ALFF and FC Analysis

ALFF calculation was carried out using the DPABI_V4.3 software. All voxels were converted from the time domain to frequency domain through the use of the Fast Fourier Transform (Luo et al., 2020). The ALFF value of each voxel was obtained by averaging the square root of the power spectrum in the range of 0.01–0.10 Hz. Finally, the ALFF value of whole brain voxels was divided by mean ALFF value of all voxels in order to obtain a standardized ALFF map.

Regions that had significant ALFF differences between before and after surgery in PD patients were identified as regions of interest (ROIs). A seed-based FC analysis was carried out through the use of Resting-State f-MRI Data Analysis Toolkit (REST_V1.8, Beijing, China)³ (Song et al., 2011). The average time series of each ROIs was extracted, and correlation between the ROIs and the time series of each voxel in the whole brain was calculated in order to obtain FC maps. Finally, Fisher’s z transformation (Long et al., 2016) helped normalize all FC maps in order to improve the normality of the data distribution.