Supervised Pectoral Muscle Removal in Mammography Images

In this paper, we provide the segmentation masks of the pectoral muscle for INbreast, MIAS, and a CBIS-DDSM subset datasets, which will enable the development of supervised methods and the utilization of deep learning for pectoral muscle removal from mammography images. We trained AU-Net separately on the INbreast and CBIS-DDSM subset for the segmentation of the pectoral muscle. We used cross-dataset testing to evaluate the performance of the models on an unseen dataset. The experimental results show that cross-dataset testing achieves a comparable performance to the same-dataset experiments. In addition, the models were tested on the entire MIAS dataset, and they outperformed previous methods. The segmentation masks are available at https://github.com/Parvaneh-Aliniya/pectoral_muscle_groundtruth_segmentation .