방사선의학 웹진

바로가기  >

Abstract
OBJECTIVE:
Digital brain template and atlas designed for a specific group provide advantages for the analysis and interpretation of neuroimaging data, but require a significant workload for development. We developed a simple method to create customized brain atlas for diffeomorphic anatomical registration through exponentiated lie algebra (DARTEL) tool using FreeSurfer-generated volume-of-interest (FSVOI) images and validated.

METHODS:
18F-florbetaben positron emission tomography (PET) and magnetic resonance (MR) imaging data were obtained from 248 participants of Alzheimer's disease spectrum (from cognitively normal to Alzheimer's disease dementia). To create a customized atlas, MR images of 84 amyloid-negative controls were first processed with FreeSurfer to obtain individual FSVOI and with DARTEL tool to create DARTEL template. Individual FSVOI images were spatially normalized, and each voxel was then labelled with a VOI label with maximum probability. Using these template and atlas, all images were normalized, and the regional standardized uptake value ratios (SUVR) were measured.

RESULTS:
18F-florbetaben SUVR values measured with customized atlas showed excellent one-to-one correlation with SUVR measured with individual FSVOI in all regions, and thereby showed almost identical between-group comparison results and outperformed the classic methods.

CONCLUSIONS:
Customized FreeSurfer-based brain atlas for DARTEL tool is easy to create and useful for the analysis of PET and MR images with high adaptability and reliability for broad research purposes.

Author information

Baek MS1, Cho H1, Ryu YH2, Lyoo CH3.
1
Department of Neurology, Gangnam Severance Hospital, Yonsei University College of Medicine, 20 Eonjuro 63-gil, Gangnam-gu, Seoul, Republic of Korea.
2
Department of Nuclear Medicine, Gangnam Severance Hospital, Yonsei University College of Medicine, 211 Eonjuro, Gangnam-gu, Seoul, Republic of Korea. ryuyh@yuhs.ac.
3
Department of Neurology, Gangnam Severance Hospital, Yonsei University College of Medicine, 20 Eonjuro 63-gil, Gangnam-gu, Seoul, Republic of Korea. lyoochel@yuhs.ac.