Table 2.

AI model performance.

	Precision	Recall (sensitivity)	F1-score*	Supporta
Healthy-NSR	0.78 (0.76–0.80)	0.64 (0.62–0.66)	0.70 (0.68–0.72)	2,096
PAF-NSR	0.69 (0.67–0.71)	0.82 (0.80–0.84)	0.75 (0.73–0.77)	2,074
Accuracy			0.73 (0.72–0.74)	4,170
Macro avg	0.74 (0.73–0.75)	0.73 (0.72–0.74)	0.73 (0.72–0.74)	4,671
Weighted avg	0.74 (0.73–0.75)	0.73 (0.72–0.74)	0.73 (0.72–0.74)	4,671

Data in parentheses represent the 95% confidence intervals. AI artificial intelligence; avg average; PAF paroxysmal atrial fibrillation.

*F1 Score (balanced F-score) is the harmonic mean of precision and recall, and calculated as follows: F1 score = 2 (precision × recall) / (precision + recall).

^aSupport is the number of occurrences of each class when it is true.