Color and texture feature fusion for breast cancer classification using histopathology images
Dipti Deb, Ratnakar Dash, Durga Prasad Mohapatra · 2023
Recent advancements in the fields of artificial intelligence and computer vision have made it possible to create computer-aided diagnosis (CAD) systems to detect breast cancer using histopathology images. In this research, a classification framework for breast cancer has been developed using histopathology images. In this regard, the color and texture features of the images are utilized for classification using several machine learning (ML) classifiers. For texture, three handcrafted feature extraction techniques, such as Gray-Level Co-occurrence Matrix (GLCM), Local Binary Pattern (LBP), and Gabor have been investigated. Similarly, for color features, HSV histogram features and H, S, and V values of the HSV color model are used to classify the images. The features extracted from each method are evaluated individually. The best derived methods are combined and feed into different ML classifiers. The performance of each classifier is evaluated using performance metrics such as accuracy, precision, specificity, sensitivity, and F1-score. It is observed that LBP and HSV histogram with XGB classifier obtained an accuracy of 87.50% and 83.62%, respectively. After fusing the LBP and HSV histogram features with the XGB classifier obtained an accuracy of 93.34%. The experiment is carried out on the BreakHis dataset. The proposed methodology is useful in classifying breast cancer histopathology images with benign and malignant.