Page 66 - PIECES OF THE PUZZLE Eline Vissia

CHAPTER 3
S3.1 Detailed image acquisition and analysis methods
Image acquisition
Identifying individuals with posttraumatic stress disorder (PTSD), with or without dissociative identity disorder (DID), willing and able to participate in a neuroimaging study is known to provide a great challenge. As minimizing travel time increases the likelihood of participation, participants were scanned in the closest of two (University Medical Center Groningen (UMCG) and of the Amsterdam Medical Center (AMC)) 3T MR scanners (Philips Medical Systems, Best, NL) in The Netherlands. Prior to starting, a reproducibility study was conducted that resulted in an optimized structural MRI protocol with a high contrast-to-noise ratio (important for manual and automated segmentation procedures) and a high reproducibility between the two centers (Chalavi et al. 2012). At both centers T1-weighted anatomical MR scans were acquired (MPRAGE, TR=9.95ms, TE=5.6ms, flip-angle=8o, 1x1x1mm voxels, number of slices=160, total scan-time=10m14s). All-PTSD patients and their matched HC were scanned interleaved within a short time interval to avoid an interaction between group and time dependent scanner fluctuations. The samples were balanced over the two centers: twenty All-PTSD patients (ten PTSD-DID, ten PTSD-only) and nineteen HC were scanned at UMCG. Two structural MRI scans were collected from each subject whenever possible (fifteen PTSD-DID patients and fourteen HC). Where both scans were artifact-free, the first scan was used. Four HC subjects were excluded due to the presence of (motion) artifacts in the MRI scan. A total of 17 PTSD-DID, 16 PTSD-only and 28 HC subjects were included in the demographic and morphological analyses.
Image analysis
Manual measures of global volume and shape analysis of the hippocampus
Non-cortical tissue was removed from the MR images using the Brain Extraction Tool (BET) (Smith 2002) and head alignment was standardized by rigidly aligning the individual MR images with the average brain template (ICBM452) using FSL-FLIRT. The hippocampi were manually traced using MultiTracer (Woods 2003) by a single rater (SC), who was blind to all demographical and clinical variables and was trained by an expert (JHC) in this field and obtained
128