Summary

In this chapter a new and automatic method was described to determine ocular torsion (OT) from digitized video images (Video-oculography). We developed this method based on the tracking of iris patterns. Instead of quantifying OT by means of cross-correlation of circular iris samples, which is commonly applied, this new method automatically selects and recovers a set of 36 significant patterns in the iris by a technique of template matching. Each relocated landmark results in a single estimate of the torsion angle. A robust algorithm estimates OT from this total set of individually determined torsion angles, thereby largely correcting for errors which may arise as due to misjudgments of the rotation center. In a prepared set of images of an artificial eye the new method reproduced OT with an accuracy of 0.1º. In a sample of 256 images of human eyes, a practical reliability of 0.25º was achieved.

Chapter 2

The objective of this experiment was to assess evidence for vestibular adaptation to prolonged hypergravity in human subjects as to substantiate previously described effects, such as postural imbalance and motion sickness (Sickness Induced by Centrifugation, SIC). We measured the ocular torsion response in eleven subjects during static and dynamic body tilt, once before and once after an one-hour centrifuge run of +3Gx. The OT response to static tilt (in the range of 0 to 57º to either side) showed a 10% decrease, suggesting a reduced otolith gain. The otolith-canal interaction was examined by comparing the dynamic OT response to sinusoidal body roll (frequency of 0.25 Hz and amplitude of 25º) about an earth-horizontal rotation axis (stimulation of both otoliths and canals) and about an earth-vertical rotation axis (stimulation of canals). After centrifugation, the gain of the slow component velocity increased in both conditions in all but four subjects, who showed a decrease in the supine condition (but not in the upright condition). These four subjects developed symptoms of SIC, so that the different behavior of their SCV gain was likely due to a declined state of alertness which specifically may have occurred in the supine condition. In addition to the OT data, the horizontal VOR was measured in response to a velocity step rotation about the vertical yaw axis. The mean gain of the horizontal VOR was unaffected by centrifugation, but the dominant time constant was significantly reduced. Because the time constant of the horizontal VOR is centrally controlled by the velocity storage mechanism, this result provides evidence for vestibular adaptation at a central level.

Chapter 3

In the experiment of Chapter 2 the dynamic OT response (or torsional VOR) was measured to study possible effects of centrifugation on the canal-otolith interaction. However, the response did not show a clear otolith contribution. Supposedly, the stimulus frequency of 0.25 Hz had been too high to reveal an otolith component. The response was therefore studied in more detail at a wider frequency range. The ocular torsion response was examined during passive sinusoidal body roll in five human subjects. To separate the otolith organ and semicircular canal contributions, again the axis of rotation was varied between earth-horizontal and earth-vertical. At a fixed amplitude of 25°, the stimulus frequency was varied from 0.05 to 0.4 Hz. Additionally, at a fixed frequency of 0.2 Hz, the response was also measured at the amplitudes of 12.5° and 50°. The results showed that the gain and phase of the slow component velocity (SCV) did not depend on stimulus amplitude, indicating a linear response. Contribution of the otoliths affected the ocular torsion response in three different ways. First, the gain of the SCV was slightly but consistently higher during earth-horizontal rotation than during earth-vertical rotation. In the supine orientation the average gain increased from 0.10 to 0.26. In the upright orientation the average gain increased from 0.14 to 0.37. Second and more substantially, modulation of the otolith inputs improved the response dynamics by reducing the phase lead at frequencies up to 0.2 Hz. Third, the nystagmus showed considerably less anti-compensatory saccades in upright conditions than in supine conditions, even though the SCV gain was lower in the latter. As a consequence, the average excursion of torsional eye position was highest during earth-horizontal rotation. This effect was observed in the entire frequency range. Thus, the otoliths did not only control the torsional VOR at low stimulus frequencies by keeping the slow component in phase with head motion, but also in a wider frequency range by modulating the saccadic behavior as to increase the excursion range of torsional eye position. We conclude that, during head tilt, the primary concern of the otolith-oculomotor system is to stabilize eye position in space, rather than to prevent retinal blur. This confirms that tilt otolith-induced ocular responses subserve spatial orientation.

Chapter 4

This chapter describes a study on the effectiveness of a highly polarized visual environment to induce sensations of self-motion and self-tilt in a stationary observer. The subject were immersed in an 8 foot cubic room which could be fully rotated about an earth-horizontal axis. The interior of the room was filled with common visual features such as a door, a window, and a great variety of objects which indicated up and down (visual polarity). When the room was tilted about the roll axis of an erect observer it produced illusory self-tilt by virtue of its visual polarity alone. Although the effect was larger than is known from the literature, the experienced self-tilt did not linearly increase with room tilt. At higher angles of room tilt (80 and 120º) the judgment of verticality became more variable and depended less on the visual scene. The room induced complete self-rotation in more than 80% of the cases when it was rotated at constant velocity about a stationary subject in various body positions. This strong effect was attributed to both its motion and its visual polarity.

Chapter 5

This chapter describes a study on the visual-vestibular interaction in the judgment of the body orientation relative to gravity. Illusory self-tilt and self-motion (vection) produced by rotation of a full-field non-polarized visual scene about the subject’s roll axis was measured as a function of the presence or absence of actual rotation of the subject during visual acceleration. Subject rotation was at two levels of acceleration and with or without a delay between initial rotation and subsequent return (washout) to the vertical position. In one set of conditions, visual motion and self-motion were in opposite directions (concordant) and in another set they were in the same direction (discordant). For concordant motion the main effect of body rotation was to reduce the time taken by the subject to indicate self-rotation. The magnitude of self-tilt was increased by actual body tilt as could be expected from addition of the perceived actual body tilt and illusory body tilt induced by visual rotation. This effect of augmented body tilt did not persist after the body was returned to the vertical. The magnitude of vection was not influenced by body rotation and washout. For discordant motion of the body and the visual scene, subjects were confused and their responses were very variable. This suggests a non-linear visual-vestibular interaction, in which perceived self-tilt and self-motion are strongly determined by visual inputs, except for discordant accelerations of the body and the visual surroundings. Then the perception is determined by the vestibular inputs.