Three-fold cross-validation of parkinsonian brain patterns

Novel methodology for detection and prediction of mild cognitive impairment using resting-state EEG

BACKGROUND

Early discrimination and prediction of cognitive decline are crucial for the study of neurodegenerative mechanisms and interventions to promote cognitive resiliency.

METHODS

Our research is based on resting-state electroencephalography (EEG) and the current dataset includes 137 consensus-diagnosed, community-dwelling Black Americans (ages 60-90 years, 84 healthy controls [HC]; 53 mild cognitive impairment [MCI]) recruited through Wayne State University and Michigan Alzheimer's Disease Research Center. We conducted multiscale analysis on time-varying brain functional connectivity and developed an innovative soft discrimination model in which each decision on HC or MCI also comes with a connectivity-based score.

RESULTS

The leave-one-out cross-validation accuracy is 91.97% and 3-fold accuracy is 91.17%. The 9 to 18 months' progression trend prediction accuracy over an availability-limited subset sample is 84.61%.

CONCLUSION

The EEG-based soft discrimination model demonstrates high sensitivity and reliability for MCI detection and shows promising capability in proactive prediction of people at risk of MCI before clinical symptoms may occur.

Classification of Parkinsonian syndromes from FDG-PET brain data using decision trees with SSM/PCA features

Medical imaging techniques like fluorodeoxyglucose positron emission tomography (FDG-PET) have been used to aid in the differential diagnosis of neurodegenerative brain diseases. In this study, the objective is to classify FDG-PET brain scans of subjects with Parkinsonian syndromes (Parkinson's disease, multiple system atrophy, and progressive supranuclear palsy) compared to healthy controls. The scaled subprofile model/principal component analysis (SSM/PCA) method was applied to FDG-PET brain image data to obtain covariance patterns and corresponding subject scores. The latter were used as features for supervised classification by the C4.5 decision tree method. Leave-one-out cross validation was applied to determine classifier performance. We carried out a comparison with other types of classifiers. The big advantage of decision tree classification is that the results are easy to understand by humans. A visual representation of decision trees strongly supports the interpretation process, which is very important in the context of medical diagnosis. Further improvements are suggested based on enlarging the number of the training data, enhancing the decision tree method by bagging, and adding additional features based on (f)MRI data.

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.

Scaled subprofile modeling of resting state imaging data in Parkinson's disease: methodological issues

Consistent functional brain abnormalities in Parkinson's disease (PD) are difficult to pinpoint because differences from the normal state are often subtle. In this regard, the application of multivariate methods of analysis has been successful but not devoid of misinterpretation and controversy. The Scaled Subprofile Model (SSM), a principal components analysis (PCA)-based spatial covariance method, has yielded critical information regarding the characteristic abnormalities of functional brain organization that underlie PD and other neurodegenerative disorders. However, the relevance of disease-related spatial covariance patterns (metabolic brain networks) and the most effective methods for their derivation has been a subject of debate. We address these issues here and discuss the inherent advantages of proper application as well as the effects of the misapplication of this methodology. We show that ratio pre-normalization using the mean global metabolic rate (GMR) or regional values from a "reference" brain region (e.g. cerebellum) that may be required in univariate analytical approaches is obviated in SSM. We discuss deviations of the methodology that may yield erroneous or confounding factors.