Automated Feature Extraction techniques for Histopathology Images with Machine Learning based Classification Techniques

Cancer is a leading cause of death worldwide. It is the primary reason why research in this area is difficult. Pathologists most commonly use histopathological images to detect cancer. The shape, morphology, intensity, and texture of histopathology images can be used to classify the cancer status. Because of the large amount of data, using full high-resolution histopathology images will require more time to extract all information. The classification results are compared in this study using different feature extraction algorithms capable of extracting various features from histopathological image texture. The successful feature extraction algorithms GLCM, LBP, PCA, and DCP have been chosen for this study. These features are classified using NB, ANN, DT, and K-NN Classifiers. The most successful feature extraction algorithm and classification algorithm for histopathological images are determined.