Methods for detecting functional classifications in neuroimaging data

Abstract

Data-driven statistical methods are useful for examining the spatial organization of human brain function. Cluster analysis is one approach that aims to identify spatial classifications of temporal brain activity profiles. Numerous clustering algorithms are available, and no one method is optimal for all areas of application because an algorithm's performance depends on specific characteristics of the data. K-means and fuzzy clustering are popular for neuroimaging analyses, and select hierarchical procedures also appear in the literature. It is unclear which clustering methods perform best for neuroimaging data. We conduct a simulation study, based on PET neuroimaging data, to evaluate the performances of several clustering algorithms, including a new procedure that builds on the kth nearest neighbor method. We also examine three stopping rules that assist in determining the optimal number of clusters. Five hierarchical clustering algorithms perform best in our study, some of which are new to neuroimaging analyses, with Ward's and the beta-flexible methods exhibiting the strongest performances. Furthermore, Ward's and the beta-flexible methods yield the best performances for noisy data, and the popular K-means and fuzzy clustering procedures also perform reasonably well. The stopping rules also exhibit good performances for the top five clustering algorithms, and the pseudo-T2 and pseudo-F stopping rules are superior for noisy data. Based on our simulations for both noisy and unscaled PET neuroimaging data, we recommend the combined use of the pseudo-F or pseudo-T2 stopping rule along with either Ward's or the beta-flexible clustering algorithm.

Do we need to revise the tripartite subdivision hypothesis of the human subthalamic nucleus (STN)? Response to Alkemade and Forstmann

“…The main body of the text states that the tripartite division was the optimal solution in 58% of subjects tested. Using this as supportive evidence, we then extracted a tripartite parcellation for each and every individual using ward linkage clustering, a hard clustering technique adopted by many other groups (Bowman et al, 2004;Chen et al, 2012;Palomero-Gallagher et al, 2009;Blumensath et al, 2013;Caspers et al, 2013;Bzdok et al, 2013). Furthermore, we provided individual results for tripartite parcellations for six STN pairs in the supplementary material.…”

Brain Imaging Analysis

“…There are well established methods in the statistical literature for performing cluster analysis. These data-driven methods have been successfully applied in brain imaging, including K-means (2, 44, 45), fuzzy clustering (3, 35, 36, 80), hierarchical clustering methods (45, 12, 13, 26, 81), a hybrid hierarchical K-means approach (37), and dynamical cluster analysis (DCA) (4), among others. Hierarchical clustering generally begins with each brain region (or node) as a single cluster, calculates functional distances between all pairs of brain regions, e.g.…”

Sparse geostatistical analysis in clustering fMRI time series

“…Step three -Clustering using k means algorithm: The k means algorithm is common in neuroimaging applications because of its computational advantages: computations are fast, the algorithm does not require retention of all distances, and convergence occurs quickly (Bowman et al, 2004). For a given number of clusters k, the algorithm iteratively minimizes the within-class variance by 1.…”