Identification of Hypertrophic Cardiomyopathy on Electrocardiographic Images with Deep Learning

Figures

• 

  Figure 1.. Model Development and Study Design. Abbreviations: CMR, Cardiac Magnetic Resonance Imaging; HCM, Hypertrophic Cardiomyopathy; IVSd, End-diastolic Interventricular Septal Thickness

• 

  Figure 2.. Model performance measures (A) Receiver operating characteristic curves across image formats in the held-out test set. B) Diagnostic odds ratios across age, gender, and race subgroups on standard format images in the held-out test set. Abbreviations: AUROC, area under receiver-operating characteristic curve. The box plots display the odds ratio, with the upper whiskers extending to 95th percentile of the odds ratio, and the lower whiskers extending to the 5th percentile of the odds ratio.* *Note: Diagnostic odds ratios were not available for Hispanic, Asian, or other Races because there were no FN screens in the held-out test set

• 

  Figure 3.. Gradient-weighted Class Activation Mapping (Grad-CAMs) across Electrocardiogram formats. A) Standard format B) Two rhythm leads C) Standard shuffled format D) Alternate format. The heatmaps represent averages of the 25 positive cases with the most confident model predictions for HCM.

• 

  Figure 4.. Receiver-Operating Curves for external validation sites Abbreviations: AUROC, area under receiver-operating characteristic curve; AUMC, Amsterdam University Medical Center; UKB, UK Biobank.

• 

  Figure 5.. (A) Distribution of interventricular septal thickness in Yale New Haven Hospital patients without hypertrophic cardiomyopathy. (B). Distribution of left ventricular mass index in UK Biobank without hypertrophic cardiomyopathy.