and compared with the axis of stimulation and the spatial vertical (gravity axis). With the head upright, the axis of e ye rotation during yaw axis OKN was coincident with the stimulus axis and the spatial vertical. With the head tilted, a significant vertical component of eye velocity appeared during yaw axis stimulation. As a result the axis of eye rotation shifted from the stimulus axis toward the spatial vertical. Vertical components developed within 1-2 s of st imulus onset and persisted until the end of stimulation. In the six su bjects there was a mean shift of the axis of eye rotation during yaw a xis stimulation of approximate to 18 degrees with the head tilted 45 d egrees on the body. Oblique optokinetic stimulationwith the head uprig ht was associated with a mean shift of the axis of eye rotation toward the spatial vertical of 9.2 degrees. When the head was tilted and the same oblique stimulation was given, the axis of eye rotation rotated to the other side of the spatial vertical by 5.4 degrees. This counter rotation of the axis of eye rotation is similar to the ''Muller (E) ef fect,'' in which the perception of the upright is counterrotated to th e opposite side of the spatial vertical when subjects are tilted in da rkness. The data were simulated by a model of OKN with a ''direct'' an d ''indirect'' pathway. It was assumed that the direct visual pathway is oriented in a body, not a spatial frame of reference. Despite the s hort optokinetic after-nystagmus time constants, strong horizontal to vertical cross-coupling could be produced if the horizontal and vertic al time constants were in proper ratio and there were no suppression o f nystagmus in directions orthogonal to the stimulus direction. The mo del demonstrates that the spatial orientation of OKN can be achieved b y restructuring the system matrix of velocity storage. We conclude tha t an important function of velocity storage is to orient slow-phase ve locity toward the spatial vertical during movement in a terrestrial en vironment.