Case Study: Emory University predicts sepsis faster with Google Cloud for Education

Emory University predicts sepsis 4-6 hours early with Google for Education

Emory University researchers faced the challenge of predicting the onset of sepsis, a deadly and expensive condition, in intensive care patients. They needed a reliable way to analyze vast amounts of real-time patient data for early detection. To address this, they turned to Google for Education, utilizing Google Cloud and TensorFlow to build an AI engine.

The solution implemented on Google Cloud used an integrated set of tools, including App Engine, to process data and run a TensorFlow-based prediction algorithm in real time. This resulted in an 85% accuracy rate for predicting sepsis four to six hours before onset. By leveraging the scalable infrastructure from Google for Education, the team was able to abstract away deployment requirements and focus on improving their algorithm, with the potential to save lives and reduce medical costs.

Emory University