Non-Invasive estimation of Cerebral Metabolic Rate of Glucose Using Simultaneous Estimation and Cluster Analysis: A feasibility Study

Wong, K.-P., Feng, D.D., Meikle, S.R. and Fulham, M.J.

    Quantitative PET studies usually require invasive blood sampling from a peripheral artery to obtain an input function for accurate modelling. However, blood sampling is impractical in clinical PET studies. We recently proposed a non-invasive modelling approach that can simultaneously estimate parameters which describe both the input and output functions using two or more regions of interest (ROIs). However, this approach is still limited by manual delineation of ROIs which is subjective and time-consuming. In this work, we present an extension to our method where ROI delineation is performed automatically by cluster analysis so that subjectivity is reduced while at the same time ensuring that the extracted time-activity curves have distinct kinetics. Our aim was to investigate the feasibility of using the kinetics extracted by cluster analysis for non-invasive quantification of physiological parameters. The estimates and the fitted curves obtained by simultaneous estimation are in good agreement with those obtained by model fitting with the measured input function (gold standard method). We conclude that cluster analysis is able to identify distinct kinetics and hence can be used for the non-invasive quantification of physiological parameters. 1
Cite as: Wong, K.-P., Feng, D.D., Meikle, S.R. and Fulham, M.J. (2001). Non-Invasive estimation of Cerebral Metabolic Rate of Glucose Using Simultaneous Estimation and Cluster Analysis: A feasibility Study. In Proc. Selected papers from Pan-Sydney Area Workshop on Visual Information Processing (VIP2000), Sydney, Australia. CRPIT, 2. Eades, P. and Jin, J., Eds. ACS. 109-111.
pdf (from crpit.com) pdf (local if available) BibTeX EndNote GS