35 Using a new technique of ASL | Trigonocephaly MRI Acquisition All MRI data were acquired using a 1.5 Tesla MR Unit (General Electric Healthcare, Milwaukee, WI, United States), and the imaging protocol included a three-dimensional spoiled gradient T1w MR sequence. Imaging parameters included slice thickness of 2 mm, no slice gap, a field of view 22.4 cm, matrix size 224 x 224, in plane resolution 1 mm, echo time 3.1 ms and repetition time 9.9 ms. The pseudo-continuous ASL sequences with the following imaging parameters: 3D fast spin-echo spiral readout with a stack of 8 spirals and 3 averages, repetition time 4,604 ms, echo time 10.7 ms, voxel size 3.75 mm x 3.75 mm x 4.0 mm, field of view 24.0 cm x 24.0 cm, labeling duration 1,450 ms, post-labeling delay 1,025 ms, and background suppression. This protocol was identical in both trigonocephaly patients and controls. Both groups underwent deep sedation or sevoflurane-induced anesthesia during the MRI procedure. Image Processing Data processing and analysis were performed with ExploreASL, a Matlab-based toolbox (MathWorks, MA, United States) developed to facilitate quality control and analyses for single or multicenter ASL studies.15-17 This toolbox is based on Statistical Parametric Mapping (SPM) 12 (Wellcome Trust Centre for Neuroimaging, University College London, United Kingdom).18 The preprocessing included the default 3D ASL processing of ExploreASL except for the ASL-T1w registration part, which includes the quantification of the M0 and CBF image.15 Afterwards, we compared four approaches: (1) normalization via segmentation and spatial normalization of the T1w images (regT1); a direct registration of ASL to MNI using either (2) rigid body (regASLrigid), (3) affine (regASLaffine), or (4) a nonlinear Direct Cosine Transform (DCT) transformation (regASLdct). For regT1, the T1w image was segmented – to GM, WM, and cerebrospinal fluid – and registered to MNI space with ExploreASL default settings except for using SPM12 with the NITRC 2.3 brain template of 1-year-olds. The 1-year-old template was chosen for all subjects since this was the closest match for the majority of our subjects.19 The M0 and T1w images were rigidly aligned and applied to ASL to align it with T1w. The spatial normalization of ASL for regT1 was then obtained as the joint transformation of ASL to T1w and T1w to MNI space. For the regASLrigid, regASLaffine, regASLdct, a pseudo-CBF image was constructed in the MNI space using the GM and WM probability maps from the NITRC template resampled to 1.5 mm x 1.5 mm x 1.5 mm voxels, multiplying them by 60 and 20 ml/ min/100 g, respectively, and finally smoothing them to the ASL resolution. The native ASL CBF image was then registered with the pseudo-CBF image using the respective 2
RkJQdWJsaXNoZXIy MTk4NDMw