The field of medical imaging and analysis is rapidly advancing with the application of deep learning techniques. Recent developments have shown significant improvements in image segmentation, classification, and detection of various diseases. One notable direction is the use of transformer-based architectures, such as the Vision Transformer, for image classification tasks. Additionally, the integration of skeletal representations and contrastive learning has enhanced the performance of models in tasks like yoga pose classification and electrocardiogram analysis. Noteworthy papers in this area include the introduction of SAM3-UNet, a simplified variant of the Segment Anything Model 3, which achieves state-of-the-art performance in image segmentation tasks while requiring less computational resources. Another significant contribution is the development of CLEF, a clinically-guided contrastive learning approach for electrocardiogram foundation models, which demonstrates improved performance in cardiac abnormality detection. Overall, these advances have the potential to improve disease diagnosis, patient outcomes, and healthcare efficiency.