This research looks at the application of Convolutional Neural Networks (CNNs) for breast cancer diagnosis with special reference to Invasive Ductal Carcinoma (IDC) that is the most prevalent type of breast cancer. Using a dataset of the image patches sourced from the Kaggle repository, the work of the research deploys the enhanced CNN models while contrasting their results to the typical machine learning classifiers, including Logistic Regression, KNN, and SVM. The findings highlight the fact that CNNs perform much better than all these traditional approaches in diagnosing IDC while minimizing on errors that may be made by doctors. This work therefore highlights the possibility of using these CNNs to improve the medical image analysis and hence improve diagnosis of diseases to improve the patient’s quality of life and treatment outcomes.