“University of Nottingham epidemiologist Stephen Weng and his colleagues have created an algorithm that can outperform standard methods at predicting heart attacks. The team tested four AI algorithms (random forest, logistic regresion, gradient boosting, and neural networks) against the American College of Cardiology/American Heart Association (ACC/AHA) guidelines for predicting a patient’s risk of experiencing a cardiovascular event in the next decade. All four algorithms outperformed these standard guidelines, with neural networks correctly predicting 7.6% better than the ACC/AHA method. Interestingly, several factors the algorithms identified as strong predictors were not on the list of the ACC/AHA guidelines (i.e. oral corticosteroid use), while top 10 risk factors under the ACC/AHA like diabetes were not considered…”
#future = #REALnews #health #medicine #wellness #tech #innovation #science #design #biotech #biology #singularity #engineering #ai #artificialintelligence #robots #automation
