Could Big Data really help Predict a Heart Attack?

 In Big Data

AHA-scientific-sessions-2014Our CTO, Craig G. Rusin, PhD, recently presented his research at the American Heart Association Scientific Sessions national meeting in Chicago, IL. More than 17,000 cardiovascular experts from more than 100 countries met to discuss basic, translational, clinical and population science with an additional 1.5 million professionals attending virtually. The five-day meeting included presentations from the world’s leaders in cardiovascular disease, 14 of which were given or moderated by Texas Children’s Heart Center experts.

Craig presented on the “Prediction of Imminent Deterioration of Children after Stage 1 Palliation Using Real-Time Processing of Physiological Data”. Using his research and the technology developed at MIC, he has developed the technology of predicting imminent cardiac arrests for children with certain forms of congenital heart disease.

At Texas Children’s Hospital, Sickbay® captures high-resolution physiological data from bedside monitors and other ancillary devices. Research scientists then develop algorithms for specific clinical conditions using the Sickbay research interface. This interface allows them to rapidly prototype algorithms and assess their performance across thousands of patients.

This work demonstrates that predicting the onset of critical deterioration is actually possible by leveraging the data measured at the patient bedside combined with big data analytics. Our Sickbay® platform is the engine that can both enable the development of such real-time decision support algorithms as well as deploy them for effective clinical use.

This algorithm, currently being used in research and development, is on the leading edge in clinical analytics. It hopefully gets those in the medical profession thinking about what other possibilities exist around the development of algorithms applied to continuous physiological patient data.

Can you imagine what the impact would be at your institution if you could predict the onset of cardiac arrests 1-2 hours before they occur?

 

Facebooktwitterlinkedinmail