Objective  To explore the correlation between MRI microstructure and perfusion parameters and cognitive function in patients with Parkinson disease (PD).

Methods  The clinical data of 62 patients with PD treated in Lianyungang First People's Hospital, Xuzhou Medical University from December 2019 to December 2020 were retrospectively analyzed, including 22 males and 40 females, with an average age of (61.91±8.41) years. Patients were divided into PD-MCI group (n=33) and control group (n=29) in accordance with whether they had mild cognitive impairment (MCI). The Montreal Cognitive Assessment Scale (MoCA) was used to evaluate their cognitive function. The apparent diffusion coefficient (ADC) and fractional anisotropy (FA) of magnetic resonance diffusion tensor imaging (DTI) were used to reflect the microstructure of brain tissue. Changes in relative cerebral blood flow (rCBF) were examined by magnetic resonance arterial spin-labeled imaging. Independent sample t test was used to compare measurement data consistent with normal distribution. Chi-square test was used to compare count data. Pearson's correlation analysis was used to analyze the correlation among ADC, FA, rCBF, and MoCA. The receiver operating characteristic (ROC) curve was used to evaluate the predictive ability of ADC, FA, and rCBF on cognitive function in patients with PD.

Results  The MoCA score of the PD-MCI group (20.79±0.51) was lower than that of the control group (27.99±0.70), and the difference was statistically significant (t=−46.449, P<0.001). The ADC (0.914±0.079, 0.802±0.035, 0.793±0.042, and 0.745±0.041) and rCBF (0.753±0.051, 0.774±0.078, 0.869±0.054, and 0.836±0.063) of the substantia nigra region of interest (ROI), caudate nucleus ROI, thalamic ROI, and hippocampus ROI in the PD-MCI group were higher than those in the control group (0.716±0.068, 0.671±0.044, 0.675±0.034, 0.641±0.047; 0.611±0.048, 0.605±0.081, 0.732±0.059, 0.721±0.046), whereas the FA (0.356±0.047, 0.341±0.022, 0.324±0.039, 0.317±0.027) was lower than that in the control group (0.497±0.068, 0.431±0.046, 0.445±0.051, 0.437±0.043), with statistical significance (t=−13.326–13.044, all P<0.001). The results of Pearson's correlation analysis showed that the ADC and rCBF of the substantia nigra ROI, caudate ROI, thalamus ROI, and hippocampus ROI were negatively correlated with MoCA, whereas FA was positively correlated with MoCA (r=−0.856–0.847, all P<0.001). The ROC curve results showed that the AUCs of ADC, FA, and rCBF in the substantia nigra ROI were 0.979(95%CI: 0.953~1.000), 0.963(95%CI: 0.919~1.000) and 0.977(95%CI: 0.946~1.000), respectively (all P<0.05). The AUC of ADC, FA and rCBF in caudate ROI were 0.993(95%CI: 0.977~1.000), 0.965(95%CI: 0.920~1.000) and 0.964(95%CI: 0.927~1.000), respectively (all P<0.05). The ADC, FA and rCBF AUC of thalamic ROI were 0.997(95%CI: 0.989~1.000), 0.973(95%CI: 0.941~1.000) and 0.975(95%CI: 0.945~1.000), respectively (all P<0.05). The AUC of ADC, FA and rCBF of hippocampal ROI were 0.970(95%CI: 0.933~1.000), 0.99(95%CI: 0.985~1.000) and 0.917(95%CI: 0.845~0.989), respectively (all P<0.05).

Conclusions  MRI microstructure and perfusion parameters are closely related to the cognitive function of patients with PD. They can better predict the cognitive function of patients with PD.