Segmentation Validation Engine

The Segmentation Validation Engine (SVE) provides an automated online framework for performing validation studies of skull-stripping methods. Registered users may download 40 T1 MRI volumes, skull-strip them with the algorithm of their choice, and upload their segmentation results to the SVE website (http://sve.bmap.ucla.edu), hosted at the UCLA Brain Mapping Center (http://www.bmap.ucla.edu). The server will then compare the 40 skull-stripped results against a set of manually generated brain masks. The server computes a series of measures for the uploaded data, including Jaccard and Dice measures, and produces images of the spatial location of segmentation errors relative to a common space. Results are archived on the server, and the measures are viewable by visitors to the site. Details on the methodology can be found in:

Shattuck DW, Prasad G, Mirza M, Narr KL, and Toga AW (2009) Online Resource for Validation of Brain Segmentation Algorithms, NeuroImage(doi:10.1016/j.neuroimage.2008.10.066).