DATAx presents: How deep learning is impacting the world in 2019 Articles Big Data

Another simple, yet incredibly useful way deep learning is impacting healthcare is through the categorization of electronic health records (EHRs). Deep learning is already used in many natural language processing (NLP) programs, but EHRs present a uniquely complex problem, even for deep learning networks. Free text notes are often completed in a rush meaning they are messy, full of medical jargon, sometimes incomplete or filled out by multiple people, rendering them inconsistent. However, deep learning still represents the best method we currently have to analyze these records. A deep learning model developed by Google in 2018 was capable of predicting clinical outcomes, such as mortality and unexpected readmissions, better than traditional models after it had analyzed 216,000 patient EHRs across two hospitals.