Predicting Clinical Re-admission using Discharge Summaries (PCRUDS)

Before days in hospitals used to prescribe medicines in paper format. As the technology increasing in most of the fields to make the process easy handling & robust. The hospital clinical notes are digitalized in electronic medical/health record (EMR) format. This notes which is being very helpful for analyzing for future predictions. The data is created by combing the covid-19 issues and MIMIC-I to predict the re-admission cases through NLP (Natural language processing) techniques. In this paper implemented covid-19 possibility of re-admitting due to some symptoms like fever, cold etc. in histogram form. Used MIMIC-III database which consists of free text in unstructured form and converting in to structured form using natural language processing (NLP) technique and python flask to make a web-app which identifies important keywords and probable diagnosis from the data.