Table 1.

Summary of literature review.

Author	Models	Disease	Datasets	Accuracy
Altan et al. [8]	Deep belief networks	Coronary artery disease	Made a dataset from collecting data	98.88%
Ali et al. [17]	CNN, LSTM, RNN	Arrythmia classification	Combination of different publicly available datasets	-
Naz et al. [18]	Pretrained CNNs	ECG classification	MIT-BIH database	91.2
Wu et al. [19]	Convolutional neural networks	Arrhythmia	MIT-BIH database	97.41
Patro et al. [21]	Artificial neural network	Feature extraction from ECG signals.	MIT-BIH ECG ID database signal	-
Acharya et al. [23]	Gaussian Mixture Model (GMM)	Coronary artery disease	The CAD datasets from the University California Irvine a database	95%
Acharya et al. [24]	Convolution neural network	Coronary artery disease	Physio net databases	95.11%
Bhyri et al. [25]		heart diseases	CSE ECG database	around 99%
Lin et al. [26]	Deep convolutional neural networks	coronary artery disease	Combination of datasets	95%
Akella et al. [27]	SVM, K-NN, artificial neural network	coronary artery disease	UCI dataset	93.03%
Yıldırım et al. [29]	16-layer standard CNN	Arrhythmia	MIT-BIH Arrhythmia database	86.67%
Luz et al. [30]		Arrhythmia	MIT-BIH, EDB, AHA, CU, NST databases	-
Gayathri et al. [31]	Relevance vector machine	Arrhythmia	MIT/BIH database	RVM boosts generalization capability
Rajpurkar et al. [32]	34-layer convolutional neural network	Arrhythmia	Own dataset with a combination of datasets
Li et al. [33]	CNN-based classification on ECG signals.	ECG classification	MIT-BIH arrhythmia database,	99.1%
Avanzato et al. [34]	Convolutional neural networks	coronary artery disease	MIT-BIH arrhythmia database	98.33%
Alizadehsani et al. [35]	ML algorithms	Coronary artery disease	Combination of different datasets	-
Acharya et al. [36]	11-layer deep convolutional neural network	congestive heart failure	BIDMC: Congestive Heart Failure Database, Fantasia Database, MIT-BIH database	99.01%
Acharya et al. [37]	Time level and frequency domain analysis	Coronary artery disease	CAD dataset	96.8