Abstract #4638
Fast and Accurate Brain Tissue Segmentation with Polarity Categorization (POLCAT)
Steven Kecskemeti 1,2 and Andrew L Alexander 3,4
1
Waisman Center, University of Wisconsin,
Madison, WI, United States,
2
Radiology,
University of Wisconsin, Madison, WI, United States,
3
Medical
Physics, University of Wisconsin, WI, United States,
4
Psychiatry,
University of Wisconsin, Madison, United States
Intensity-based brain tissue segmentation algorithms
rely on post-hoc image intensity at a single point along
the relaxation recovery curve of MPRAGE exams, making
them very sensitive to unexpected signal variations such
as the spatial heterogeneity of radio-frequency (RF)
coil sensitivities. This work develops a novel, robust
and efficient method for brain tissue segmentation that
relies on intrinsic properties such as T1 and is
insensitive to variations in RF receiver coil bias. The
method assigns the tissue class according to the sign of
the real signal intensity after voxel-wise
complex-multiplication of inversion recovery images with
different inversion times.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
