Abstract #2157
Accuracy of Texture Analysis to Detect Hepatic Fibrosis on T1 Weighted Contrast Enhanced MR
Hei Shun Yu 1 , Baojun Li 2 , Karen Buch 2 , Naznin Daginawala 2 , Brian Tischler 2 , and Stephan Anderson 2
1
Boston University Medical Center, Boston,
MA, United States,
2
Boston
University Medical Center, MA, United States
The purpose of this study was to diagnose and stage
hepatic fibrosis by performing a texture analysis on
post-gadolinium T1 weighted images of the liver
parenchyma in the portal venous and equilibrium phases.
A MATLAB-based texture analysis program was employed to
extract 42 texture features for thirty patients. These
texture parameters were analyzed using a linear
discriminant analysis to calculate sensitivities and
specificities. Findings of this study demosntrate
specificities of 72% and 82% discriminating lower from
higher grades of hepatic fibrosis on equilibrium and
venous phases. The ability to non-invasively assess and
quantify the severity of hepatic fibrosis is of
diagnostic clinical importance and could potentially
minimize invasive biopsy techniques to assess severity
of fibrosis.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
