Meeting Banner
Abstract #0781

Across 1.5T and 3.0T MR Field-Strengths: Comparison of Regional Brain Volumes Derived from Atlas-Based Parcellation Data

Adolf Pfefferbaum1, 2, Torsten Rohlfing1, Margaret J. Rosenbloom1, 2, Edith V. Sullivan2

1Neuroscience Program, SRI International, Menlo Park, CA, United States; 2Psychiatry & Behavioral Sciences, Stanford University School of Medicine, Stanford, CA, United States

To determine whether structural T1-weighted data acquired across MR field strengths could be merged, archival SPGR data acquired at 1.5T and 3.0T within 3 weeks in 114 adults were compared using intraclass correlation (ICC) computed on 24 cortical, subcortical, and CSF-filled volumes derived from an automated, unsupervised atlas registration/parcellation method. Application of regression correction functions for each region yielded significant ICC increases (mean improvement=.087; post-correction ICC range=.599 in globus pallidus to .998 in lateral ventricles). Thus, selected T1-weighted regional anatomical brain data can be reliably combined across 1.5T and 3.0T field strengths with the application of an appropriate correction procedure.

Abstract PDF

Keywords

ability accordingly adjusted adults aging agreement analyses anatomical application applied archival argues array assessed atlas automated behavioral biometrics body brain broadening channel clinical coil collected combine comparability computed coronal correction correlation correspondence cortex cortical declined dependent derived designed determination determine determined development disease dual enhancements ensure examined expand failed fair fast field forthcoming function functions general greater head healthy identity improve improved improvement independent indicated insubstantially integrity international intraclass inversion investigations iron lack larger lateral longitudinal lowest matrix measured median merged merging moderate pair park perfect possibilities post potential prepared procedure progression projects proposition psychiatry quadrature quantification rather reach reasons recovery regional registration regression reliability reliably remained replacements representing resolution rich sample samples scanned scatter school selected significantly skip slices spaces spontaneous strength strengths structural studies substantial system thick tissue turn twin types typically unchanged uncorrected unsupervised utility variety various ventricles volume volumes weeks whereby whether whole women worse years yielded