We propose to investigate the validity and applicability of fuzzy clustering (FCM) for the identification of dynamic functional connectivity (dFC) patterns in resting-state fMRI data, and comparing it with two approaches that have been used in this context (PCA and K-means). For such purpose, all methods were applied to data simulating either the joint or separate expression of dFC patterns, and to empirical data, collected from epilepsy patients. Both clustering methods, particularly FCM, outperformed PCA. Concomitantly, results from empirical data indicated that the occurrence of epileptic activity of patients was separately expressed by the dFC patterns.
This abstract and the presentation materials are available to members only; a login is required.