New version page

Reliability of MRI-derived

This preview shows page 1-2-3 out of 10 pages.

View Full Document
View Full Document

End of preview. Want to read all 10 pages?

Upload your study docs or become a GradeBuddy member to access this document.

View Full Document
Unformatted text preview:

Reliability of MRI-derived cortical and subcortical morphometric measures: Effects of pulse seq.....IntroductionMethodsParticipantsApparatus and parametersData processingStatistical analysisResultsTest–retest reliability of traditional T1 sequenceEffects of voxel sizeEffects of parallel imaging accelerationEffects of multiecho sequenceComprehensive statistical analysisDiscussionConclusionsDisclosure statementAcknowledgmentsReferencesReliability of MRI-derived cortical and subcortical morphometric measures: Effects ofpulse sequence, voxel geometry, and parallel imagingJ.S. Wonderlicka,b,⁎, D.A. Zieglera,b, P. Hosseini-Varnamkhastia,b, J.J. Locasciod, A. Bakkourc,e,A. van der Kouwec,f, C. Triantafylloub, S. Corkina,b,c, B.C. Dickersonc,daDepartment of Brain and Cognitive Sciences, Massachusetts Institute of Technology, USAbAthinoula A. Martinos Imaging Center at the McGovern Institute, Massachusetts Institute of Technology, USAcMGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging, USAdDepartments of Neurology, Massachusetts General Hospital, Harvard Medical School, USAeDepartments of Psychiatry, Massachusetts General Hospital, Harvard Medical School, USAfDepartments of Radiology, Massachusetts General Hospital, Harvard Medical School, USAabstractarticle infoArticle history:Received 18 July 2008Revised 24 September 2008Accepted 21 October 2008Available online 7 November 2008Advances in magnetic resonance imaging (MRI) have contributed greatly to the study of neurodegenerativeprocesses, psychiatric disorders, and normal human development, but the effect of such improvements onthe reliability of downstream morphometric measures has not been extensively studied. We examined howMRI-derived neurostructural measures are affected by three technolog ical advancements: parallelacceleration, increased spatial resolution, and the use of a high bandwidth multiecho sequence. Test–retestdata were collected from 11 healthy participants during 2 imaging sessions occurring approximately 2 weeksapart. We acquired 4 T1-weighted MP-RAGE sequences during each session: a non-accelerated anisotropicsequence (MPR), a non-accelerated isotropic sequence (ISO), an accelerated isotropic sequence (ISH), and anaccelerated isotropic high bandwidth multiecho sequence (MEM). Cortical thickness and volumetricmeasures were computed for each sequence to assess test–retest reliability and measurement bias. Reliabilitywas extremely high for most measures and similar across imaging parameters. Significant measurement biaswas observed, however, between MPR and all isotropic sequences for all cortical regions and somesubcortical structures. These results suggest that these improvements in MRI acquisition technology do notcompromise data reproducibility, but that consistency should be maintained in choosing imaging parametersfor structural MRI studies.© 2008 Elsevier Inc. All rights reserved.IntroductionRapid improvements in magnetic resonance imaging (MRI)technology continue to provide new opportunities to deepen ourunderstanding of brain structure and function in health and disease.Technologic developments include methods for accelerating theacquisition of MRI data (Griswold et al., 2002; Katscher et al., 2003;McDougall and Wright, 2005; Pruessmann et al., 1999; Tsao et al.,2003), improving the spatial resolution of MRI data (Augustinack et al.,2005), and reducing spatial distortions within and between types ofsequences (Fischl et al., 2004a; van der Kouwe et al., 2008). Althoughthese new techniques may provide theoretical advantages for studiesof patients with neurologic or psychiatric disorders, few studies haveexamined the impact of such techniques on the quantitative measuresthus derived.Because the growing number of tools used to perform computa-tional analysis on MRI data rely upon subtle differences in imagesignal intensity and tissue contrast to determine neuroanatomicalboundaries (Fischl and Dale, 2000), slight differences in imagingmethods could have a considerable impact on the reproducibility ofmorphometric measures. In addition, reliability may differ acrossbrain structures due to variability in tissue contrast profiles anddivergent modeling algorithms (e.g., cortical surface-based or voxel-based segmentation methods). The goal of this study was to assess theimpact of novel MRI technologic parameters on reliability, using avariety of morphometric measures as outcome variables.The present study examined the effect of three elements of MRIdata acquisition technology on the reliability of neuroanatomicalmeasures: geometric reduction in voxel size (higher resolution),acceleration through parallel acquisition (Carlson and Minemura,1993), and use of a high bandwidth, multiecho T1-weighted sequence(Fischl et al., 2004a; van der Kouwe et al., 2008). First, parallelacquisition (i.e., the use of phased array head coils to acquire data frommultiple points in space simultaneously) can reduce scanning time,with relatively small decreases in the signal-to-noise ratio (SNR)NeuroImage 44 (2009) 1324–1333⁎ Corresponding author. Department of Brain and Cognitive Sciences, MassachusettsInstitute of Technology, Cambridge, MA 02139, USA.E-mail address: [email protected] (J.S. Wonderlick).1053-8119/$ – see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.neuroimage.2008.10.037Contents lists available at ScienceDirectNeuroImagejournal homepage: www.elsevier.com/locate/ynimg(Griswold et al., 2002; Roemer et al., 1990). The use of parallelacquisition has become increasingly common with the proliferation ofacceleration-capable MRI equipment, but its effect on reliability hasnot been thoroughly studied, parti cula rly with respect to thequantitative morphometric measures of interest in clinical investiga-tion. Second, voxel geometry is often manipulated to gain advantagesin scanning time. Larger voxel volumes require less scanning time andprovide increased image SNR over smaller volumes (Edelstein et al.,1986), but they also increase partial volume effects whereby the signalmeasured from a single voxel may consist of contributions from morethan one type of tissue. Advances in head coil and parallel acquisitiontechnology have led to the capacity to collect data of higher spatialresolution with minimal reductions in SNR, but the impact of voxelsize on morphometric measures has not been quantified system-atically. Third, high bandwidth multiecho sequences have emergedthat promise decreased


Loading Unlocking...
Login

Join to view Reliability of MRI-derived and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Reliability of MRI-derived and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?