Download or read book Classification of Mammographic Images Using Support Vector Machine written by Amirali Asgari and published by . This book was released on 2020 with total page 50 pages. Available in PDF, EPUB and Kindle. Book excerpt: Breast cancer today is the leading cause of death worldwide. In developed countries, it is the most common type of cancer in women, and is the second or third common malignancy in developing countries. In this study, an automatic diagnostic algorithm for breast cancer presents mammographic images based on features extracted from the GLCM, local binary patterns, and zernic moment and fusion in the intelligent classifiers. For this purpose, a data set of mammogram images from the database is extracted in two healthy and cancerous classes. The images are subjected to segmentation (fuzzy, thresholding) after the preprocessing, so that the desired area can be obtained. The zoned images are considered as inputs of a feature extraction block. In this block, the proposed algorithm consists of three types of attributes extracted from the coincidence matrix, local binary patterns, and Zernik Moment. The optimal features of the feature selection methods (such as UTA or statistical methods) and subsequent diminishing methods (such as principal component analysis and linear differential analysis) are selected and reduced later. Characteristics are considered as inputs of linear classification structures (such as backup machines) and non-linear (nerve networks), and in the next step, fusion methods at the class level (such as bagging or boosting or Other innovative methods will be considered for the implementation of a council machine from weak floors, and the output of the classification class will be a healthy or cancerous label. The results of the classification of linear and nonlinear methods with the combined structure of the Soviet machine for the various characteristics and the characteristics of reduced and selected dimension by comparing the classification indices (accuracy, sensitivity and specificity index), and the optimal structure of the choice Gets The results of this study showed that the combination at the level of the classifier provides a more than 90% mean acc