Fig. 3From: Classification of COVID-19 electrocardiograms by using hexaxial feature mapping and deep learningOne-way ANOVA results within a box plot for each related GLCM features. All normalized GLCM features obtained from binary ECG images were verified to be statistically significant relative to each other (\(p < 0.0001\) for all binary groups). Each group has a total of 4500 samples: 18-lead x 250 paper-based ECG reports (C: COVID-19, N: No-Findings)Back to article page