8 February 2017 Multi-scales region segmentation for ROI separation in digital mammograms
Author Affiliations +
Proceedings Volume 10225, Eighth International Conference on Graphic and Image Processing (ICGIP 2016); 1022515 (2017) https://doi.org/10.1117/12.2267046
Event: Eighth International Conference on Graphic and Image Processing, 2016, Tokyo, Japan
Abstract
Mammography is currently the most effective imaging modality used by radiologists for the screening of breast cancer. Segmentation is one of the key steps in the process of developing anatomical models for calculation of safe medical dose of radiation. This paper explores the potential of the statistical region merging segmentation technique for Breast segmentation in digital mammograms. First, the mammograms are pre-processing for regions enhancement, then the enhanced images are segmented using SRM with multi scales, finally these segmentations are combined for region of interest (ROI) separation and edge detection. The proposed algorithm uses multi-scales region segmentation in order to: separate breast region from background region, region edge detection and ROIs separation. The experiments are performed using a data set of mammograms from different patients, demonstrating the validity of the proposed criterion. Results show that, the statistical region merging segmentation algorithm actually can work on the segmentation of medical image and more accurate than another methods. And the outcome shows that the technique has a great potential to become a method of choice for segmentation of mammograms.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Dapeng Zhang, Di Zhang, Yue Li, Wei Wang, "Multi-scales region segmentation for ROI separation in digital mammograms", Proc. SPIE 10225, Eighth International Conference on Graphic and Image Processing (ICGIP 2016), 1022515 (8 February 2017); doi: 10.1117/12.2267046; https://doi.org/10.1117/12.2267046
PROCEEDINGS
5 PAGES


SHARE
Back to Top