Premium
Robust tissue–air volume segmentation of MR images based on the statistics of phase and magnitude: Its applications in the display of susceptibility‐weighted imaging of the brain
Author(s) -
Du Yiping P.,
Jin Zhaoyang
Publication year - 2009
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.21910
Subject(s) - segmentation , standard deviation , magnitude (astronomy) , volume (thermodynamics) , artificial intelligence , phase (matter) , pattern recognition (psychology) , computer science , intensity (physics) , image segmentation , kernel (algebra) , mathematics , statistics , physics , optics , quantum mechanics , astronomy , combinatorics
Purpose To develop a robust algorithm for tissue–air segmentation in magnetic resonance imaging (MRI) using the statistics of phase and magnitude of the images. Materials and Methods A multivariate measure based on the statistics of phase and magnitude was constructed for tissue–air volume segmentation. The standard deviation of first‐order phase difference and the standard deviation of magnitude were calculated in a 3 × 3 × 3 kernel in the image domain. To improve differentiation accuracy, the uniformity of phase distribution in the kernel was also calculated and linear background phase introduced by field inhomogeneity was corrected. The effectiveness of the proposed volume segmentation technique was compared to a conventional approach that uses the magnitude data alone. Results The proposed algorithm was shown to be more effective and robust in volume segmentation in both synthetic phantom and susceptibility‐weighted images of human brain. Using our proposed volume segmentation method, veins in the peripheral regions of the brain were well depicted in the minimum‐intensity projection of the susceptibility‐weighted images. Conclusion Using the additional statistics of phase, tissue–air volume segmentation can be substantially improved compared to that using the statistics of magnitude data alone. J. Magn. Reson. Imaging 2009;30:722–731. © 2009 Wiley‐Liss, Inc.