Distorted gaze direction input to attentional priority map in spatial neglect

Abstract

A contribution of the gaze signals to the attention imbalance in spatial neglect is presumed. Direct evidence however, is still lacking. Theoretical models for spatial attention posit an internal representation of locations that are selected in the competition for neural processing resources - an attentional priority map. Following up on our recent research showing an imbalance in the allocation of attention after an oculoproprioceptive perturbation in healthy volunteers, we investigated here whether the lesion in spatial neglect distorts the gaze direction input to this representation. Information about one's own direction of gaze is critical for the coordinate transformation between retinotopic and hand proprioceptive locations. To assess the gaze direction input to the attentional priority map, patients with left spatial neglect performed a cross-modal attention task in their normal, right hemispace. They discriminated visual targets whose location was cued by the patient's right index finger hidden from view. The locus of attention in response to the cue was defined as the location with the largest decrease in reaction time for visual discrimination in the presence vs. absence of the cue. In two control groups consisting of healthy elderly and patients with a right hemisphere lesion without neglect, the loci of attention were at the exact location of the cues. In contrast, neglect patients allocated attention at 0.5⁰-2⁰ rightward of the finger for all tested locations. A control task using reaching to visual targets in the absence of visual hand feedback ruled out a general error in visual localization. These findings demonstrate that in spatial neglect the gaze direction input to the attentional priority map is distorted. This observation supports the emerging view that attention and gaze are coupled and suggests that interventions that target gaze signals could alleviate spatial neglect.

Keywords: Attention; Coordinate transformation; Spatial neglect; Stroke.

Graphical abstract

Fig. 1

Individual lesions of the patients with spatial neglect. The colour maps indicating the lesion are superposed on a single-subject T1 template coregistered with the MNI152 template (International Consortium for Brain Mapping). The figure shows the vertical, z coordinate for each transversal slice. The left side of the brain is shown to the left. The demographics of the individual patients are listed in Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Individual lesions of the patients in the control group. Conventions like in Fig. 1.

Fig. 3

Setup and task for the cross-modal attention experiment. A. Participants fixated at +19° from the body/head midline. A hand proprioceptive cue (the participants' right index finger, hidden from view) was positioned at one out of four possible locations: 8°, 18°, 20° or 30° to the right from body-midline. B. A target letter, ‘A’ or ‘H’ was presented for 100 ms at one out of seven possible locations, at −3, −2, −1, 0, +1, +2 and +3 degrees horizontally from the finger. Participants named the letter as fast and accurately as possible.

Fig. 4

Setup and task for the visual open-loop pointing experiment. A.

Fig. 5

Reaction time and accuracy for visual discrimination in the right body hemispace in the absence of a visual cue. Neglect patients' (●) voice reaction time (A) and accuracy (B) was not significantly different than that of the patient control group () and healthy controls (◯) for targets located further than 11° to the right of the body midline (error bars show one standard deviation, * denotes p< 0.05, independent samples t-tests, neglect vs. control, for either control group).

Fig. 6

Reaction time for visual discrimination in the cross-modal attention task in the presence () and the absence of the finger position cue (). The arrow indicates the location of the cue in trials when the cue was present. NEG – patients with spatial neglect, HEC – healthy elderly controls, PCG – patient control group. Error bars show one standard deviation.

Fig. 7

Neglect patients show a rightward shift of the loci of attention in the visual space relative to the finger position cue. The locus of attention was defined as the location with the largest advantage in reaction time for visual discrimination in the presence of the cue. The cueing error in neglect patients was in average 1.9°, 1.4° and 0.5° when the cue was presented at 18°, 20° and 30° from the body midline respectively. This was significantly different from the error in the control groups (p<0.001, one-tailed, independent samples t-tests and Mann-Whitney U test). NEG-neglect; individual patients with spatial neglect KS-, ASG-, SG-, UH-; PCG - patient control group; HEC – healthy elderly controls.

Fig. 8

Neglect patients show a rightward shift that is selective for the allocation of attention. Their locus of attention in the visual space was in average at 1.25° right of the finger position across all tested locations (18°, 20° and 30° from the body midline). There was no statistically significant difference open-loop pointing error between neglect and control groups at any of these locations, which rules out a general error in coordinate transformations as an explanation of the results. NEG-neglect; individual patients with spatial neglect KS-, ASG-, SG-, UH-; PCG-patient control group; HEC – healthy elderly controls.