Abstract
Introduction: Visual field loss can result from damage to the primary visual cortex (V1). In spite of this visual field loss, some patients show residual visual function in their blind region, likely due to extrastriate visual area function. Therefore, the location of the lesion may be linked to residual vision. Here we measured the lesion size and location across a large group of patients with hemianopia and correlated this to residual vision. Methods: Four observers independently created lesion masks on structural MRI scans for 35 patients with hemianopia (mean age=50.4; 10 female). Observer masks were combined and thresholded to produce ‘consensus masks’ that only included voxels identified by at least 3 of the observers. The extent to which each area was affected was calculated by using the intersection of the ‘consensus masks’ and visual areas V1, V4 and V5 (Jülich atlas). Residual vision was determined using two 2AFC detection tasks presented in the perimetry-defined blind field: a high-contrast (50% and 100%) Gabor patch detection task (n=35) and a motion detection task (n=26; 5 degree patch of coherent black dots moving at different speeds on mid-grey background). Percentage correct was calculated and related to the proportion of damage to visual areas V1, V4 and V5. Results: Residual visual performance on both tasks was inversely correlated to the proportion of V1 (Contrast: r=-0.43;p=0.01; Motion: r=-0.50;p=0.01) and V5 (Contrast: r=-0.50;p=0.003; Motion: r=-0.40;p=0.04) damaged. In contrast, while V4 damage correlated with residual behaviour on the contrast task (r=-0.40;p=0.02), this was not the case with the motion detection task (r=-0.34;p=0.09). Conclusion: The extent of damage to V1, V4 and V5 is significantly related to the ability to detect contrast-defined stimuli in individuals with visual field loss. However, detecting moving stimuli related only to damage in V1 and V5, consistent with the known visual motion pathway.