Factor analysis of regional cerebral glucose metabolic rates in healthy men

Principal component analysis of the dynamic response measured by fMRI: a generalized linear systems framework

Principal component analysis (PCA) is one of several structure-seeking multivariate statistical techniques, exploratory as well as inferential, that have been proposed recently for the characterization and detection of activation in both PET and fMRI time series data. In particular, PCA is data driven and does not assume that the neural or hemodynamic response reaches some steady state, nor does it involve correlation with any pre-defined or exogenous experimental design template. In this paper, we present a generalized linear systems framework for PCA based on the singular value decomposition (SVD) model for representation of spatio-temporal fMRI data sets. Statistical inference procedures for PCA, including point and interval estimation will be introduced without the constraint of explicit hypotheses about specific task-dependent effects. The principal eigenvectors capture both the spatial and temporal aspects of fMRI data in a progressive fashion; they are inherently matched to unique and uncorrelated features and are ranked in order of the amount of variance explained. PCA also acts as a variation reduction technique, relegating most of the random noise to the trailing components while collecting systematic structure into the leading ones. Features summarizing variability may not directly be those that are the most useful. Further analysis is facilitated through linear subspace methods involving PC rotation and strategies of projection pursuit utilizing a reduced, lower-dimensional natural basis representation that retains most of the information. These properties will be illustrated in the setting of dynamic time-series response data from fMRI experiments involving pharmacological stimulation of the dopaminergic nigro-striatal system in primates.

Visualizing fronto-striatal circuitry and neuroleptic effects in schizophrenia

Disturbances in fronto-striatal circuitry have been postulated to be important in schizophrenia. Positron emission tomography typically shows decreased metabolic rates in these areas relative to other brain areas in schizophrenia. After treatment with typical neuroleptics, striatal metabolic rates are increased, but other brain areas tend not to show significant changes. Atypical neuroleptics less markedly affect striatal metabolic rates, but show wider cortical effects. In order to examine fronto-striatal circuitry, a technique for visualizing the correlations between metabolic rates in all brain areas was applied in 33 controls and 27 unmedicated schizophrenic patients. Correlation images revealed strong fronto-striatal connections in controls, but weak fronto-striatal links in schizophrenic patients. Changes in striatal circuits, also reflected in recent anatomical studies, may be important for understanding antipsychotic effects.

Brain networks affected by synchronized sleep visualized by positron emission tomography

Nineteen lightly sleep-deprived healthy volunteers were examined with H2(15)O and positron emission tomography (PET). Scanning was performed during wakefulness and after the subjects had fallen asleep. Sleep stage was graded retrospectively from electroencephalogram (EEG) recordings, and scans were divided into two groups: wakefulness or synchronized sleep. Global flow was quantified, revealing no difference between sleep and wakefulness. A pixel-by-pixel-blocked one-way analysis of variance (ANOVA) was performed after correcting for differences in anatomy and global flow. The sum of squares of the z-score distribution showed a highly significant (P < 0.00001) omnibus difference between sleep and wakefulness. The z-score images indicated decreased flow in the thalamus and the frontal and parietal association cortices and increased flow in the cerebellum during sleep. A principal component (PC) analysis was performed on data after correction for global flow and block effects, and a multivariate analysis of variance (MANOVA) on all PC scores revealed significant (P = 0.00004) differences between sleep and wakefulness. Principal component's 2 and 5 correlated to sleep and revealed distinct networks consisting of PC 2, cerebellum and frontal and parietal association cortices, and PC 5, thalamus.

Delineation and quantitation of brain lesions by fuzzy clustering in positron emission tomography

In this study, we investigate the application of the fuzzy clustering to the anatomical localization and quantitation of brain lesions in Positron Emission Tomography (PET) images. The method is based on the Fuzzy C-Means (FCM) algorithm. The algorithm segments the PET image data points into a given number of clusters. Each cluster is an homogeneous region of the brain (e.g. tumor). A feature vector is assigned to a cluster which has the highest membership degree. Having the label affected by the FCM algorithm to a cluster, one may easily compute the corresponding spatial localization, area and perimeter. Studies concerning the evolution of a tumor after different treatments in two patients are presented.

Commentary and opinion: I. Principal component analysis, variance partitioning, and "functional connectivity"

We briefly review the need for careful study of "variance partitioning" and "optimal model selection" in functional positron emission tomography (PET) data analysis, emphasizing the use of principal component analysis (PCA) and the importance of data analytic techniques that allow for heterogeneous spatial covariance structures. Using an [15O]water dataset, we demonstrate that--even after data processing--the intrasubject signal component of primary interest in baseline activation studies constitutes a very small fraction of the intersubject variance. This small intrasubject variance component is subtly but significantly changed by using analysis of covariance instead of scaled subprofile model processing before applying PCA. Finally, we argue that the concept of "functional connectivity" should be interpreted very generally until the relative roles of inter- and intrasubject variability in both disease and normal PET datasets are better understood.