Medical Text Annotation Tool based on IBM Watson Platform

A critical sum of critical data is often found, as it were, in the unstructured portion of patient accounts in Electronic Health Records (EHRs), making it hard to handle and use it for tasks such as evidence-based health care or medical research. The proposed method uses a novel tool for the annotation of medical text. This is based on IBM Watson Platform and uses the UMLS as a source of medical expressions and their classifications; the experiments are applied on 100 medical documents collected from the MIMIC data set. The results obtained are promising in this area of research as we obtained a precision and recall value 91 % for the tested data.