Medical text classification based on convolutional neural network: A review

Medical text classification has a significant impact on disease diagnosis, medical research, and the automatic development of disease ontology, acquiring knowledge of clinical results recorded in the medical literature. Hence, medical text classification is challenging because it contains terminologies that describe medical concepts and terminologies. Furthermore, the medical data mostly does not follow natural language grammar; it has inadequate grammatical sentences. The techniques used for text classification give different results comparing to medical text classifications, as extracting text and training sets are different. One of the most significant text classification models in general and medical text classification specifically is CNN-based models. In this paper, many papers on medical text classification have been reviewed, and the details of each article, such as algorithms, or approaches used, databases, classification techniques, and outcomes obtained, are evaluated and outlined thoroughly. Besides, discussions were carried out on all the studied papers, which profoundly influenced medical documents classification.