32 channel radiofrequency receive head coil

The UK Biobank conducted MRI across sites at Reading, Newcastle, and Cheadle Manchester (22 ). Scanning was performed using a Siemens Skyra 3T scanner running on VD13A SP4 software with a Siemens 32-channel radiofrequency receive head coil (22 ) (Supplemental Text S3). Scanning was conducted from the top of the head to the neck using a 256-cm superior–inferior field of view (22 ). The protocol consisted of sagittal T1-weighted images [1 × 1 × 1 mm resolution, time of repetition (TR) = 2000 ms, time of echo (TE) = 2 ms] and T2 fluid-attenuated inversion recovery (FLAIR) images with fat saturation (1.05 × 1 × 1 mm resolution, TR = 5000 ms, TE = 395 ms) (26 ). All acquired images were preprocessed and then checked for quality (Supplemental Text S3) (22 ). Although total brain, gray matter, white matter, and hippocampal volumes were extracted from processed T1 images only, white matter hyperintensities were identified from both processed T2 FLAIR images and T1-weighted images (22 ). All brain volumes were calculated using FreeSurfer software and normalized for the head size using a T1-based head sizing scaling factor (scaled brain volume = brain volume * head size scaling factor) (22 ).

For the current analyses, we utilized varieties of brain imaging phenotypes in the UK Biobank as outcomes, which were acquired from quality-controlling T1-weighted imaging. The brain structure images were scanned by a standard Siemens Skyra 3T running VD13A SP4, with a standard Siemens 32-channel radio frequency receive head coil. During MRI pre-processing, brain structure was segmented into grey matter, white matter, and cerebrospinal fluid (CSF), each of which were anatomically registered and ultimately normalized. For the imaging data, image-derived phenotypes were generated via FMRIB Software Library version 5.0.10 and FreeSurfer version 6.0, which can represent and quantify the structure and functions of the brain. All selected imaging data were presented in cubic millimeters (mm³), encompassing the volumes of 7 subcortical and 33 cortical phenotypes. These data were extracted from FreeSurfer desikan white (category 192) and subcortical volumes (FIRST) (category 1102), respectively. The data fields corresponding to each brain grey matter phenotype can be found in the online showcase of the UK Biobank accessed on 20 July 2023 (https://biobank.ctsu.ox.ac.uk/).

MRI imaging protocols were designed by the UK Biobank Imaging Working Group (http://www.ukbiobank.ac.uk/expert-working-groups). Details regarding image acquisition and processing can be assessed in the previously published article [14 (link)]. Briefly, MRI data were acquired in a single Siemens Skyra 3T scanner with a 32-channel radiofrequency receive head coil located at the recruitment center at Stockport. The T1-weighted scans were obtained using three-dimensional magnetization-prepared rapid gradient-echo (resolution 1 mm³ isotropic voxels) and analyzed with the Functional Magnetic Resonance Imaging of the Brain Software Library (http://fsl.fmrib.ox.ac.uk/fsl). The volumes of the whole hippocampus and its subfield regions (i.e., CA1, CA3, CA4, fimbria, granule layer-molecular layer-dentate gyrus boundary (GC-ML-DG), hippocampal-amygdaloid transition area (HATA), hippocampal tail, hippocampal fissure, paralaminar nucleus, subiculum, and presubiculum) were subsequently generated (N = 23,714). Fractional anisotropy (FA) values of 27 white matter tracts were imaged using diffusion tensor imaging (DTI), where higher FA values suggest better brain integrity. Outlier data points, defined as being further than ± 4 SD from the mean, were excluded (< 1% of values).