Arrhythmia Classification Techniques Using Deep Neural Network

Table 3

Top arrhythmia classification techniques.

Ref.	Publication year	Database	Classification techniques	Optimization	Accuracy (%)
[39]	2018	PTB	FCN	MFB-CNN	99.95
[40]	2018	PTB	CNN	CNN	100
[41]	2018	MITDB	CNN	—	84
[42]	2018	PhysioBank	CNN	—	84
[43]	2018	PhysioNet challenge	FCN	FFT	81
[44]	2018	MITDB AFIB	DNN	Fourier transform	98.34
[45]	2018	PhysioBank	CNN	—	96.36
		PhysioNet challenge
[46]	2018	PhysioNet challenge	MLP-CNN	—	76.79
[47]	2018	MITDB	CNN	—	98.6
[48]	2018	MITDB	AlexNet VGGNet	Transformation	99.05
[49]	2017	PhysioNet challenge	ResNet	—	72.1
[50]	2018	MITDB	FCN-CNN	—	91.33
[51]	2019	Multiple DB	CNN	—	95.98
[52]	2017	MITDB	FCN	1D-CNN	97.5
[53]	2017	MITDB PhysioNet 2000	RNN	CNN	94.03
[54]	2017	PAF	CNN	—	93.6
[55]	2019	Multiple	RNN	—	99.26
[56]	2018	MITDB	SVM	GD-DBM	99.5
[57]	2017	MITDB	RNN	LSTM	99.4
[58]	2017	MITDB	RNN	—	95