Figure 11. Visualization Approach for Decoding Results of the Principle Gradient of Connectivity

A) Principal gradient of functional connectivity of the cortical surface. B) Segmentation of the principal gradient using a KMeans approach for a segment solution equal to five. C) Results of the functional decoding as shown by a radar plot and a word cloud using an LDA-based meta-analytic map generated using NeuroQuery. Radar plots present the top ten functional terms from statistically significant meta-analytic maps associated with each segment, sorted according to their values, where the values represent correlation coefficients ranging from -1 to 1, with the maps that yielded the highest correlation plotted in the angle equal to zero degrees in polar coordinates. Note that the second and third gradient shows less than 10 terms as they were the only functional and significant terms for those gradient maps. Word cloud plots were generated using a frequency estimated by the normalized probability of the topic given a word weighted by the correlation coefficient of the corresponding maps.