Premium
P2‐450: Hemizygous deletion of synaptojanin1 ameliorates learning and memory impairments in a mouse model of Alzheimer's disease
Author(s) -
McIntire Laura Beth J.,
Berman Diego,
Myaeng Jennifer,
Staniszewski Agniezka,
Arancio Ottavio,
Di Paolo Gilbert,
Kim Tae-Wan
Publication year - 2010
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2010.08.024
Subject(s) - long term potentiation , creb , dendritic spine , synaptic plasticity , hippocampal formation , neuroscience , phosphorylation , alzheimer's disease , synapse , biology , microbiology and biotechnology , endocrinology , medicine , disease , biochemistry , receptor , gene , transcription factor
Background: Cortical grey matter (cGM) thickness change (cGMtc) and ventricular volume (VV) change (VVc) are two measures of regional atrophy that can be used to quantify degenerative changes in AD. To increase precision, we developed longitudinal methods to quantify cGMtc and VVc directly from paired MRI scans. Objective To compare the precision of longitudinal vs. cross-sectional measures of cGMtc and VVc. Methods: As part of the MRI site activation process for two ongoing clinical trials, each participating center scanned a volunteer, removed them from the magnet and then rescanned them using a standardized T1-weighted protocol. Processing of images included “nonbrain”-constrained registration between scan pairs. Cortical thickness was computed as previously described (Chen et al., NeuroImage 2004). cGMtc was calculated from the difference between mean cortical thickness measured over all cortical surface voxels common to both scans, expressed as a percentage of the mean overall cGM thickness. Ventricles were segmented using a manual thresholding technique. VVc was computed by propagating the ventricular mask from the first to the second timepoint, computing the mean orthogonal displacement for each ventricular boundary voxel, calculating the mean displacement over all ventricular voxels and converting this to percentage change using a calibration factor computed from the displacement yielded by a known ventricular volume change (Smith et al. NeuroImage 2002). Both processes were executed symmetrically to ensure unbiased results.Results: Data were obtained from 22 centers using scanners from 3 manufacturers and 2 field strengths. Compared to the truth, which should be 0 (see Figure), the mean (SD) of the absolute cGMtc was 0.34% (0.32%) using the longitudinalmethod, and 1.1% (0.72%) using the difference of cross-sectional measures. The absolute VVc was 0.27% (0.17%) using the longitudinal method, and 1.96% (1.20) using the cross-sectional differences. Conclusions: The mean absolute errors for our longitudinal measures were 1/3 that of the difference of cross-sectional measurements for cGMtc, and less than 1/7 for VVc. These results demonstrate the enhanced precision of our longitudinal measures over repeated cross-sectional measures, and suggest that the additional cost and complexity of phantom correction is not necessary.