Voxel-based Morphometry of Alzheimer's Disease Using a Localizer Image: A Comparative Study with Magnetization Prepared Rapid Acquisition with Gradient Echo

Shohei Inui; Daita Kaneda; Keita Sakurai; Yuto Uchida; Osamu Abe; Yoshio Hashizume

doi:10.2463/mrms.mp.2023-0093

Voxel-based Morphometry of Alzheimer's Disease Using a Localizer Image: A Comparative Study with Magnetization Prepared Rapid Acquisition with Gradient Echo

Magn Reson Med Sci. 2024 Jan 16. doi: 10.2463/mrms.mp.2023-0093. Online ahead of print.

Authors

Shohei Inui¹, Daita Kaneda², Keita Sakurai³, Yuto Uchida⁴, Osamu Abe¹, Yoshio Hashizume²

Affiliations

¹ Department of Radiology, The University of Tokyo, Tokyo, Japan.
² Choju Medical Institute, Fukushimura Hospital, Toyohashi, Aichi, Japan.
³ Department of Radiology, National Center for Geriatrics and Gerontology, Obu, Aichi, Japan.
⁴ Department of Neurology, Nagoya City University Graduate School of Medical Sciences, Nagoya, Aichi, Japan.

PMID: 38233191
DOI: 10.2463/mrms.mp.2023-0093

Abstract

Purpose: Magnetization prepared rapid acquisition with gradient echo (MPRAGE) sequence is a gold-standard technique for voxel-based morphometry (VBM) because of high spatial resolution and excellent tissue contrast, especially between gray matter (GM) and white matter (WM). Despite its benefits, MPRAGE exhibits distinct challenge for VBM in some patients with neurological disease because of long scan time and motion artifacts. Speedily acquired localizer images may alleviate this problem. This study aimed to evaluate the feasibility of VBM using 3D Fast Low Angle Shot image captured for localizer (L3DFLASH).

Methods: Consecutive 13 patients with pathologically confirmed Alzheimer's disease (AD) (82 ± 9 years) and 21 healthy controls (HC) (79 ± 4 years) were included in this study. Whole-brain L3DFLASH and MPRAGE were captured and preprocessed using the Computational Anatomy Toolbox 12 (CAT12). Agreement with MPRAGE was evaluated for L3DFLASH using regional normalized volume for segmented brain areas. In addition to brain volume difference on VBM and Bland-Altman analysis, atrophic pattern of AD on VBM was evaluated using L3DFLASH and MPRAGE.

Results: Acquisition time was 18 s for L3DFLASH and 288 s for MPRAGE. There was a slight systematic difference in all regional normalized volumes from L3DFLASH and MPRAGE. For the whole cohort, GM volume measured from MPRAGE was greater than that from L3DFLASH in most of the region on VBM. When AD and HC were compared, AD-related atrophic pattern was demonstrated in both L3DFLASH and MPRAGE on VBM, although the difference was noted in significant clusters between them.

Conclusion: Although systematic difference was noted in regional brain volume measured from L3DFLASH and MPRAGE, AD-related atrophic pattern was preserved in L3DFLASH on VBM. VBM, using speedily acquired localizer image, may provide limited but useful information for evaluating brain atrophy.

Keywords: 3D T1-weighted image; Alzheimer’s disease; fast Low Angle Shot image; magnetization prepared rapid acquisition with gradient echo; voxel-based morphometry.