Dr. Ruan Chunyang & Dr. Luo Guangsheng published “Semantic-Aware Graph Convolutional Networks for Clinical Auxiliary Diagnosis and Treatment of Traditional Chinese Medicine”paper in SCI journal IEEE Access(Volume: 9,Page: 8797-8807, ISSN: 2169-3536, Digital Object Identifier: 10.1109/ACCESS.2020.3048932, Impact Factor:3.745, SCI Zone 1).
The paper starts from the demand of clinical aid diagnosis in TCM, constructs a training corpus from large-scale unstructured Chinese medical record data, and exploits the hot Graph Convolutional Network (GCN) theory, Self-attention Mechanism and the emerging heterogeneous graph Meta-graph. A feature extraction method is used to explore the hidden knowledge of prescriptions by mining the corpus for multiple relationships such as 'drug-sickness-disease'. Through quantitative and qualitative experimental analysis, the validity of the knowledge discovery results is demonstrated, which can assist physicians in clinical diagnosis.
