基于MRF-MAP和模糊质量图的相位展开算法

MRF-MAP and fuzzy quality map based phase unwrapping algorithm

相位展开是磁共振成像技术应用中最关键的环节之一,可以为磁共振的某些重要临床应用提供精确的相位信息。然而,由于临床磁共振成像过程中,部分区域真实的相位存在急剧变化,同时伴有不同性态的噪声污染,导致相位展开时存在信息的高度不一致性。为了有效地解决上述难题,基于马尔可夫-最大后验(Markov Random Field& Maximum A Posterioi,MRF-MAP)模型,首次将相位展开看作计算机视觉中的标记问题,并结合磁共振相位数据的特点,设计出相位图的模糊质量图,完成相位展开的能量函数构建。针对能量函数的优化求解,采用高效的图割算法进行。

Phase unwrapping is one of the most key steps in magnetic resonance imaging （MRI） applications in which accurate phase is required.The main difficulty in phase unwrapping stems from the noise and（or） abrupt variation in the true phase data. To address this problem,we present a method based on Markov Random Field ＆ Maximum A Posterioi （MRF-MAP）.We regard phase unwrapping as labeling problem in computer vision for the first time.Considering of the characteristic of MRI phase data, we design a fuzzy quality map and then obtain an energy function which is suitable to be minimize by an efficient algorithm based on graph cuts.