With the rapid advancement of artificial intelligence technology, computer vision and deep learning have revolutionized medical image analysis. This review systematically outlines the latest developments, key challenges, and future directions of deep learning techniques in medical image analysis. First, we introduce the applicability and improvement strategies of foundational models—including convolutional neural networks (CNNs), generative adversarial networks (GANs), and Transformers—in medical imaging. Subsequently, we conduct a detailed analysis of the practical applications and performance of these technologies in core tasks such as disease classification, image segmentation, super-resolution reconstruction, and diagnostic assistance. Our findings reveal that CNN-based models achieve up to 98.7% accuracy in lung nodule detection, while GAN-augmented datasets enhance the specificity of Alzheimer's disease diagnostic models by 12.3%.