Validation of Fully-Automated Deep Learning-Based Fibroglandular Tissue Segmentation for Efficient and Reliable Quantitation of Background Parenchymal Enhancement in Breast MRI

TL;DR

This study validates a fully-automated deep learning method for segmenting fibroglandular tissue in breast MRI, demonstrating improved accuracy and efficiency over semi-automated methods for quantifying background parenchymal enhancement, a potential breast cancer risk marker.

Contribution

It provides evidence that a deep learning-based segmentation approach outperforms semi-automated methods in accuracy and correlation with qualitative assessments for BPE quantification.

Findings

DL-based segmentation correlates more strongly with qualitative BPE assessments.

DL-based method scored higher in segmentation quality by radiologists.

Automated method improves efficiency and standardization in BPE measurement.

Abstract

Background parenchymal enhancement (BPE) on breast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) shows potential as a breast cancer risk marker. Clinically, BPE is qualitatively assessed by radiologists, but quantitative BPE measures offer potential for more precise risk evaluation. This study evaluated an existing open-source, fully-automated deep learning-based (DL-based) method for segmenting fibroglandular tissue (FGT) to quantify BPE and compared it to a semi-automated fuzzy c-means method. Using breast MRI examinations from 100 women, we evaluated segmentation agreement, concordance across quantitative BPE metrics, and associations with qualitative BPE. The quality of FGT segmentations from both methods was scored by a radiologist. While the DL-based and semi-automated methods showed good agreement for quantitative BPE measurements, DL-based measures more strongly correlated with qualitative BPE assessments and DL-based segmentations were scored as higher quality by the radiologist. Our findings suggest that DL-based FGT segmentation enhances efficiency for objective BPE quantification and may improve standardized breast cancer risk assessment.