Marieke van Rosmalen

Prognostic value of quantitative MRI in chronic inflammatory neuropathies 115 7 Table 7.1 MRI parameters Parameter DTI T2 mapping T1 DIXON 3D TSE Acquisition 2D SE-EPI 2D TSE 3D FFE 3D TSE Field of view 240*180*150 mm 3 240*180*52.5 mm 3 288*288-200.25 mm 3 336*336*170 mm 3 Matrix size 96*71 96*96 192*192 224*223 Slice thickness 2.5 mm 2.5 mm - - Voxel size 2.5*2.5*2.5 mm 3 2.5*2.5*2.5 mm 3 0.75*0.75*0.75 mm 3 0.75*0.75*1 mm 3 Echo time 60 ms 7.6 ms 1.186 ms 206 ms Number of echoes - 17 3 - Repetition time 8595 ms 3242 ms 5615 ms 2200 ms Flip angle - - 16 o - Turbo spin echo factor - - - 76 Sensitivity encoding factor 2.5 2.3 2 (AP); 1 (FH) 3 (RL); 1.5 (AP) Fat suppression SPAIR - - SPIR Gradient directions 37 - - - b values (s/mm 3 ) 0, 50, 100, 150, 300, 400, 600 - - - Acquisition time 05:43 minutes 04:45 minutes 01:56 minutes 03:59 minutes Abbreviations: DTI = diffusion tensor imaging; TSE = turbo spin echo; SPAIR = spectral attenuated inversion recovery; SPIR = spectral presaturation with inversion recovery; SE EPI = spin echo-echo planar imaging; FFE = fast field echo; mm = millimeter; ms = milliseconds; AP = anterior/posterior; FH = foot/head; RL = right/left. Data processing We processed MRI data semi-automatically using a two-step custom-build processing pipeline, as described in detail previously. 25 In short, the processing pipeline provided whole volume fiber tractography of the brachial plexus and reconstruction of cervical nerve roots C5, C6 and C7. This allowed us to extract diffusion parameters (i.e. fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD)), T2 relaxation times and fat fraction from a standardized site, i.e. over a 1 cm segment next to the ganglion. Statistical analysis We used IBM SPSS Statistics (Version 26, Armonk, NewYork, United States) for statistical analysis. To compare patient characteristics, we used an independent samples t test. The mean difference over time for the quantitative parameters (i.e. diffusion parameters, T2 relaxation times and fat fraction) was estimated between t 0 and t 1 . We used a linear mixed model to correct for the multiple paired measurements per subject and calculated the mean difference with 95% confidence intervals and p values . A p value < 0.05 was considered significant. Correlations between the quantitative parameters and clinical data were analyzed using an independent samples t test for categorical data and using the Pearson correlation coefficient r for continuous data. We considered r ≤ 0.35 as a weak correlation, 0.36 – 0.70 as moderate, 0.70 – 0.89 as high and ≥ 0.90 as a very high correlation. 28