December 4, 2025Open Access

Graph Neural Networks in Medical Imaging: Methods, Applications and Future Directions

Key Points

Segmentation and classification were enhanced with deep learning techniques in medical imaging, indicating significant performance gains.
Key findings include the effective use of Graph Neural Networks for reconstruction tasks and the integration of advanced architectures.
Assessment of various GNN architectures revealed challenges like model generalization and real-world application constraints in medical imaging.
Current trends call for investigations into self-supervised graph learning and federated learning to improve medical imaging outcomes.

Abstract

Graph neural networks (GNNs) extend deep learning to non-Euclidean domains, offering a robust framework for modeling the spatial, structural, and functional relationships inherent in medical imaging. This paper reviews recent progress in GNN architectures, including recurrent, convolutional, attention-based, autoencoding, and spatiotemporal designs, and examines how these models have been applied to core medical imaging tasks, such as segmentation, classification, registration, reconstruction, and multimodal fusion. The review further identifies current challenges and limitations in applying GNNs to medical imaging and discusses emerging trends, including graph–transformer integration, self-supervised graph learning, and federated GNNs. This paper provides a concise and comprehensive reference for advancing reliable and generalizable GNN-based medical imaging systems.

Read Full Paperexternally

Bookmark

View Full Paper

Cite This Study

Mienye et al. (Mon,) studied this question.

synapsesocial.com/papers/6930e8c6ea1aef094cca35f0 https://doi.org/https://doi.org/10.3390/info16121051

Bookmark

View Full Paper