Artificial intelligence (AI) is applied across many disciplines including medical imaging: computer-aided diagnosis, automatic segmentation of anatomical structures, lesion detection, multi-modal data fusion, and image-guided interventions. This course introduces deep learning, a subset of AI, for biomedical engineering students who want to apply AI to imaging data (ultrasound, MRI, CT, PET, optical, and microscopy) but who may be new to programming. We start from the ground up: the first two weeks are a hands-on Python primer so every student, regardless of coding background, can build and train models by the end of the course. In this course we will evaluate and implement several deep learning architectures: convolutional neural networks, autoencoders and variational autoencoders, generative adversarial networks, U-Net segmentation networks, transformers and vision transformers, and diffusion models. We use PyTorch throughout, and we anchor each method in a real biomedical imaging problem.