Gaze, Head Rotation, and Neuroanatomic Correlates of Reading Errors in Neglect Dyslexia - PROJECT SUMMARY/ABSTRACT Neglect dyslexia (ND) is an acquired reading disorder associated with spatial neglect (SN), commonly experienced by right hemisphere stroke survivors. ND causes whole-word omissions of left-sided words and errors on the left-sided letters of individual words (unilateral paralexias). These two distinct error types can be explained by dysfunctional egocentric (i.e., viewer-centered) and allocentric (i.e., object-centered) spatial cognition, respectively. Gaze asymmetry has long been considered a hallmark symptom of SN, and recent research suggests that it might contribute to these errors. However, no study to date has directly correlated gaze and reading in this population. Gaze involves the coordination of the head and eyes to orient toward a target. However, most gaze-related research in SN and ND thus far has required the head to be stabilized at midline. Therefore, relative contributions of the head and eyes to the asymmetry of gaze in SN and ND is unknown. Current neuroanatomical models of SN and ND suggest that it is caused by disruptions to the interaction of a bilateral dorsal attention network, which subserves spatial attention and eye movements, and a unilateral right hemisphere ventral attention network, which supports arousal, vigilance, detection of novel stimuli, and reorienting attention to novel stimuli. It is unknown how a network disruption leads to a specific type of ND. In this project, we will examine the predictive value of gaze on ND reading errors, the relative contributions of the head and eyes to gaze orientation, and the disrupted neural networks that underlie specific gaze patterns and egocentric and allocentric ND error patterns. (1) To characterize the association between gaze and reading, we will conduct two experiments that measure gaze with a wearable eye tracker during single-word reading. We expect to find a strong correlative relationship between gaze accuracy and reading accuracy in participants with ND, which would support the potential for the use of gaze metrics for ND diagnosis and classification. (2) To determine the relative contributions of the head and eyes to gaze deficits, we will conduct an experiment using eye tracking-enabled immersive virtual reality to measure gaze while manipulating the effect of head rotation. We expect participants with a greater contribution of the head to have poorer egocentric gaze toward left-sided words and more whole-word omissions. (3) To identify the neural networks that underlie gaze asymmetry and reading errors, we will lesion map brain scans of participants with and without ND, then use voxel-wise lesion- symptom mapping to correlate the degree of lesioned brain areas to gaze accuracy and reading accuracy. We anticipate that lesions disrupting the dorsal network impair egocentric gaze orientation and associate with whole- word omissions, and that lesions disrupting the ventral network impair allocentric gaze orientation and associate with unilateral paralexias.
