Absolute Quantitation in Neurological PET: Do We Need it?

Abstract

Summary: This article addresses the question posed in the title by examining the effects of parameters tradition ally associated with improved absolute quantitation, on the analysis of 12 acquired immune deficiency syndrome dementia complex (ADC) patients compared to a normal control group. Results are discussed within the frame work of the subprofile scaling model (SSM) for analyzing patterns of regional covariation. It is demonstrated that the ability to extract measures of group discrimination and disease progression are unaffected by (1) limited improvements in image resolution, (2) the use of trans mission scan smoothing, (3) the application of a scatter deconvolution correction, and (4) converting region-Of interest measurements of counts per voxel to mea surements of regional CMRg!c. This "robustness" of the ---_ .--. __ ..

The influence of image resolution on the positron emission tomographic measurement of caudate glucose consumption

“…Recently, Strother et al [17] have investigated the influence of improvements in image resolution, of the smoothing of transmission scans, and of the application of scatter correction on the outcome of group comparisons of rCMRGlc. These authors used a factor analytic approach, the so-called scaled subprofile model [SSM; 18], describing group differences by SSM metabolic covariance patterns.…”

Quantitative Analysis in Functional Brain Imaging

“…81 Also the implementation of neural networks for analysis of perfusion and receptor data may enhance clinical analysis of datasets as has been shown by several authors. [82][83][84][85] Spatial activation patterns can also be investigated by ANCOVA-adjusted voxel-based principal component analysis (PCA) and PCA with a Scaled Subprofile Model (SSM) 86 where all obtainable information about regional ratios can, in theory, be derived from the SSM regional covariance patterns, 87 allowing the generation of disease-specific metabolic or perfusion signatures or covariance patterns.…”

Discussion of PET Workshop Reports, Including Recommendations of PET Data Analysis Working Group

“…Related but somewhat different viewpoints, which examined the signal-to-noise ratio conse quences of changes in absolute regional quantita tion, were considered in three separate papers by Carson (1991), Strother et al (1987Strother et al ( , 1991, and Rot tenberg et al (1991). Carson discussed the general role of quantitation in PET and the conditions under which physiological models may not be important.…”

“…Each group had cogni tive deficits that corresponded to the specific met abolic pattern. Similarly, the scaled subprofile model of Moeller et al (1991) can be used to extract group specific profiles of PET metabolic patterns that distinguish normal from pathological subjects. In the future, cognitive and drug stimulation par adigms or lesion studies in animals might be com bined with PET to examine hypotheses derived from correlations of resting metabolism or flow rates.…”

“…Carson discussed the general role of quantitation in PET and the conditions under which physiological models may not be important. Using Moeller's scaled subprofile model, Strother et al (1987Strother et al ( , 1991 and Rottenberg et al (1991) dem onstrated that successful discrimination of normals from acquired immune deficiency syndrome de mentia complex patients, and the prediction of dis ease severity, are largely unaffected by significant changes in (1) PET reconstruction parameters, (2) the noise in measured transmission scans, (3) the size of applied ROIs, or, for that matter, (4) the use of raw PET count rate data rather than glucose met abolic rates calculated by the Sokoloff model.…”