Retrospective validation of a machine learning clinical decision support tool for myocardial infarction risk stratification

Figures

• 

  FIGURE 1. Patient encounters used to train and test a machine learning algorithm to predict myocardial infarction based on electronic health data available within the first 3 h.

• 

  FIGURE 2. Area under receiving operating characteristic curves and clinical operating points for (A) machine learning and GRACE clinical prediction model of myocardial infarction diagnosis and (B) machine learning and TIMI clinical prediction model of myocardial infarction diagnosis.

• 

  FIGURE 3. Top unique feature correlations and distribution of feature importance for each patient encounter for machine learning models. Input variables are ranked in descending order of feature importance. Red indicates a high feature value and blue indicates a low feature value. Points to the right and left sides of the line of neutral contribution resulted in higher and lower prediction scores, respectively. Abbreviations: alanine aminotransferase (ALT); blood pressure (BP); heart rate (HR); international normalised ratio (INR); myocardial infarction (MI); peripheral oxygen saturation (SpO2).