Adaptation of a bimodal integration stage: visual input needed during neck muscle vibration to elicit a motion aftereffect

Abstract

Vibratory stimulation of the neck muscles can elicit illusory drift of a visual target; after vibration stops, motion in the opposite direction is perceived. This motion aftereffect (MAE) could be due to adaptation of proprioceptive mechanisms that encode head orientation, or at a stage where visual and proprioceptive information are combined. To distinguish between these two possibilities, we applied vibratory stimulation to dorsolateral neck muscles for 15-s periods alternating with 15-s periods without vibration. Twenty-six observers used a hand-held tracker to indicate perceived motion of a stationary light-emitting diode (LED) in an otherwise dark room. In the critical condition, observers were in complete darkness during vibration, and the LED was only turned on in post-vibration periods. If adaptation was purely proprioceptive, a visual MAE should have occurred in this condition, but it did not. In a follow-up experiment (N = 9), the LED was presented intermittently to determine if there was a position aftereffect that might have been inhibited by processes signalling an absence of motion. No aftereffect occurred under these conditions either. In both experiments, a visual stimulus had to be present during the adaptation period in order to elicit an aftereffect. Results from our previous study ruled out an explanation based on suppression of eye movements. Thus, the most likely site responsible for the visual aftereffect lies with bimodal mechanisms combining proprioceptive and visual information. We conclude that the bimodal mechanisms adapted more quickly than the proprioceptive mechanisms from which they received input.

Appendix

Eye movements

Throughout the history of vision research, eye movements have been proposed to explain various motion illusions: the classical motion aftereffect (Purkinje 1825; Addams 1834; quoted in Wade 1998, p. 213–214), the oculogyral illusion (Graybiel and Hupp 1946), the autokinetic effect (see review by Levy 1972), and the illusion that is the topic of the present study (Popov et al. 1999). The crucial element in these explanations is the idea that the brain does not register involuntary or reflex eye movements, but does register the retinal slip of images arising from these eye movements. It is also assumed that the brain registers the commands issued to the extraocular muscles to pursue the target or to maintain fixation countering the reflex tendencies, interpreting them as eye movement even if the eye remains stationary.

Of the various motion illusions to which eye movement explanations were applied, the one most similar to the illusion presently discussed is the oculogyral illusion that occurs in response to vestibular stimulation. Vestibular stimulation is similar to neck-proprioceptive stimulation in that they both indicate head rotation and trigger reflex eye movements. However, unlike the vestibulo-ocular reflex, reflex eye movements triggered by proprioceptive input from the neck—such as those elicited by rotating the trunk relative to the stationary head (the cervico-ocular reflex)—are not necessarily compensatory in sign relative to the direction of head movement: for some velocities, they are actually anticompensatory, moving in the same direction as the head (Mergner et al. 1998). These eye movements are also much more variable in terms of direction across subjects than eye movements triggered by vestibular stimulation (see Fig. 3 in Mergner et al. 1998).

The eye movements elicited by vibratory stimulation of the neck muscles are also variable and therefore they are often inconsistent with the direction of illusory visual motion, which is more uniform across subjects. Similarly, eye movements that occur after adaptation through neck muscle vibration are often inconsistent with the perceived motion aftereffect (Lackner and Levine 1979; Seizova-Cajic et al. 2006). These findings are incompatible with the idea that eye movements or oculomotor commands cause the aftereffects. Details of a possible explanation of the aftereffect based on oculomotor commands are described below.

Since neck muscle vibration elicits eye movements that are suppressed when a fixation target is present, it is possible that, when vibration stops, the aftereffect of the suppression signal may be interpreted as image motion, leading to a visual aftereffect (a similar explanation was proposed in relation to the autokinetic effect that occurs following a prolonged extreme eye turn; see Levy 1972, for a review). The absence of an aftereffect in the Dark–Dot condition would thus be explained by the absence of a suppression signal during vibratory stimulation since there was no fixation target (as suggested by an anonymous reviewer). However, in our previous study (Seizova-Cajic et al. 2006) we also presented evidence that argues against the suppression signal as the cause of the aftereffect. In the remaining part of this Appendix, we clarify our method to infer the direction encoded by oculomotor commands that might suppress eye movements in the presence of a fixation target, and re-analyze data from our previous study to determine whether such motor commands could explain the aftereffect.

Although the eye movements suppressed during fixation can obviously not be measured directly, we might reasonably assume that the eye movements exhibited in the absence of a fixation target are the ones that are suppressed in the presence of a fixation target. From this information, the direction of the suppression command can be inferred and compared to the direction of illusory motion during the aftereffect.

To test the suppression explanation, we re-analyzed results from our previous study which included measurement of eye movements during neck muscle vibration in darkness as well as the direction of illusory motion after vibration in the presence of a fixation target. The methods and analysis were very similar to those described in the current study, with variations in the placement of vibrators. The top row of Fig. 11 shows results for vibrators applied simultaneously to the left and right SCM muscles, while the bottom row shows results for a single vibrator placed against the right posterior region of the neck, stimulating primarily splenius capitis and possibly the SCM. Results are shown for participants for whom both eye movements and illusory motion were significantly different from zero (taken from Figs. 7, 8 in Seizova-Cajic et al. 2006; eye movements are taken from the third column in each of the original figures, where they were shown as black lines, and pointing data is from the first column, shown there as grey lines).

Fig. 11

(Appendix). Comparison of motion aftereffect to eye movements during neck muscle vibration in the dark. Test of hypothesis that motion aftereffect is caused by an aftereffect of suppressed eye movements. Data are taken from Figs. 7, 8 in Seizova-Cajic et al (2006); see text for details. The top row shows results for vibration applied simultaneously to SCM muscles on both sides of the neck and the bottom row for vibration applied to splenius capitis on the right side of neck. Black lines show direction of the motion aftereffect following adaptation in the presence of a fixation target (pointing data). Grey lines show direction of cumulative slow-phase eye movements during vibration in complete darkness. Note that the scale for eye movements differs from the scale for pointing movements indicated on the axes for all participants except SG: for participants CB, JO, and TSC, eye movements have been multiplied by 1/2, and by 2 for JP

For all participants, and for both stimulation sites, the direction of the motion aftereffect following neck muscle vibration in the presence of a fixation target was predominantly downward, as indicated by the solid black lines in Fig. 11. The grey lines show the cumulative slow-phase eye movements recorded during neck muscle vibration in the dark (see Seizova-Cajic et al. 2006, for details). For JP, and somewhat less so for TSC, these eye movements in the dark are roughly in the same direction as the aftereffect, as predicted by the suppression account. However, the direction of eye movements differs by about 45° from the direction of illusory motion for SG, by almost 90° for CB, and by more than 90° for JO.

This extensive variation between the direction of eye movements in the dark and the direction of illusory motion is incompatible with a causal relationship. We can thus rule out the suppression motor command as the main cause of the motion aftereffect. By a similar account, we showed in our previous study that eye movements suppressed during the test phase fail to explain the aftereffect. Overall, eye movements concurrent with the illusion and aftereffect, as well as imputed oculomotor commands, vary greatly in direction, and thus fail to explain the visual phenomena, which are quite consistent.