Neurology

Pattern Classification

Pattern classification is a growing area of research in fMRI that aims at interpreting activation patterns in terms of underlying cognitive processes. Recent studies have demonstrated a high degree of pattern specificity for discrete neural processes and it is expected that pattern classification will play a major role in understanding cortical organization in individual subjects. Specifically, automatic interpretation and classification of neuroimaging data may hold important keys for understanding the human mind, which has raised interests due to the potential clinical applications. Information embedded in the spatial shape and extent of activation patterns, and differences in voxel-to-voxel time course, are not easily quantified with conventional analysis tools, such as statistical parametric mapping (SPM). Pattern classification in functional MRI (fMRI) is a novel approach, which promises to characterize subtle differences in activation patterns between different tasks. There is growing evidence that exquisite pattern specificity exists in visual cortex and other brain areas, such as motor cortex, auditory cortex and parietal cortex. However, automatic and reliable classification of patterns is challenging due to the high dimensionality of fMRI data, the small number of available data sets, inter-individual differences in activation patterns, and dependence on the image acquisition methodology. We recently introduced spatially distributed classifier for boosting to further reduce the dimensionality problem.

Our technology development is aimed at improving automated pattern classification to facilitate clinical applications of fMRI, such as fMRI-guided mental training, identification of disease markers (e.g. psychiatric disorders, epilepsy, migraine), and prediction of treatment response and relapse.

Citations

Martínez-Ramón M., Koltchinsky V., Verzi, S., Heileman, G., Posse, S.,"fMRI pattern classification using neuroanatomically constrained boosting," Neuroimage, 2006 Jul 1;31(3):1129-41.

Zheng, W., Ackley, E.S., Martínez-Ramón, M., Posse, S., "Spatially aggregated multiclass pattern classification in functional MRI using optimally selected functional brain areas," Magnetic Resonance Imaging, 2012 Aug 16.