Thomas Heldt, core faculty at MIT’s Institute for Medical Engineering & Science (IMES) and former MIT Hugh Hampton Young Fellow, believes in the power of patient data—the physiological measurements collected in intensive care units, operating rooms and emergency rooms—from medical devices logging hundreds of samples per second.
“What if we didn’t throw away all that data?” Heldt asks. He and his research colleagues at IMES analyze patient data in an effort to help clinicians deepen and personalize patient care, and potentially alert them to crises. “What if we could move away from a care paradigm that is reactive . . . to one that is predictive?” Yet the first hurdle, even in the age of big data, is harnessing the information.
“The infrastructure in the hospital was never set up to keep that data…. (Plus) devices from different vendors don’t really communicate. It’s very difficult to get data on a common time axis, so you can see what happened with that particular patient.” Partnerships with medical device manufacturers are extremely helpful. And when that doesn’t work, there is always “hacking,” Heldt says. Yet getting at the data is only the beginning. To move beyond hypotheses, Heldt’s group uses mathematical modeling and model-based data integration. Read the full article at IMES