Computer Aided Breast Cancer Detection and Classification using Optimal Deep Learning

Medical imaging is receiving considerable interest in the field of healthcare, involving breast cancer (BC). BC is one of the cancer-related death amongst women around the world. At present, analysis of histopathology image is the clinical gold standard in cancer diagnoses. But the histopathological analysis of BC is labor-intensive, non-trivial, and might results in a higher degree of disagreement amongst pathologists. Thus, an automated diagnosis method could help pathologists to enhance the efficiency of diagnosis process. Extracting features for BC classification is challenging, owing to the nature of large intraclass and small inter-class variances in BC histopathological image (H1). This study develops an automated breast cancer classification using deep learning with aquila optimizer (ABCC-DLAO) technique on histopathological images. The proposed ABCC-DLAO technique intends to identify and classify the BC on Hli. To accomplish this, the presented ABCC-DLAO technique employs contrast enhancement approach as a preprocessing step. Next, the ABCC-DLAO technique exploits EfficientNet model as feature extractor and the hyperparameter tuning process take place by AO algorithm. For BC classification, Elman neural network (ENN) model is utilized. The experimental assessment of the ABCC-DLAO technique take place on Breast Cancer Histopathological Database (BreakHis) dataset. The simulation outcomes reported that the ABCC-DLAO technique is proficient in automated classification of HIs for BC diagnosis.