Crossmodal visual-tactile extinction: Modulation by posture implicates biased competition in proprioceptively reconstructed space

Abstract

Extinction is a common consequence of unilateral brain injury: contralesional events can be perceived in isolation, yet are missed when presented concurrently with competing events on the ipsilesional side. This can arise crossmodally, where a contralateral touch is extinguished by an ipsilateral visual event. Recent studies showed that repositioning the hands in visible space, or making visual events more distant, can modulate such crossmodal extinction. Here, in a detailed single-case study, we implemented a novel spatial manipulation when assessing crossmodal extinction. This was designed not only to hold somatosensory inputs and hand/arm-posture constant, but also to hold (retinotopic) visual inputs constant, yet while still changing the spatial relationship of tactile and visual events in the external world. Our right hemisphere patient extinguished left-hand touches due to visual stimulation of the right visual field (RVF) when tested in the usual default posture with eyes/head directed straight ahead. But when her eyes/head were turned to the far left (and any visual events shifted along with this), such that the identical RVF retinal stimulation now fell at the same external location as the left-hand touch, crossmodal extinction was eliminated. Since only proprioceptive postural cues could signal this changed spatial relationship for the critical condition, our results show for the first time that such postural cues alone are sufficient to modulate crossmodal extinction. Identical somatosensory and retinal inputs can lead to severe crossmodal extinction, or none, depending on current posture.

Figure 1

Schematic diagram showing the layouts used in Expt 1. Shaded squares (shown transparent here, but opaque in the experiment) depict the small shields that occluded the tactile stimulators and stimulated fingers from the patient's view in Expt 1. Note that the right-visual-field visual stimulus becomes positioned over the left hand in the deviated posture, but that the retinotopic visual location and somatotopic tactile location of the experimental visual and tactile stimuli are themselves unchanged across postures. In Expt 2 the whole array, plus the hands, arms, and shoulders of patient JM, were occluded by a large sheet. The leftmost shaded square represents a dummy shield on which was placed the left visual-field stimulus for the deviated posture.

Figure 2

Transverse section from the cranial CT scan of JM. The darker area is consistent with infarction in the territory of the right middle cerebral artery. The lesion involves regions of the parietal, temporal, and frontal lobes (note that the bright outline circle is an artifact of the image production).

Figure 3

Percentage left tactile detections in unilateral and bilateral trials for Expt 1. Separate plots are presented for (a) left touch/right vision crossmodal trials and for (b) left touch/right touch unimodal trials. Bold lines depict performance in the deviated posture, thinner lines are for the standard posture. Note that the actual data for left touch/right touch (in b) overlap perfectly; the slight vertical offset between the two lines is added here only so that both can be seen.

Figure 4

Percentage left tactile detections in unilateral and bilateral trials for Expt 2. Separate plots are presented for left touch/right vision crossmodal trials (a) and left touch/right touch unimodal trials (b) Bold lines depict performance in the deviated posture, thinner lines are for the standard posture. Dotted lines (thick for deviated posture, thin for standard) are for when the arms and hands were completely occluded with a sheet.