Improved Breast Mass Segmentation in Mammograms with Conditional Residual U-net

TL;DR

This paper introduces CRU-Net, a novel deep learning model combining residual learning and probabilistic graphical modeling to enhance breast mass segmentation in mammograms without extensive pre- or post-processing.

Contribution

The paper presents CRU-Net, a new deep network that integrates residual learning and probabilistic graphical modeling to improve segmentation accuracy over existing methods.

Findings

CRU-Net outperforms state-of-the-art segmentation methods on INbreast and DDSM-BCRP datasets.

CRU-Net does not require pre-processing or post-processing steps.

The model achieves superior pixel-level mass segmentation performance.

Abstract

We explore the use of deep learning for breast mass segmentation in mammograms. By integrating the merits of residual learning and probabilistic graphical modelling with standard U-Net, we propose a new deep network, Conditional Residual U-Net (CRU-Net), to improve the U-Net segmentation performance. Benefiting from the advantage of probabilistic graphical modelling in the pixel-level labelling, and the structure insights of a deep residual network in the feature extraction, the CRU-Net provides excellent mass segmentation performance. Evaluations based on INbreast and DDSM-BCRP datasets demonstrate that the CRU-Net achieves the best mass segmentation performance compared to the state-of-art methodologies. Moreover, neither tedious pre-processing nor post-processing techniques are not required in our algorithm.