Researchers create repository of waveform and health data to drive AI innovation

Editor's Note: 

Researchers from UCLA and UC Irvine have created the first comprehensive online repository of physiological waveform and health record data from surgeries which is intended to serve as a resource for training and testing AI algorithms. The findings were published in JAMIA Open on October 17. 

In development since 2012, the Medical Informatics Operating Room Vitals and Events Repository (MOVER) contains data from 59,000 adult patients who underwent 83,500 surgeries at the University of California, Irvine Medical Center from 2015 to 2022. The data is drawn from physiological waveforms which have then been matched with patients’ EMR data. Waveforms involve data from monitors (such as EKGs) that measure a patient’s physiology during surgical procedures.

Prior to the development of this repository, physician-scientists have not had access to a single repository where they can access a large volume of data that includes surgical patients’ physiological waveforms. Beyond sharing the resource in the short term, a larger focus is to ultimately create a single repository across multiple institutions under the National Institutes of Health initiative called Bridge to AI in order to develop more precise AI tools to improve surgical care. 

Although the data includes general information about patients' medical history, including details about procedures, medicines, and postoperative complications, it has been stripped of all patient identifiers to protect patient privacy and comply with privacy laws.