The field of image analysis is witnessing a significant shift towards the adoption of graph-based methods, which are proving to be highly effective in capturing complex relationships and structures within images. These methods are being applied to a wide range of tasks, including image classification, object detection, and segmentation, and are showing promising results. The use of graph neural networks, in particular, is allowing for the incorporation of prior knowledge and constraints into the analysis process, leading to improved accuracy and robustness. Notably, the integration of graph-based methods with other techniques, such as self-supervised learning and semi-supervised learning, is enabling the development of more efficient and effective image analysis pipelines. Some noteworthy papers in this area include: Morphology-Aware KOA Classification, which proposes a novel framework for combining anatomical structure with radiographic features for knee osteoarthritis diagnosis. Beyond Augmentation, which introduces a graph-based mechanism for self-supervised representation learning that captures inter-instance relationships. SemiETPicker, which presents a fast and label-efficient semi-supervised framework for particle picking in CryoET tomography. FastJAM, which leverages graph-based methods for rapid joint alignment of images. iPac, which introduces a new graph representation of images for medical image classification. A Re-node Self-training Approach, which proposes a graph-based semi-supervised learning approach for multi-view image data.