Breast cancer is one of the leading causes to women mortality in the world. Clustered microcalcifications (MCs) in mammograms can be an important early sign of breast cancer, the detection is important to prevent and treat the disease. In this work, we present a novel method for the detection of MCs in mammograms which consists of regions of Interest (ROIs) segmentation, based on a spatial filter that allows the detection of small and large microcalcifications, clustering and classification of MCs by Artificial Neural Network. The system has been tested on a public dataset of digital images and compared with previous approaches. The results demonstrate that the proposed approach could achieve significantly higher FROC curves: our CAD system achieve a cluster-based sensitivity of 70, 80, and 90 % at 0.31, 0.69, and 1.6 FPs/image, respectively.
|Title of host publication||2018 IEEE Nuclear Science Symposium and Medical Imaging Conference, NSS/MIC 2018 - Proceedings|
|Number of pages||4|
|Publication status||Published - 2018|
All Science Journal Classification (ASJC) codes
- Radiology Nuclear Medicine and imaging
- Nuclear and High Energy Physics