Quantitative susceptibility mapping analyses of white matter in Parkinson's disease using susceptibility separation technique

Abstract

Introduction

To apply susceptibility separation on quantitative susceptibility mapping (QSM) images of Parkinson's disease (PD) patients to obtain more accurate images and gain pathophysiological insights.

Methods

This retrospective study included subjects who underwent head MRI, including QSM between March 2016 and March 2018. Patients with PD were categorized as having mild cognitive impairment (PD-MCI), or normal cognition (PD-CN); healthy controls (HC) were also enrolled. Susceptibility separation generated positive (QSM+) and negative susceptibility (QSM-) labels. Voxel-based whole-brain susceptibility and atlas-based susceptibility were compared among groups on white matter. Correlations between susceptibility and Montreal Cognitive Assessment (MoCA) scores were analyzed.

Results

Overall, 65 subjects (mean age 72 years ±6, 35 men) were included. White-matter regions with significant (P < 0.05) group differences were found for QSM+ (HC > PD-MCI, PD-CN > PD-MCI) and QSM- (PD-MCI > HC, PD-MCI > PD-CN). In the atlas-based analyses, PD-MCI exhibited lower QSM + values (vs. HC; P = 0.002, vs. PD-CN; P = 0.001), and higher QSM-values (vs. HC; P = 0.02, vs. PD-CN; P = 0.03) in the genu of corpus callosum (gCC). QSM+ and QSM-showed significant positive and negative correlations with MoCA (P < 0.05). In the gCC, partial correlation analyses revealed a positive correlation between QSM+ and MoCA (R = 0.458, P < 0.001) and a negative correlation between QSM- and MoCA (R = −0.316, P = 0.01).

Conclusion

QSM utilizing susceptibility separation is valuable for assessing white matter in PD patients, where nerve fiber loss potentially influences cognitive function.