Deep learning has quickly evolved over the past several years from a promising to a feasible solution for medical image analysis. This is a great topic for research because more and more people are using medical imaging to treat and diagnose. A significant benefit of deep learning is the ability to use enormous volumes of data to eliminate the painstaking handcrafting of features, which needs strong domain expertise. This study discusses how convolutional neural networks (CNNs) are utilized in the field, including detection, classification, registration, segmentation, and picture enhancement. It also gives some important information about how CNNs can be used to analyze medical images of the brain, eye, breast, chest, and skin.