Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future

• Published in: Sensors (Basel, Switzerland) Switzerland), 2021-07, Vol.21 (14), p.4758
• Main Authors: Ahmedt-Aristizabal, David, Armin, Mohammad Ali, Denman, Simon, Fookes, Clinton, Petersson, Lars
• Format: Article
• Language: English
• Subjects: anatomical structure analysis; Attention; Automation; Brain; brain functional connectivity; Deep Learning; Electrical junctions; graph convolutional networks; Graph neural networks; graph representation; Graph representations; Health care; Machine Learning; Magnetic resonance imaging; Medical diagnosis; Neural networks; Neural Networks, Computer; Review; Semantics; Signal processing
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s21144758

With the advances of data-driven machine learning research, a wide variety of prediction problems have been tackled. It has become critical to explore how machine learning and specifically deep learning methods can be exploited to analyse healthcare data. A major limitation of existing methods has been the focus on grid-like data; however, the structure of physiological recordings are often irregular and unordered, which makes it difficult to conceptualise them as a matrix. As such, graph neural networks have attracted significant attention by exploiting implicit information that resides in a biological system, with interacting nodes connected by edges whose weights can be determined by either temporal associations or anatomical junctions. In this survey, we thoroughly review the different types of graph architectures and their applications in healthcare. We provide an overview of these methods in a systematic manner, organized by their domain of application including functional connectivity, anatomical structure, and electrical-based analysis. We also outline the limitations of existing techniques and discuss potential directions for future research.