Ethical approval and subjects description
This study was approved by the ethics committee at Qilu Hospital of Shandong University (Qingdao) and performed in accordance with the Helsinki Declaration of the world medical association. Participants were consecutively recruited at Movement Disorder Center of Neurology between Jan 2015 and July 2020 whose brain scanning including SWI sequence, as part of our previous of study. Clinical diagnoses of MSA-P and PD were made according to established criteria[8,9] by two clinicians (CP Zhao and JY Zhang) with movement disorders professional experience for more than 10 years. The disease duration was defined from the occurrence of the motor symptoms. We aimed to figure out the image difference of MSA-P and PD in relative early stage when it is difficult to differentiate clinically, we grouped patients with cut-off point of 2 years’ disease duration at the time of performing MR according to previous study. In the light of the consensus guidelines in 2008, our MSA-P patients were with clinical probable MSA-P and regularly followed in our clinics. As a control group (CG), we selected the previously retrieved patients from Picture and Communication System (PACS) with same methods and exclusion criteria of our previous study except that we also included the subjects scanned on 1.5T MR. At last, 17 patients with MSA-P (14 on 3.0T, 3 on 1.5T MR), 17 age-sex and MR model matched PD with disease duration no more than 2 years and 17 CG were enrolled in this study.
Scanning model and parameters
The axial scans were set parallel to the intercommissural line. Scanning parameters on 3.0T MR scanner (Ingenia scanner, Philips Medical Systems, Netherlands) were as follows: slice thickness = 2 mm; TR = 20 ms; TE = 27 ms; flip angle 15°; FOV=220 mm; number of signal acquisitions 1; and matrix size 284 × 230. Scanning parameters on 1.5T MR (Achiva, scanner, Philips Medical Systems, Netherlands) were as follows: slice thickness = 2mm; TE=10 ms; TR= shortest; flip angle 10°; FOV=230 mm; number of signal acquisitions 1; and voxel size = 1mm× 1mm × 0.6mm .
Region of interest and morphometric index extraction
In the PACS system, the longest horizontal line (LL), the short line (SL), the calculated ratio of SL/LL (SLLr), and the area, the mean signal intensity (SIm_LN) and the standard deviation of the signal intensity (SIsd_LN) of the manually sketched boundary area of the LN were recorded by two experienced radiologists (QG Ren and XM Nan) in the magnitude image axis plane according to our previous study. Those above indexes of both sides were recorded, the uniformity of these two radiologists measurement results were estimated by the mean value of both sides, moreover, the uniformity of the smaller side of SL by these two radiologists were also estimated. The cerebrospinal fluid signal intensity of the 4th ventricle was also measured as SIm_CSF and SIsd_CSF. Then, the SIm_LN of each side was normalized to SIm_CSF with a signal intensity of 200 (nSIm) according to previous study. Then, we calculated the mean measured value of each side by these two radiologists separately. According to our previous experience, the MSA-P patients are characterized by narrowing morphology and the inhomogeneous signal intensity of the posterior region of LN. So we chose the smaller SL side as corrected SL (cSL) and calculated the corrected ratio (cSLLr) by LL with the same side of cSL. as shown in Fig 1. We also chose the bigger SIsd_LN side as corrected cSIsd_LN,. Given all the other above indexes we calculated the mean values of the left and the right for the following statistical analysis.
The Statistical Package for the Social Sciences (SPSS 22.0, Chicago, Illinois) was used for statistical analysis. Continuous variables were expressed as the mean ± SD. One way analysis of variance (ANOVA) with post-hoc multiple comparisons conducted by Least Significant Difference (LSD) was used for groups and subgroups comparison when variables conformed to the normal distribution by Shapiro-Wilk test, otherwise nonparametric test (Kruskal-Wallis or Mann-Whitney U) was used. All statistical significance was defined as P value＜.05. Receiver operating characteristic (ROC) curves were plotted to assess the value of the significant different index in differentiating MSA-P from PD and health controls, in which cutoff values were determined by using Youden's index . Kappa analysis was used to assess the uniformity of the smaller side of SL by the radiologists and Intraclass correlation coefficient (ICC) was to assess the uniformity of the mean values of the left and the right measured data by them.