"Medical Transcriptions and UMLS-Based Disease Inference and Risk Assessment Using Machine Learning"
This research was conducted by Thamizharuvi Arikrishnan and S. Swamynathan. It was published in the Advances in Intelligent Systems and Computing book series. The study proposes a system for disease inference by extracting symptoms and mapping the metadata using the Unified Medical Language System (UMLS) to have the disease code. The symptoms and metadata are extracted from medical transcripts using natural language processing (NLP), and the extracted information has been categorized into chief complaints, present illness, and past history. These extracted symptoms are mapped using UMLS for disease code inference. Based on the disease code, the risk classifier classifies the level of risk. The proposed system based on deep learning and UMLS for disease inference resulted in a significant improvement in accuracy.
.svg)