Breast cancer is the most pervasive form of cancers in women, therefore automated algorithms for cancer detection, and analysis of hematoxylin and eosin stained breast-cancer histopathology images are being actively developed worldwide. In this paper, we propose multi-scale voting classifiers which operate on clinically-relevant features extracted from such images, and apply them to classify real-life breast-cancer data. The extensive experiments (encompassing cross-validation scenarios backed up with statistical tests) showed that our models deliver high-quality classification, can be learned quickly, and offer instant operation.
CITATION STYLE
Nalepa, J., Piechaczek, S., Myller, M., & Hrynczenko, K. (2019). Multi-scale Voting Classifiers for Breast-Cancer Histology Images. In Lecture Notes on Data Engineering and Communications Technologies (Vol. 23, pp. 526–534). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-319-98557-2_49
Mendeley helps you to discover research relevant for your work.