1.5 tesla scanner

MRI images were acquired in a General Electric 1.5 Tesla scanner (General Electric Medical Systems, Miwaukee, WI, USA) with a homogeneous birdcage head coil. The participants lay supine with their heads snugly fixed by a belt and pads were used to minimize head motion. Resting-state functional images (T2^* weighted images) were obtained by gradient-recalled echo-planar imaging (GRE-EPI) sequence: number of slices = 30, thickness = 4.0 mm, gap = 0 mm, in-plane resolution = 3.75 × 3.75 mm², TR = 3,000 ms, TE = 40 ms, flip angle = 90°, acquisition matrix = 64 × 64, FOV = 240 × 240 mm. This acquisition sequence generated 142 volumes in 7 min and 6 s. In addition, three-dimensional T1-weighted axial images covering the whole brain were obtained using a spoiled gradient echo (SPGR) sequence: TR = 9.9 ms, TE = 2.1 ms, thickness = 2.0 mm, gap = 0 mm, flip angle = 15°, FOV = 240 × 240 mm, acquisition matrix = 256 × 192. Participants were instructed to keep their eyes closed, bodies aplanatic and not to think systematically or fall asleep during the scanning.

CMR imaging was performed with a 1.5 Tesla scanner (GE). Delayed-enhancement images for detection of fibrosis were obtained 10 minutes after intravenous administration of gadolinium–DTPA (0.2 mmol/kg) [25] (link). Total of 6 short axis images were analyzed to qualify and quantify LGE. LGE was defined by a signal intensity of > 6 standard deviations above normal myocardium [26] (link). The LGE area was measured by manual planimetry using ImageJ software (National Institutes of Health, Bethesda, Maryland) in each short-axis image [27] (link). The area of the left ventricle was determined by manually tracing epicardial and endocardial borders in each short-axis image. The percentage area of LGE was then calculated by dividing the sum of the LGE areas by that of the total left ventricular area [27] (link). Data were presented both as percentage of the left ventricular area and as a binomial variable; positive/negative (Figures 1E and F). CMR data were analyzed by 2 experienced investigators (E.M., Y.N.) with consensus without knowledge of the clinical information of the patients.

Subjects were scanned on a GE 1.5 Tesla scanner in the Department of Radiology MR Research Center. T1-weighted, T2-weighted, and axial proton density images were obtained, as previously described [21 (link)]. Regions of interest were drawn using BRAINS2 [39 (link)], a program that uses a semiautomated segmentation approach to provide reliable and valid structural volumetric measurements. Two raters who were blind to subject identity and risk group status traced the volumes of the cerebellar lobes and intracranial volume (ICV) according to the guidelines established by Pierson et al. [40 (link)]. The anterior, inferior posterior, and superior posterior lobes were traced along with the corpus medullare. The corpus medullare consists of the central white matter and the output nuclei of the cerebellum. Deep within the corpus medullare, these small gray matter nuclei provide much of the output of the cerebellum (Fig. 1).Fig. 1

Regions of interest traces are shown in the sagittal plane. Measured ROIs included the anterior, superior posterior, and inferior posterior lobes and the corpus medullare. Figure from Pierson et al. [40 (link)]. The yellow Xs (oval cluster) represent the outline of the corpus medullare with the remaining yellow Xs indicating the horizontal fissure

Brain MRI data were acquired with a 1.5-Tesla scanner (GE Healthcare, Milwaukee, Wisconsin) using an eight-channel head coil during a 30-min brain imaging protocol that was previously described in detail (Ikram et al. 2011 (link); Jones et al. 1999 (link)). Two trained technicians performed all examinations in a standardized way. This protocol included high-resolution axial fluid-attenuated inversion recovery (FLAIR), T1-weighted, and T2-weighted sequences. The T1-weighted image was used for amygdala segmentation and consisted of a 3D spoiled gradient-recalled echo (SPGR) scan, with voxel volume of 0.49 × 0.49 × 0.80 mm³. Automatic segmentation of subcortical brain structures, including the amygdala and hippocampus left and right, was performed on T1-weighted images using Freesurfer software (version 4.5.0) (FreeSurfer 2013 ). This rendered volumetric measures of gray matter (in mm³). Exact processing details are described elsewhere (Reuter et al. 2012 (link); Desikan et al. 2006 (link); Fischl et al. 2002 (link)).