The specificity of visual recognition impairments following focal brain damage

Although the visual system is perhaps the most well understood system in the human brain, the precise organization of the neural system whose activity gives rise to higher order functions like visual recognition remains unknown. Furthermore, the manner in which damage involving this system relates to deficit, or the extent to which other factors modulate this relationship is unknown. Building on prior research in this laboratory and elsewhere, which has related focal brain damage to deficits in visual recognition pertaining to particular categories of stimuli, the present study examined both the specificity of lesion-deficit associations, and the relation between damage to the neural systems subserving visual recognition and the severity of a patient's impairment.

In the first part, I employed a novel method to address the specificity of visual recognition impairments in relation to the categories of faces, animals, fruits/vegetables and tools/utensils. By using voxelwise logistic regression to parse out variance that could be attributed to deficits across multiple categories, I was able to identify areas that were uniquely predicted by impairment in a single category. In the second part, I examined the relation between the extent of damage in these "category-specific" regions and the severity of the recognition impairment in the same four categories, as well as potential modulating effects from various demographic (e.g., sex, handedness), neuropsychological (e.g., premorbid intellectual functioning, visual-spatial and visuoperceptual ability), and lesion (e.g., age at onset, time elapsed since onset, extent of damage in other ROIs, lesion size) variables. The findings indicated that the largest factor accounting for performance in the recognition of these entities was the extent of damage in the respective category-specific regions. However, within each of the categories, there were additional factors that were also associated with performance, which helped explain some of the additional variance in recognition performance that could not be explained by extent of damage alone. With regard to the latter, I found that damage in certain category-specific regions was related to the severity of deficit across multiple categories, thereby reinforcing the notion of relative specialization within the visual system.