TY - GEN
T1 - An adaptive fuzzy segmentation algorithm for three-dimensional magnetic resonance images
AU - Pham, Dzung L.
AU - Prince, Jerry L.
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 1999.
PY - 1999
Y1 - 1999
N2 - An algorithm is proposed for the fuzzy segmentation of two and three-dimensional multispectral magnetic resonance (MR) images that have been corrupted by intensity inhomogeneities, also known as shading artifacts. The algorithm is an extension of the two-dimensional adaptive fuzzy C-means algorithm (2-D AFCM) presented in previous work by the authors. This algorithm models the intensity inhomogeneities as a gain field that causes image intensities to smoothly and slowly vary through the image space. It iteratively adapts to the intensity inhomo-geneities and is completely automated. In this paper, we fully generalize 2-D AFCM to three-dimensional (3-D) multispectral images. Because of the potential size of 3-D image data, we also describe a new, faster multigrid-based algorithm for its implementation. We show using simulated MR data that 3-D AFCM yields significantly lower error rates than both the standard fuzzy C—means algorithm and several other competing methods when segmenting corrupted images. Its efficacy is further demonstrated using real 3-D scalar and multispectral MR brain images.
AB - An algorithm is proposed for the fuzzy segmentation of two and three-dimensional multispectral magnetic resonance (MR) images that have been corrupted by intensity inhomogeneities, also known as shading artifacts. The algorithm is an extension of the two-dimensional adaptive fuzzy C-means algorithm (2-D AFCM) presented in previous work by the authors. This algorithm models the intensity inhomogeneities as a gain field that causes image intensities to smoothly and slowly vary through the image space. It iteratively adapts to the intensity inhomo-geneities and is completely automated. In this paper, we fully generalize 2-D AFCM to three-dimensional (3-D) multispectral images. Because of the potential size of 3-D image data, we also describe a new, faster multigrid-based algorithm for its implementation. We show using simulated MR data that 3-D AFCM yields significantly lower error rates than both the standard fuzzy C—means algorithm and several other competing methods when segmenting corrupted images. Its efficacy is further demonstrated using real 3-D scalar and multispectral MR brain images.
UR - http://www.scopus.com/inward/record.url?scp=84947418652&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84947418652&partnerID=8YFLogxK
U2 - 10.1007/3-540-48714-x_11
DO - 10.1007/3-540-48714-x_11
M3 - Conference contribution
AN - SCOPUS:84947418652
SN - 3540661670
SN - 9783540661672
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 140
EP - 153
BT - Information Processing in Medical Imaging - 16th InternationalConference, IPMI 1999, Proceedings
A2 - Kuba, Attila
A2 - Samal, Martin
A2 - Todd-Pokropek, Andrew
PB - Springer Verlag
T2 - 16th International conference on Information Processing in Medical Imaging, IPMI 1999
Y2 - 28 June 1999 through 2 July 1999
ER -