Breast Mass Detection and Segmentation Using Multi-scale Morphological Sifting and K-Means Clustering

A mass detection and segmentation approach that consists of five main steps for region segmentation, which are pre-processing, region candidate generation using multi-scale morphological sifting, mean shift filtering, k-means clustering, and finally post-processing. ucasML tool was used for obtaining a classification score using 10-fold cross validation for each region candidate for a binary classification task as it is robust to class imbalance. This project also includes feature extraction process and evaluation methods used.

You can find the source code in the following Github repository.