According to our review, most studies applied knowledge graphs to specific topics such as disease detection, localization, and report generation.

To bridge this gap, we conducted a systematic review on knowledge graph applications in medical imaging analysis. Though there are comprehensive survey papers for knowledge graph applications in biomedical informatics, we have not seen any literature survey for knowledge graph applications in medical imaging analysis. In the medical imaging analysis domain, knowledge graphs have drawn a lot of research attention. The algorithm learns this with large amounts of data, and after analyzing thousands of iterations of different images and diagnoses, it eventually will learn to make some diagnoses. Medical record histories, previous diagnoses made by pathologists or radiologists, are used to train the algorithms. We hope that our article will provide the readers with aggregated documentation of the state-of-the-art knowledge graph applications for medical imaging to encourage future research.ĭeep learning methods try to train algorithms to identify abnormal regions and tissue variations in a manner similar to human beings. We further identified the potential future directions according to the identified limitations, including employing semisupervised frameworks to alleviate the need for annotated data and exploring task-agnostic models to provide better generalizability. We also identified limitations of current work, such as the limited amount of available annotated data and weak generalizability to other tasks. We looked at four applications in medical imaging analysis, including disease classification, disease localization and segmentation, report generation, and image retrieval.

After screening, evaluating, and reviewing the selected articles, we performed a systematic analysis. We systematically searched over five databases to find relevant articles that applied knowledge graphs to medical imaging analysis. Knowledge graphs are widely used to model prior knowledge in the form of nodes and edges to represent semantically connected knowledge entities, which several works have adopted into different medical imaging applications. There is an increasing trend to represent domain knowledge in structured graphs, which provide efficient knowledge representations for many downstream tasks.