The first implant and mammary gland segmentation dataset for surgical planning of breast reconstruction based on MRI images
Breast cancer is the most common cancer in the female population globally, and reconstructive breast surgery has been widely adopted to improve the quality of life for patients who have undergone mastectomy. One of the most popular techniques in reconstructive breast surgery is implant- based reconstruction, in which segmentation of implant and mammary glands are critical for surgical planning. Recently, segmentation of breast magnetic resonance imaging (MRI) has been widely used for implant segmentation and evaluation. How- ever, most of the existing works focus on manual segmentation or semi-automatic segmentation, and automatic segmentation and evaluation of the implant and mammary gland are rarely explored. In this paper, we propose ImageBR, the first implant and mammary gland segmentation dataset for surgical planning of breast reconstruction based on MRI images. ImageBR contains 44 MRI images which are labelled by a team of four experienced surgeons from Department of Breast Oncology with segmentation labels including implant and mammary gland. Particularly, the implant is labeled in the MRI images based on clinical knowledge which can possibly best fit the actual implant. We also present a baseline method with anatomical prior integration and weighted loss optimization for automatic implant and mammary gland segmentation. Results show that the baseline framework can achieve a segmentation accuracy of 83.0% and 74.4%, and there is still big room for further improvement. We hope that ImageBR can stimulate further research and lead to innovative and generic solutions that would have an impact in multiple domains. Our dataset and code are released to the public.
Please send emails to me xiao.wei.xu@foxmail.com for the link and the password to download the dataset and the benchmark.