Skip to main content
. 2012 Feb;2(1):33–43. doi: 10.1089/brain.2011.0026

FIG. 2.

FIG. 2.

Simulation for constrained SBM. (a) Simulated seed region (red circular) in the GM. (b) Simulated GM and WM images. Each of the three regions (white circular) was generated within the GM and WM images, and three different networks were simulated. Network 1 was formed by region 1 and 2 in GM and the linked region 4 and 6 in WM; network 2 was the region 3 in GM; and network 3 was the region 5 in WM. 100 GM and WM images were created separately representing images for 100 subjects, and Gaussian noise was then added to the images. (c) Constrained SBM analysis results. Utilizing the seed regions as the constraint on the GM and separating most of the Gaussian noise into its own component, a constrained SBM analysis on the 200 images successfully detected network 1 containing both GM regions and related WM regions.