March 3, 2026Open Access

ER2Net: an evidential reasoning rule–enabled neural network for reliable triple-negative breast cancer tumor segmentation in magnetic resonance imaging

Key Points

Segmentation accuracy significantly improved, enhancing reliability for tumor detection.
Key performance enhancements include better sensitivity, confidence calibration, and uncertainty estimation.
Assessment using enhanced fusion techniques and an uncertainty metric was employed in the analysis.
Highlights the potential for MRI-based tools to aid in personalized treatment for breast cancer patients.

Abstract

ER2Net introduces WR segmentation and BF as enhanced fusion techniques and PWRE as a uncertainty metric. Together, these advances improve segmentation accuracy, sensitivity, confidence calibration, and uncertainty estimation, paving the way for reliable MRI-based tools to support personalized treatment planning and response assessment in TNBC.

ER2Net: an evidential reasoning rule–enabled neural network for reliable triple-negative breast cancer tumor segmentation in magnetic resonance imaging

Key Points

Abstract

Cite This Study