Automated arrhythmia detection using novel hexadecimal local pattern and multilevel wavelet transform with ECG signals

A Multitier Deep Learning Model for Arrhythmia Detection

“…Recently, focus on ECG rhythm (ECGr) classification has similarly been on the increase. ECGr classification can be grouped into areas that focus on finding effective extraction methods, [13,14] improving classification outcomes, [15][16][17][18][19] and utilization of deep learning methods to enhance the performance of classification [20][21][22][23][24][25][26][27][28][29].…”

“…As reported therein, their strategy yielded an average classification accuracy of 99%. Tuncer et al [19], presented an automated method for arrhythmia detection using hexadecimal local pattern (HLP) and discrete wavelet transform (DWT). For classification, they used 1 nearest neighborhood (1NN) classifier and obtained an accuracy of 99.7%.…”

“…However, all these ML methods [15][16][17][18][19] reported good performances, when faced with a big data of ECG records their results might be less effective. This is attributed to the inherent shortcoming attributed to training ML models on limited or small data sets.…”

“…• The focus of the majority of these papers on designing a method for preprocessing of ECG signals, feature extraction, feature selection and classification, as in [15][16][17][18][19]. In this paper, only DNN model optimized by genetic algorithm (GA) is proposed.…”

“…In this paper, only DNN model optimized by genetic algorithm (GA) is proposed. • The majority of these works used small datasets, hence the result not robust as in [17,18,19]. In addition, several methods obtained low accuracy on big data, which is not suitable for real applications as in [16,17,19].…”