Table 13.

Performance comparison of the present work with respect to state-of-the-art methods for the AF: TD1, VF: TD2-A, ST: TD3, and VF: TD2-B data sets

	Accuracy	Sensitivity	F1 score	Specificity
proposed methodology (AF: TD1)	96.47	0.93	0.92	0.90
Chanthercrob et al. [39] (AF: TD1)	—	0.97	—	1.00
Gopika et al. [31] (AF: TD1)	96.47	0.93	0.92	0.90
proposed methodology (VF: TD2-A)	90.88	0.91	0.95	0.90
Boreiko [40] (VF: TD2-A)	—	0.83	—	—
Sohail et al. [41] (VF: TD2-A, VF: TD2-B)	98.03%	—	—	—
proposed methodology (ST: TD3)	27.97	0.28	0.44	0.24
Gopika et al. [31] MIT arrhythmia (N, S, V, F, Q)	97.94	0.98	—	—
proposed methodology (VF: TD2-B-lead I)	94.71	0.95	0.97	0.95
Proposed Methodology (VF: TD2-B-lead II)	94.18	0.94	0.97	0.94
Mohanty et al. [42] (VF: TD2-A/VF: TD2-B)	99.18	0.97	—	0.99
Andreotti et al. [24] (AF: TD1)	—	—	0.83	—
Hannun et al. [26] (AF: TD1)	—	—	0.837	—
Andersen et al. [25] (AF: TD1, VF: TD2, VT, AF, etc.)	—	0.86	—	0.98
Shashikumar et al. [27] (AF: TD1)	AUC-94%	—	—	—