Skip to main content

Extracting meaningful radiology insights from natural language using Amazon Comprehend Medical (Level 300)

The insights needed to optimize the use of scarce and in-demand clinical resources are often hidden plain sight within unstructured clinical reports. This session explains how to integrate machine learning and analytics technology into their applications and automate processes to optimize the use of clinical resources. We show the use of near-real-time Apache Spark pipeline, with Amazon Comprehend Medical, to capture radiology examinations as they are added to the hospital's clinical data repositories. Learn how to classify natural language clinical notes, and translate the clinical entities into relational views built on standard SNOMED Clinical Terminology. We conclude by showcasing how general-purpose visualization and analysis tools can enable your users to access the data insights. Download slides »
Speaker: Craig Roach, Principal Solutions Architect, AWS
Duration: 30mins