DefinePK

Abstract

Hippocampal volume change over time, as evaluated by MRI, offers excellent potential as a marker for Alzheimer's “disease. In this study, we consider, using publicly available Statistical Parametric Mapping (SPM) software, a fully automated and computationally efficient processing pipeline for atlas-based hippocampal volumetry in 75 amnestic subjects with mild cognitive impairment (MCI) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Subjects were split into two groups: MCI stable and MCI to probable Alzheimer's disease (AD) converters, based on follow-up diagnoses at 0, 6, and 12 months. From the baseline T1-weighted MRI, the hippocampal grey matter volume (HGMV) was measured and adjusted for age and total intracranial volume. The average processing time per subject on a typical PC was less than 4 minutes. To identify MCI to likely AD converters, the area under the receiver operator characteristic curves of the corrected hippocampal grey matter volume right (HGMVR) and hippocampal grey matter volume left (HGMVL) were determined. Using one-way ANOVA data on the MMSE score, HGMVR, and HGMVL, a ROC CURVE is computed to compare the control group versus MCI and the control group versus AD.