Application of graph theory in liver research: A review

Abstract

Abstract Graph theory has emerged as a valuable tool in liver research, aiding in the assessment of complex interactions underlying liver diseases at different organizational levels. This has allowed significant advancements in the detection, treatment, and control of liver disorders. Particularly, graph theory is useful in identifying different liver diseases. Graph theory can be used to analyze liver networks and identify altered nodes and edges, which may serve as potential noninvasive biomarkers for disease detection. Furthermore, graph‐based techniques, including graph neural networks and graph knowledge, have been employed to construct interaction networks, providing insights into the communication involved in focal liver diseases and related conditions such as coronavirus disease 2019 (COVID‐19), hepatic muscular atrophy, and hepatic necrosis. Functional networks for the liver have also been developed with graph‐based methods, providing insights into the metabolic processes occurring in the liver and the functional organization of the liver. Graph theory is also useful for image analysis, with applications such as image segmentation, registration, synthesis, and object identification. By representing images as graphs, it is possible to analyze and process them with graph‐based algorithms, handling complex relationships among pixels and making them useful in boundary extraction and texture analysis. Overall, graph theory is an essential research tool for liver research, providing valuable insights into the complex interactions underlying liver diseases as well as innovative approaches for diagnosis and treatment.

Keywords

• disease
• graph knowledge
• graph theory
• liver
• neural network