<p>Breast cancer is a well-known reason for mortality in women, underscoring the significance of early detection for saving lives. Enhancing survival rates hinges on timely identification. Nonetheless, conventional breast cancer detection methods like mammography possess shortcomings. They can prove time-intensive and uncomfortable for patients, and their efficacy may falter when detecting minute tumors or those concealed within the dense breast tissue. It has been demonstrated that artificial intelligence performs exceptionally well in image recognition tasks. Deep learning applications for breast cancer diagnosis have recently grown in popularity, which discern patterns within images that may elude human observation. With less chance of false positive results, this capability may improve the sensitivity and specificity of breast cancer diagnosis. A systematic literature review of new improvements in deep learning for detecting breast cancer is presented in this article. The review found that deep learning models can achieve accuracies up to 93.8%; this is far more accurate than what can be achieved using traditional approaches. This manuscript provides a comprehensive overview of breast cancer detection, covering its fundamental processes, different methods, associated challenges, and emerging trends. It also explores recent research that has applied deep-learning techniques to address issues related to breast cancer. Furthermore, the study evaluates the performance of various models on different datasets. This research offers a clear and concise summary of breast cancer detection techniques, including current challenges, ongoing issues, and future directions.</p>