Michigan Medicine makes AI, machine learning a top tech priority
The academic medical center of the University of Michigan is leveraging investments in artificial intelligence, machine learning and advanced analytics to unlock the value of its health data.
According to Andrew Rosenberg, MD, chief information officer for Michigan Medicine, the organization currently has 34 ongoing AI and machine leaning projects, 28 of which have principal investigators.
“There’s a lot of collaboration around these projects—as there should be for the diversity of thought and background needed to deal with complex problems—working with at least seven other U of M schools,” Rosenberg told the Machine Learning for Health Care conference on Friday in Ann Arbor, Mich. “That’s one of the powers that we enjoy.”
One of the machine learning projects cited by Rosenberg leverages a combination of electronic health records, monitor data and analytics to predict acute hemodynamic instability—when blood flow drops and deprives the body of oxygen—which is one of the most common causes of death for critically ill or injured patients.
A team at the Michigan Center for Integrative Research in Critical Care developed an automated computer algorithm that utilizes data from a single lead of a non-invasive electrocardiograph signal for analysis and early identification of hemodynamic decline.
Rosenberg contends that the technology—called Analytic for Hemodynamic Instability—is a novel application of continuous nonlinear pattern recognition through the analysis of heart rate variability, which can detect signs of hemodynamic decompensation prior to overt changes in vital signs over very short periods of time.
Another project highlighted by Rosenberg targets clostridium difficile, an aggressive gut-infecting bacteria that is resistant to many common antibiotics. Michigan researchers teamed with investigators from Massachusetts General Hospital and the Massachusetts Institute of Technology to develop machine learning models that can predict a patient’s risk of developing C. difficile much earlier than can be diagnosed with current methods.
“It’s an extremely resistant-to-cleaning type of infection that is in some ways very preventable, but when someone gets it, there’s a really high morbidity and even mortality,” said Rosenberg. The good news, he added, is that “there’s enough features” in Michigan’s clinical data that are amenable to the modeling.
In this big data approach, analysis of EHRs is able to predict a patient’s C. difficile risk throughout the course of hospitalization. In addition, the models are specifically tailored to individual institutions to accommodate different EHR systms and patient populations.