Abstract
INTRODUCTION: The basal ganglia (BG) modulate inferior temporal activity, which may account for visual hallucinations of complex items (bodies, faces) experienced by many patients with BG disorders like Parkinson's Disease (PD; Middleton & Strick, 1996). This project tested the hypothesis that BG mediate perceptual organization (PO), specifically with respect to two themes we identified from a review of PD clinical reports: part-whole integration and facial expression perception. METHODS: This project consisted of two studies with complementary methods: a re-analysis of an fMRI study on part-whole perception that showed BG activation in healthy participants (N=17), and a behavioral study with PD patients and age-matched controls (age > 40; both N=5). fMRI stimuli were 2D shapes composed of different local contour features. Three distinct block design conditions presented different subsets of these shapes, which permitted analysis of part-whole contingencies. Behavioral tests were the L-POST (Torfs et al., 2014), which assesses a range of PO processes; and a custom emotional expression recognition test based on the Karolinska Directed Emotional Faces. RESULTS: Bilateral caudate nucleus body activation was tightly correlated (t > 4.8, FDR p < 0.05) with the degree to which local features appeared in multiple distinct whole shapes within a set. No BG regions were activated above fixation baseline by shape viewing in general. The PD patients in our sample were not significantly impaired relative to controls on the behavioral tasks. However, controls performed below the L-POST normative range on the embedded figures subtest. CONCLUSION: We identified a specific BG in the caudate body that was activated by a specific part-whole contingency and not by visual stimulation. This activation may reflect the increased salience of local parts produced when the same part appears in multiple whole shapes. This in turn may relate to embedded figure perception, which was impaired in our participants.
Meeting abstract presented at VSS 2016