Purpose: The aim of this study was to evaluate the visual assessment of positron emission tomography images of N-[methyl-11C]2-(4'-methylaminophenyl)-6-hydroxybenzothiazole ([11C]PIB) in a patient population with mild to moderate memory impairment or dementia.
Methods: We compared the visual ratings of two readers using kappa statistics and correlated the results of visual and quantitative region of interest (ROI) analyses. The one reader had good experience in evaluating PIB images and the other had little previous experience. The sensitivity and specificity of the visual assessment was determined using quantitative data from 18 healthy controls previously examined: [11C]PIB uptake was considered as abnormal if it was more than 2 SD above the mean of the healthy subjects.
Results: The evaluation of visual classification as "normal" or "abnormal" showed good interobserver agreement (kappa = 0.90). There was a clear correlation between visual and quantitative analysis (r = 0.47-0.79, p < 0.001). The most difficult visually assessed brain area was the putamen (kappa = 0.11; correlation with quantitative analysis: reader A r = 0.22; reader B r = 0.60).
Conclusion: Our study shows that visual evaluation of [(11)C]PIB images conforms with quantitative analyses also in a clinical patient population supporting the feasibility of visual evaluation in clinical settings.