Main Article Content
Breast cancer is the world's second leading malignant tumor, which is susceptible toboth men and women,butit'sfarmorecommoninwomen.Multipledatamining, machinelearning and deep learningtechniques were developed andapplied for classification. Despite significant improvements in medicine, pathological image analysis remains the most common method for diagnosing breast cancer.In pathological image analysis, Computer-Aided Diagnosis (CAD) is commonly used to assist pathologists improve diagnosis efficiency, accuracy, and consistency.Deep learning strategies have been investigated in recent studies to boost the effectiveness of pathological CAD. In this proposed work, an optimized deep ResNet structure is employed to extract more affluent and finercharacteristics from clinical images. The proposed technique is evaluated on the publicly availableBreakHis dataset. The results indicated that the proposed model in all the magnification levels outperforms the baseline techniques significantly.