Extended Data Fig. 4: Diagnostic performance of CoDE- ACS scores at presentation in the derivation cohort across patient subgroups.
From: Machine learning for diagnosis of myocardial infarction using cardiac troponin concentrations
Data are presented as a central estimate with 95% confidence intervals based on the Clopper-Pearson method. (a) CoDE-ACS low probability score of less than 3. (b) CoDE-ACS high probability score of 61 or more.
