Maximize uniformity summation heuristic (MUSH): a highly accurate simple method for intracranial delineation

摘要: A common procedure performed by many groups in the analysis of neuroimaging data is separating brain from other tissues. This often utilized both volumetric studies as well functional imaging studies. Regardless of the intent, an accurate, robust method identifying or cranial vault imperative. While this a common requirement, there are relatively few tools to perform task. Most these require T1 weighted image and are therefore not able accurately define region that includes surface CSF. In paper, we have developed novel brain extraction technique termed Maximize Uniformity Summation Heuristic (MUSH) optimization. The algorithm was designed for extraction CSF multi-modal magnetic resonance (MR) study. The method forms linear combination MR make signal intensity within as uniform possible. resulting thresholded simple morphological operators generate the resulting representation brain. was applied sample 20 scans compared to results generated 3dSkullStrip, 3dIntracranial, BET, BET2. average Jaccard metrics twenty subjects 0.66 (BET), 0.61 (BET2), 0.88 (3dIntracranial), 0.91 (3dSkullStrip), 0.94 (MUSH).