Image preprocessing analysis

All collected image data were processed in a postprocessing workstation (model No. uWS-MR-R004; Union-Imaging Healthcare). Dr. Cao Q with 12 years of medical imaging diagnosis experience, and Dr. Li Y, with 20 years of medical imaging diagnosis experience, evaluated the images, agreed on the principle of blindness, measured the relevant image indicators, and averaged the 2 measurements. The physicians evaluated the swallow tail sign based on the SWI chart and determined whether it had disappeared (Figure 1A) according to the following scoring scheme: 2 points, both sides of the swallow tail sign disappeared; 1 point, 1 side disappeared; and 0 point, no sign of disappearance of the swallow tail sign on either side. The layer-by-layer distribution of melanin in bilateral substantia nigra neurons on NM-MRI was recorded, and the volume and average signal value were calculated (Figure 1B).

Display of the substantia nigra and locus coeruleus regions and corresponding indicator observations based on multiple MRI sequences. (A) Swallow tail sign displayed on SWI (the area is indicated by the red arrows). (B) The distribution of melanin in neurons of bilateral substantia nigra on NM-MRI (green outlined area). (C) The high signal intensity in bilateral locus coeruleus nuclei on NM-MRI (the red arrows indicate the high signal area). (D) The bilateral substantia nigra areas were outlined, and T2* and R2* values were obtained based on T2*mapping (dark blue area outlined by the wireframe). VOI, volume of interest; SD, standard deviation; SP, scale pixel; MRI, magnetic resonance imaging; SWI, susceptibility-weighted imaging; NM-MRI, neuromelanin-sensitive magnetic resonance imaging.

Before images were collected, the appropriate window width and position were fixed to reduce visual and signal differences. We only delineated the high signal areas of the substantia nigra, without drawing the areas of attenuation. Images were collected without any gaps. The volume of high signal intensity areas in the distribution of melanin neurons was calculated layer by layer based on voxels, the volume of distribution of melanin in bilateral substantia nigra, and bilateral average value and smaller value; meanwhile, the measured signal was determined as the average value of the sketched area. Given that the locus coeruleus is small and the signal intensity of the locus coeruleus is higher than that of the surrounding tissues, the maximum signal value of its voxel level was first obtained by measuring the signals of the bilateral locus coeruleus areas. Then, the signals of the bilateral locus coeruleus areas were compared in each patient, and their smaller and average values were recorded (Figure 1C).

For T2*mapping data processing, some cases showed heavy skull-base magnetic susceptibility artifacts (3 cases in the PD group). In 4 cases in the PD group, multiple time to echo (TE) data points caused by skull-base plane motion could not be processed with motion correction, and thus these patients were not included. Ultimately, in 33 patients in the PD group and 40 patients in the healthy control group, bilateral substantia nigra areas were drawn while T2* and R2* values were calculated (Figure 1D) based on T2*mapping to quantitatively analyze iron deposition in the substantia nigra (14).