Meeting Banner
Abstract #1862

Bias Correction for Improved Segmentation and Background Parenchymal Enhancement Calculation in Multi-Center Breast MRI Trials

Alex Anh-Tu Nguyen1, Fredrik Strand1,2, Nola M Hylton1, and David C Newitt1

1Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, CA, United States, 2Department of Oncology and Pathology, Karolinska Institutet, Stockholm, Sweden

The presence of bias field inhomogeneity can negatively impact segmentation of breast fibroglandular tissue on MRI and subsequent quantification of background parenchymal enhancement. This can be particularly problematic in multi-center trials utilizing multiple imaging platforms. We have implemented the N4ITK algorithm for bias correction and evaluated the agreement between semi-automatic and semi-manual segmentation methods. Our results show that bias correction produces tissue segmentations and BPE estimates with better agreement with a reference manual segmentation method than non-corrected images.

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

Join Here