COMPUTER AIDED DIAGNOSIS MODEL FOR BREAST CANCER DIAGNOSIS AND CLASSIFICATION USING DEEP LEARNING TECHNIQUES

Breast Cancer is a commonly occurring deadly disease, which increases the mortality rate of women to a maximum extent. The breast cancer diagnosis and classification is not an easier task and consumes more time. Therefore, it is needed to design a computer aided diagnosis (CAD) model for the detection of breast cancer in the earlier stages using machine learning (ML) and deep learning (DL) techniques. This paper presents a new CAD model using DL for breast cancer diagnosis. The diagnosis and classification of breast cancer involve a set of steps namely preprocessing, segmentation, feature extraction, and classification. Primarily, wiener filter (WF) with contrast limited adaptive histogram equalization (CLAHE) is employed as a preprocessing step. In addition, the adaptive weighted segmentation (AWS) technique is applied for identifying the affected portions in the mammogram. Afterward, Inception with Residual Network-v2 is utilized for feature extraction. Finally, two classifiers namely Adaptive Boosting (Adaboost) and decision tree (DT) are applied namely IRAD and IR-DT to determine the appropriate class labels. The simulation process of the proposed model is validated using the benchmark dataset and the experimentation is compared with state of art models. The resultant outcome stated that the IR-AB model outperforms the existing techniques with the maximum accuracy of 94.38%