CUDA-based Real-Time Brain Extraction Method from Cerebral MRI Volume

S. Jiang, Z. Chen, Y. Wang, S. Yang, and W. Chen (PR China)


brain extraction, BET, CUDA, MRI volume


In this paper, we propose a GPU-based acceleration method to speed up brain extraction method from cerebral MRI volume using Compute Unified Device Architecture (CUDA). This method is based on the famous brain extraction method-BET. In order to greatly reduce the computational time for real-time processing, we perform the BET with parallel way by assigning one thread in GPU to calculate the new position of one vertex on the brain’s surface, thus all the positions of the vertices on the brain’s surface are renewed in the same time. Experiment shows the computational time of this parallel method is less than 1 second and much less than that of normal BET.

Important Links:

Go Back