Hybrid Deep Learning for Breast Cancer Diagnosis: Evaluating CNN and ANN on BreakHis_v1_400X

Recent medical research has significantly emphasized the need for early illness diagnosis due to the fast rise of the world's population. Since there are now more people in the world, the danger of dying from breast cancer is rising at an exponential rate. Breast cancer is currently the second deadliest form of cancer out of all the forms that are known. When it comes to illness diagnosis, having access to an automated disease detection system is very helpful to medical professionals since it enables them to provide dependable, effective, and quick reactions while also lowering the mortality rate. This research assesses the hybridization of two deep learning models, a convolutional neural network (CNN) and an artificial neural network (ANN)—the well-known BreakHis_v1_400X dataset (a dataset for breast cancer histopathological images) to work with. The performance of these approaches is evaluated based on several criteria, including accuracy, sensitivity, specificity, precision, false-positive rate, and F1 score. In addition, we evaluate these methods by calculating the area under the curve for accuracy and recall, as well as the receiver operating characteristic curve. According to the findings, the hybridized CNN-ANN achieved the most remarkable accuracy, precision, AUC, and F1 score, with respective values of 89.47%, 86.18%, 89.46%, and 89.08%.