NEUROVIA is pleased to announce a Brain Extraction Evaluation Web
Service (BEE). The goal of this service is to provide interested users with
(i) a quantitative evaluation of the performance of their "in-house" brain extraction algorithms on a downloadable test dataset and (ii) a performance comparison against McStrip, BSE, and BET. Each algorithm can be evaluated against two sets of expertly hand-stripped masks for 15 test volumes* with respect to percent correct boundary captured, average distance between brain-mask and target boundaries, and misclassified tissue volume. The location of the results will be e-mailed to you as soon as they become available.
----------------------
*Acquired on a Siemens 1.5T scanner using 3D FLASH: TR 35 ms, TE 6 ms, FA 45 degrees, 1 NEX, FOV 165 x 220 mm, matrix 192 x 256, voxel dimensions 0.86 x 0.86 x 1.0mm. These subject volumes have undergone "facial modification" to ensure anonymity. The BET, BSE and McStrip algorithms tended to perform slightly worse with the anonymized volumes than with the originals; the differences were on the order of two percent on boundary and similarity
metrics.