The field of medical image analysis is rapidly advancing, with a focus on developing innovative methods for image denoising, segmentation, and classification. Recent research has explored the use of deep learning techniques, such as U-Nets and transformers, to improve image fidelity and accuracy. Additionally, there is a growing interest in unsupervised learning methods, which can help reduce the need for manual labeling and improve the scalability of medical image analysis. Another key area of research is the development of explainable models, which can provide insights into the decision-making process and increase clinical trust. Notable papers in this area include: PC-UNet, which proposes a Poisson Consistent U-Net model for Positron Emission Tomography denoising, and DB-FGA-Net, which achieves state-of-the-art performance in brain tumor classification without data augmentation. Overall, these advances have the potential to significantly improve the accuracy and reliability of medical image analysis, leading to better patient outcomes and more effective treatment strategies.