The very early typing machine at the top of this article represented a significant advance for the written word. Without pushing the comparison too far, Amazon Web Services (AWS) recently introduced a natural language processing service developed to transform unstructured medical information into usable formats for clinical care support, financial and clinical trial management, and data privacy.
Amazon Comprehend Medical‘s core purpose is “to improve patient outcomes through technology.” The HIPAA-eligible machine learning service can glean relevant information directly from medical notes, prescriptions, audio transcripts, and lab reports according to AWS. The Seattle-based Fred Hutchingson Cancer Research Center, for example, used Amazon Comprehend Medical to process millions of clinical notes to extract and index data on diagnosis, medications, and treatments. The process took just seconds for each document instead of hours.
We have covered a range of machine learning applications in health technology from various institutions using AI analytical tools, from identifying suicidal adolescents to predicting pain level to uncovering hidden insights about diabetes. Machine learning generally involves teaching the machine highly structured rules to analyze data or running huge datasets for the machine to analyze to build its own rules. AWS’s Comprehend Medical service, which processes data at a per-item fee, enables developer customers to proceed quickly from raw data to actionable insight and analysis without writing and maintaining their own set of rules. AWS owns the service, but the customers own the data and the resulting output. Amazon Comprehend Medical service offers medical researchers, clinicians, and financial administrators a means to obtain useful information in a standard format quickly and efficiently.
