Objective: We used fMRI and a visual cognitive task to study brain mechanisms underlying visual processing at different cognitive loads in patients with PD and PD with mild cognitive impairment (PD_MCI) as compared with healthy controls (HC).

Background: Visual processing deficits can be found in PD_MCI.

Methods: Altogether, 95 subjects (55 HC, 16 PD and 24 PD_MCI) performed a visual object matching task in a 3T MR scanner. The task consisted of pairs of conventional view images, unconventional view (spatially rotated) images, and a control task. Participants pressed a YES button if the second object of the paired images was the same as the first object (regardless of spatial orientation) or a NO button if they were different. The number of correct answers was an outcome measure. The effect of stimulation was computed using a general linear model implemented in SPM12. T statistic maps were computed to assess the effects of activation with respect to task conditions. A one-way ANOVA was used to assess the differences across groups followed by post-hoc t-tests. Group results were assessed using cluster level inference at p (FWE) < 0.05 at a height threshold of p (uncor) < 0.001. For group classification we used an ROC analysis.

Results: We found significant differences between groups only for the unconventional view task condition. The post-hoc analysis revealed that PD_MCI differed from HC (p=0.011). Based on behavioral results we evaluated our fMRI data using the unconventional vs. conventional condition contrast only. ANOVA revealed significant differences in the anterior cingulate cortex (ACC, p=0.013) and in the superior parietal lobule (SPL, p=0.017). Using a post-hoc t-test we found that PD_MCI and HC groups differed in activation of ACC (decreased in PD_MCI) while PD_MCI as compared to PD showed significantly decreased activation of both ACC (p=0.047) and SPL (p=0.004). ROC analysis of fMRI contrast in SPL distinguished PD_MCI from PD with 87.5% sensitivity and 86.98% specificity, AUC=0.94 (0.86-1.00).

Conclusions: Results of fMRI analysis using a visual object matching task with unconventional vs. conventional view contrasts revealed differences between PD_MCI and PD primarily in a component of the dorsal visual pathway implicated in mental rotation (SPL). ROC analysis of fMRI data distinguished PD_MCI from PD with > 85% sensitivity and specificity Our fMRI paradigm could be used as a biomarker for PD_MCI diagnosis.

« Back to 2016 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/brain-mechanisms-underlying-visual-processing-in-parkinsons-disease/