Objective: To determine whether voxel-based morphometry of the cerebrospinal fluid (CSF) space can accurately differentiate iNPH from other related neurological disorders.
Methods: Nineteen patients with surgically proven iNPH and 24 patients with Alzheimer's disease (AD), 18 patients with Parkinson's disease (PD), and 14 healthy elderlies were examined. Regions of interest (ROIs) in the ventricular/sylvian (VS) and high convexity/midline (HCM) areas were used in VBM, and differences in CSF volume within these ROIs among the 4 groups were examined.
Results: The relative CSF volume in the VS-ROI in the iNPH group (.64) was significantly higher than that in the AD, PD, and healthy control groups (.26, .29, and .17, respectively; P < .001), whereas the relative CSF volume in the HCM-ROI was significantly reduced in the iNPH (.10) than in the other (.17, .16, and .16, respectively) groups (P < .001). Furthermore, the ratio of the 2 areas (VS/HCM) was markedly higher in the iNPH (6.87) than in the other (1.65, 1.74, .97, respectively) groups (P < .001), with sensitivity and specificity of .89 and 1.00, respectively.
Conclusions: The VBM-based CSF space analysis can detect disproportionate changes in CSF space and differentiate iNPH patients from those with AD or PD and healthy elderlies accurately.
Keywords: Alzheimer's disease; Idiopathic normal pressure hydrocephalus; Parkinson's disease; cerebrospinal fluid space; voxel-based morphometry.
Copyright © 2013 by the American Society of Neuroimaging.