Phoebe G. Spetsieris

Metabolic resting-state brain networks in health and disease

Significance We present an innovative approach to evaluate default mode network (DMN) activity in individual subjects using metabolic imaging. After characterizing a distinct set of metabolic resting state networks (RSNs) in healthy subjects, network activity was tracked over time in patients with neurodegenerative disorders, such as Parkinson’s disease and Alzheimer’s disease. We found that the dominant normal metabolic RSN, which corresponded to the DMN, is preserved in early-stage Parkinson’s disease patients. Although significant DMN reductions developed later, these changes were reversible in part by dopamine treatment. This finding contrasts with Alzheimer’s disease, in which DMN loss is rapid and continuous, beginning before clinical diagnosis. Metabolic imaging can provide a versatile, quantitative means of assessing brain disease at the network level. The delineation of resting state networks (RSNs) in the human brain relies on the analysis of temporal fluctuations in functional MRI signal, representing a small fraction of total neuronal activity. Here, we used metabolic PET, which maps nonfluctuating signals related to total activity, to identify and validate reproducible RSN topographies in healthy and disease populations. In healthy subjects, the dominant (first component) metabolic RSN was topographically similar to the default mode network (DMN). In contrast, in Parkinson’s disease (PD), this RSN was subordinated to an independent disease-related pattern. Network functionality was assessed by quantifying metabolic RSN expression in cerebral blood flow PET scans acquired at rest and during task performance. Consistent task-related deactivation of the “DMN-like” dominant metabolic RSN was observed in healthy subjects and early PD patients; in contrast, the subordinate RSNs were activated during task performance. Network deactivation was reduced in advanced PD; this abnormality was partially corrected by dopaminergic therapy. Time-course comparisons of DMN loss in longitudinal resting metabolic scans from PD and Alzheimer’s disease subjects illustrated that significant reductions appeared later for PD, in parallel with the development of cognitive dysfunction. In contrast, in Alzheimer’s disease significant reductions in network expression were already present at diagnosis, progressing over time. Metabolic imaging can directly provide useful information regarding the resting organization of the brain in health and disease.

Parkinson's Disease: Increased Motor Network Activity in the Absence of Movement

We used a network approach to assess systems-level abnormalities in motor activation in humans with Parkinsons disease (PD). This was done by measuring the expression of the normal movement-related activation pattern (NMRP), a previously validated activation network deployed by healthy subjects during motor performance. In this study, NMRP expression was prospectively quantified in 15O-water PET scans from a PD patient cohort comprised of a longitudinal early-stage group (n = 12) scanned at baseline and at two or three follow-up visits two years apart, and a moderately advanced group scanned on and off treatment with either subthalamic nucleus deep brain stimulation (n = 14) or intravenous levodopa infusion (n = 14). For each subject and condition, we measured NMRP expression during both movement and rest. Resting expression of the abnormal PD-related metabolic covariance pattern was likewise determined in the same subjects. NMRP expression was abnormally elevated (p < 0.001) in PD patients scanned in the nonmovement rest state. By contrast, network activity measured during movement did not differ from normal (p = 0.34). In the longitudinal cohort, abnormal increases in resting NMRP expression were evident at the earliest clinical stages (p < 0.05), which progressed significantly over time (p = 0.003). Analogous network changes were present at baseline in the treatment cohort (p = 0.001). These abnormalities improved with subthalamic nucleus stimulation (p < 0.005) but not levodopa (p = 0.25). In both cohorts, the changes in NMRP expression that were observed did not correlate with concurrent PD-related metabolic covariance pattern measurements (p > 0.22). Thus, the resting state in PD is characterized by changes in the activity of normal as well as pathological brain networks.

Abnormal metabolic brain networks in a nonhuman primate model of parkinsonism.

Parkinsons disease (PD) is associated with a characteristic regional metabolic covariance pattern that is modulated by treatment. To determine whether a homologous metabolic pattern is also present in nonhuman primate models of parkinsonism, 11 adult macaque monkeys with parkinsonism secondary to chronic systemic 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and 12 age-matched healthy animals were scanned with [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET). A subgroup comprising five parkinsonian and six control animals was used to identify a parkinsonism-related pattern (PRP). For validation, analogous topographies were derived from other subsets of parkinsonian and control animals. The PRP topography was characterized by metabolic increases in putamen/pallidum, thalamus, pons, and sensorimotor cortex, as well as reductions in the posterior parietal-occipital region. Pattern expression was significantly elevated in parkinsonian relative to healthy animals (P < 0.00001). Parkinsonism-related topographies identified in the other derivation sets were very similar, with significant pairwise correlations of region weights (r > 0.88; P < 0.0001) and subject scores (r > 0.74; P < 0.01). Moreover, pattern expression in parkinsonian animals correlated with motor ratings (r > 0.71; P < 0.05). Thus, homologous parkinsonism-related metabolic networks are demonstrable in PD patients and in monkeys with experimental parkinsonism. Network quantification may provide a useful biomarker for the evaluation of new therapeutic agents in preclinical models of PD.

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

The scaled subprofile model (SSM)(1-4) is a multivariate PCA-based algorithm that identifies major sources of variation in patient and control group brain image data while rejecting lesser components (Figure 1). Applied directly to voxel-by-voxel covariance data of steady-state multimodality images, an entire group image set can be reduced to a few significant linearly independent covariance patterns and corresponding subject scores. Each pattern, termed a group invariant subprofile (GIS), is an orthogonal principal component that represents a spatially distributed network of functionally interrelated brain regions. Large global mean scalar effects that can obscure smaller network-specific contributions are removed by the inherent logarithmic conversion and mean centering of the data(2,5,6). Subjects express each of these patterns to a variable degree represented by a simple scalar score that can correlate with independent clinical or psychometric descriptors(7,8). Using logistic regression analysis of subject scores (i.e. pattern expression values), linear coefficients can be derived to combine multiple principal components into single disease-related spatial covariance patterns, i.e. composite networks with improved discrimination of patients from healthy control subjects(5,6). Cross-validation within the derivation set can be performed using bootstrap resampling techniques(9). Forward validation is easily confirmed by direct score evaluation of the derived patterns in prospective datasets(10). Once validated, disease-related patterns can be used to score individual patients with respect to a fixed reference sample, often the set of healthy subjects that was used (with the disease group) in the original pattern derivation(11). These standardized values can in turn be used to assist in differential diagnosis(12,13) and to assess disease progression and treatment effects at the network level(7,14-16). We present an example of the application of this methodology to FDG PET data of Parkinsons Disease patients and normal controls using our in-house software to derive a characteristic covariance pattern biomarker of disease.

Characterization of disease-related covariance topographies with SSMPCA toolbox: Effects of spatial normalization and PET scanners

To generate imaging biomarkers from disease‐specific brain networks, we have implemented a general toolbox to rapidly perform scaled subprofile modeling (SSM) based on principal component analysis (PCA) on brain images of patients and normals. This SSMPCA toolbox can define spatial covariance patterns whose expression in individual subjects can discriminate patients from controls or predict behavioral measures. The technique may depend on differences in spatial normalization algorithms and brain imaging systems. We have evaluated the reproducibility of characteristic metabolic patterns generated by SSMPCA in patients with Parkinsons disease (PD). We used [18F]fluorodeoxyglucose PET scans from patients with PD and normal controls. Motor‐related (PDRP) and cognition‐related (PDCP) metabolic patterns were derived from images spatially normalized using four versions of SPM software (spm99, spm2, spm5, and spm8). Differences between these patterns and subject scores were compared across multiple independent groups of patients and control subjects. These patterns and subject scores were highly reproducible with different normalization programs in terms of disease discrimination and cognitive correlation. Subject scores were also comparable in patients with PD imaged across multiple PET scanners. Our findings confirm a very high degree of consistency among brain networks and their clinical correlates in PD using images normalized in four different SPM platforms. SSMPCA toolbox can be used reliably for generating disease‐specific imaging biomarkers despite the continued evolution of image preprocessing software in the neuroimaging community. Network expressions can be quantified in individual patients independent of different physical characteristics of PET cameras. Hum Brain Mapp 35:1801–1814, 2014.

Quantifying significance of topographical similarities of disease-related brain metabolic patterns.

Multivariate analytical routines have become increasingly popular in the study of cerebral function in health and in disease states. Spatial covariance analysis of functional neuroimaging data has been used to identify and validate characteristic topographies associated with specific brain disorders. Voxel-wise correlations can be used to assess similarities and differences that exist between covariance topographies. While the magnitude of the resulting topographical correlations is critical, statistical significance can be difficult to determine in the setting of large data vectors (comprised of over 100,000 voxel weights) and substantial autocorrelation effects. Here, we propose a novel method to determine the p-value of such correlations using pseudo-random network simulations.

Abnormal network topographies and changes in global activity: Absence of a causal relationship

Changes in regional brain activity can be observed following global normalization procedures to reduce variability in the data. In particular, spurious regional differences may appear when scans from patients with low global activity are compared to those from healthy subjects. It has thus been suggested that the consistent increases in subcortical activity that characterize the abnormal Parkinsons disease-related metabolic covariance pattern (PDRP) are artifacts of global normalization, and that similar topographies can be identified in scans from healthy subjects with varying global activity. To address this issue, we examined the effects of experimental reductions in global metabolic activity on PDRP expression. Ten healthy subjects underwent ¹⁸F-fluorodeoxyglucose PET in wakefulness and following sleep induction. In all subjects, the global metabolic rate (GMR) declined with sleep (mean -34%, range: -17 to -56%), exceeding the test-retest differences of the measure (p<0.001). By contrast, sleep-wake differences in PDRP expression did not differ from test-retest differences, and did not correlate (R²=0.04) with concurrent declines in global metabolic activity. Indeed, despite significant GMR reductions in sleep, PDRP values remained within the normal range. Likewise, voxel weights on the principal component patterns resulting from combined analysis of the sleep and wake scans did not correlate (R²<0.07) with the corresponding regional loadings on the PDRP topography. In aggregate, the data demonstrate that abnormal PDRP expression is not induced by reductions in global activity. Moreover, significant declines in GMR are not associated with the appearance of PDRP-like spatial topographies.

Three-fold cross-validation of parkinsonian brain patterns

Abnormal physiological networks of brain areas in disease can be identified by applying specialized multivariate computational algorithms based on principal component analysis to functional image data. Here we demonstrate the reproducibility of network patterns derived using positron emission tomography (PET) data in independent populations of parkinsonian patients for a large, clinically validated data set comprised of subjects with idiopathic Parkinsons disease (iPD), multiple system atrophy (MSA) and progressive supranuclear palsy (PSP). Correlation of voxel values of network patterns derived for the same condition in different data sets was high. To further illustrate the validity of these networks, we performed single subject differential diagnosis of prospective test subjects to determine the most probable case based on a subjects network scores expressed for each of these distinct parkinsonian syndromes. Three-fold cross-validation was performed to determine accuracy and positive predictive rates based on networks derived in separate folds of the composite data set. A logistic regression based classification algorithm was used to train in each fold and test in the remaining two folds. Combined accuracy for each of the three folds ranged from 82% to 93% in the training set and was approximately 81% for prospective test subjects.

Computer aided diagnosis of functional brain disorders using PET on a PC based platform

We describe a portable PC based platform for computer aided diagnosis (CAD) of functional brain disorders using positron emission tomography (PET) that integrates a variety of methods including quantitative and visual techniques as well as extensively automated procedures. Statistical procedures such as the scaled subprofile model (SSM) in conjunction with principal component analysis are employed to derive neuronal networks that can uniquely identify characteristic metabolic covariance patterns. By measuring the expression of these networks in diseased subjects on an individual basis we can help automate the diagnosis of conditions such as Parkinsons disease and evaluate treatment efficacy and disease progression.

New Strategies for Automated Differential Diagnosis of Degenerative Brain Disorders

New strategies are considered for automated, single-subject differential diagnosis of independent degenerative brain disorders characterized by similar clinical symptoms using functional imaging. The methodology of these strategies is described and its application in Parkinsonian movement disorders is illustrated for PET data. Using an automated diagnostic topographic profile rating (TPR) technique based on the scaled subprofile model (SSM-PCA), single-subject score values for different conditions are compared with reference values to predict diagnosis. The discriminatory parameters of reference score sets associated with significant SSM principal components referred to as group invariant subprofiles (GIS networks) are examined. It is shown that the extraction of exclusive subnetworks that stem from contrasting image features between conditions can be an effective tool for optimization that does not require expert knowledge.