A Systematic Review on Learning Based Models for Automatic Classification of Medical Documents

In healthcare domain, there is need for maintaining medical documents that contains information pertaining to patients, treatment procedures and even drug information. Such medical documents are very important to have planning and strategies for improving healthcare services. Classifying medical documents automatically is an important research area which enables healthcare professionals to perform triage, treatment planning and statistical analysis. Traditional approaches are heuristic in nature and they rely on certain procedure to deal with medical document classification. With the Artificial Intelligence (AI) in place, it is possible to have learning based approaches that are dynamic in nature in understanding medical documents and classify them. In this paper, we made a systematic review which provides valuable insights on existing learning based models used for medical document classification. PRISMA model is used for literature review which covers from 2018 to 2023. The research gaps identified in this paper could trigger further research on medical document classification.