Abstract #3840

Feasibility of utilizing heterogeneity of hepatic stiffness in 3D MR elastography to improve detection of liver fibrosis in pediatric patients with nonalcoholic fatty liver disease

Kang Wang¹, Paul Manning¹, Tanya Wolfson ², Michael S. Middleton¹, Jeffrey Schwimmer³, Kimberley Newton³, Cynthia Behling³, Janis Durelle³, Melissa Paiz³, Jorge Angeles³, Meng Yin⁴, Kevin Glaser⁴, Richard Ehman⁴, and Claude Sirlin¹

¹Liver Imaging Group, Department of Radiology, University of California, San Diego, School of Medicine, San Diego, CA, United States, ²Computational and Applied Statistics Laboratory, University of California, San Diego, San Diego, CA, United States, ³Department of Pediatric, University of California, San Diego, San Diego, CA, United States, ⁴Departments of Radiology, Mayo Clinic, Rochester, MN, United States

We evaluated the feasibility of utilizing heterogeneity of hepatic stiffness in 3D MR elastography to improve detection of liver fibrosis in a cohort of 70 children with NAFLD. Children were dichotomized into two classes of fibrosis. We characterized the heterogeneity of hepatic stiffness by fitting a bi-Gaussian model to the histogram of hepatic stiffness. Features from the bi-Gaussian model and the known class labels were used to develop a support vector machine (SVM) classification model to predict fibrosis. We demonstrated that the SVM model has better overall classification performance than the calculated mean hepatic stiffness as measured by AUROC.

This abstract and the presentation materials are available to members only; a login is required.

Join Here