High-accuracy hierarchical parallel technique for hidden Markov model-based 3D magnetic resonance image brain segmentation

摘要

Scientific applications represent a dominant sector of compute-intensive applications. Using massively parallelrnprocessing systems increases the feasibility to automate such applications because of the cooperationrnamong multiple processors to perform the designated task. This paper proposes a parallel hidden Markovrnmodel (HMM) algorithm for 3D magnetic resonance image brain segmentation using two approaches. Inrnthe first approach, a hierarchical/multilevel parallel technique is used to achieve high performance for thernrunning algorithm. This approach can speed up the computation process up to 7.8u0001 compared with a serialrnrun. The second approach is orthogonal to the first and tries to help in obtaining a minimum error for 3Drnmagnetic resonance image brain segmentation using multiple processes with different randomization pathsrnfor cooperative fast minimum error convergence. This approach achieves minimum error level for HMMrntraining not achievable by the serial HMM training on a single node. Then both approaches are combinedrnto achieve both high accuracy and high performance simultaneously. For 768 processing nodes of a BluernGene system, the combined approach, which uses both methods cooperatively, can achieve high-accuracyrnHMM parameters with 98% of the error level and 2.6u0001 speedup compared with the pure accuracy-orientedrnapproach alone.