News Report | July 22, 2018
Google AI Predicts Hospital Inpatient Death Risks With 94% Accuracy
Google's artificial intelligence (GAI) network is 94% accurate at predicting hospital inpatient death risks. This is a 9% increase in accuracy over a traditional predictive model (85% accurate).
These findings were presented in “Scalable And Accurate Deep Learning With Electronic Health Records”, by Alvin Rajkomar, Eyal Oren, and colleagues. The researchers developed a mortality logistic algorithm for computational analysis that included variables from the National Early Warning Score (NEWS) (six physiological findings and one observation: respiratory rate, oxygen saturations, supplemental oxygen, temperature, systolic blood pressure, and heart rate). Additional variables were added from 24 common lab tests, like the . . .