Premium
IC‐P‐127: Hippocampal microstructure and memory performance in Alzheimer's disease: A 7‐Tesla MRI study
Author(s) -
Kerchner Geoffrey,
Deutsch Gayle,
Zeineh Michael,
Rutt Brian
Publication year - 2011
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.2011.05.092
Subject(s) - hippocampus , hippocampal formation , audiology , parahippocampal gyrus , psychology , verbal memory , episodic memory , alzheimer's disease , entorhinal cortex , neuropsychology , neuroscience , medicine , temporal lobe , pathology , cognition , disease , epilepsy
stratum pyramidale (SP) and stratum radiatum / lacunosum-moleculare (SRLM). In prior work (Kerchner et al., 2010, Neurology 75:13811387), atrophy of the SRLM was evident in patients with mild Alzheimer’s disease (AD) relative to age-matched normal controls. In that study, manual measurements were made of stratal width. We endeavored to create a semi-automated method for measuring stratal width, and to test the resulting method on the prior data set. Methods: On raw oblique coronal 7-Tesla images oriented perpendicular to the hippocampal body, a line was drawn by the user along the center of the SRLM, and another line was drawn along the center of the SP. We wrote a MATLAB script that examined each user-specified line as follows: Normal vectors were drawn at multiple points along each line, and intensity values along those vectors were plotted against the relative distance from the user-specified line. These plots were averaged to yield a curve with a negative peak (for SRLM) or positive peak (for SP). The distance between the positive and negative peaks of the first derivative of that curve was used as the measure of stratal width. This procedure was applied to the T2*weighted 7-Tesla GRE images in the prior publication. Four subjects were excluded, because of incompatibilities of their original data with the new procedure. Results: The semi-automated measurements of SRLM and SP thickness worked robustly, correlating significantly with the manual measurements (SP: r 1⁄4 0.71, p < 0.001; SRLM: r 1⁄4 0.45, p 1⁄4 0.001). As in the prior study, the SRLM was significantly thinner among patients with AD (590 6 80 micro-m, using the new method; n 1⁄4 13) versus normal controls (670 6 110 micro-m; n 1⁄4 13; p < 0.05). Additionally, SP was significantly thinner among AD patients (1.2 6 0.4 mm) versus normal controls (1.5 6 0.2 mm; p < 0.05). Conclusions: In sum, we present a semi-automated method for extracting SP and SRLM thicknesses from 7-Tesla hippocampal cross-sectional images and demonstrate its use in revealing microanatomical features of AD.