Abstract
Neuropathologically, Alzheimer's Disease (AD) is characterized by amyloid plaques which can be visualized by using the positron emission tomography (PET) tracer [11C]Pittsburgh compound-B ([11 C]PIB). The purpose of this study was to evaluate the agreement in visual assessment of various [11 C]PIB images in a memory clinic patient population. Dynamic [11C]PIB PET scans of 90 minutes duration were obtained in 30 AD patients, 30 patients with non-AD dementia, 30 patients with mild cognitive impairments (MCI), and 30 controls. Parametric non-displaceable binding potential (BPND) images were generated using receptor parametric mapping with fixed outflux rate constant for the reference tissue (RPM2). In addition to parametric BPND images, averaged 60-90 min activity concentration (SUM) images, and standardised uptake value ratio images (SUVr) were made. All images were classified in a binary fashion as either amyloid positive or amyloid negative by two independent physicians who were blinded for diagnosis. Agreement between readers differed between analytical methods, with higher agreement for parametric BPND images (κ = 0.93) than for SUVr (κ = 0.86) and SUM images (κ = 0.78). Concordance of visual assessment between BPND images and both SUM and SUVr images were respectively 94% and 95%, with only disconcordance in the groups other than the AD group. Visual assessment of parametric RPM2 images showed higher inter-rater agreement than semi-quantitative SUVr and SUM images. Concordance between methods was 100% in the AD group, but lower in the other groups. This suggests an advantage of parametric RPM2 images, especially for assessment of amyloid pathology in aging and MCI.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call