The fifth annual PhysioNet/Computers in Cardiology Challenge focusses on this question: Is it possible to predict if (or when) an episode of atrial fibrillation will end spontaneously?
Results of fitted linear models, each adjusting for additional variable, with its β coefficients and 95% CI (corresponding to psychiatric disease presence), statistics (t and p values), and model fit
Results of multivariate Cox proportional hazard regression analysis for the pre-specified cut-off values of analyzed parameters as predictors of the EVENT during follow-up period.
