The Use of Time-Frequency Moments as Inputs of LSTM Network for ECG Signal Classification

This paper refers to the method of using the deep neural long-short-term memory (LSTM) network for the problem of electrocardiogram (ECG) signal classification. ECG signals contain a lot of subtle information analyzed by doctors to determine the type of heart dysfunction. Due to the large number of signal features that are difficult to identify, raw ECG data is usually not suitable for use in machine learning. The article presents how to transform individual ECG time series into spectral images for which two characteristics are determined, which are instantaneous frequency and spectral entropy. Feature extraction consists of converting the ECG signal into a series of spectral images using short-term Fourier transformation. Then the images were converted using Fourier transform again to two signals, which includes instantaneous frequency and spectral entropy. The data set transformed in this way was used to train the LSTM network. During the experiments, the LSTM networks were trained for both raw and spectrally transformed data. Then, the LSTM networks trained in this way were compared with each other. The obtained results prove that the transformation of input signals into images can be an effective method of improving the quality of classifiers based on deep learning.

[1]  Tomasz Rymarczyk.  The concept of the technological process control using a distributed industrial tomography system , 2018 .

[2]  David Valis,et al.  Application of selected Levy processes for degradation modelling of long range mine belt using real-time data , 2018, Archives of Civil and Mechanical Engineering.

[3]  Zhihui Lu,et al.  Selective encryption on ECG data in body sensor network based on supervised machine learning , 2020, Inf. Fusion.

[4]  Andrzej Krawczyk,et al.  Applications of Smart Textiles in Post-Stroke Rehabilitation , 2020, Sensors.

[5]  Robert Banasiak,et al.  Metrological evaluation of a 3D electrical capacitance tomography measurement system for two-phase flow fraction determination , 2013 .

[6]  Andrzej Romanowski,et al.  Big Data-Driven Contextual Processing Methods for Electrical Capacitance Tomography , 2019, IEEE Transactions on Industrial Informatics.

[7]  Antonio Pescapè,et al.  Precision Medicine and Artificial Intelligence: A Pilot Study on Deep Learning for Hypoglycemic Events Detection based on ECG , 2020, Scientific Reports.

[8]  Derya Avci,et al.  A new technique for ECG signal classification genetic algorithm Wavelet Kernel extreme learning machine , 2019, Optik.

[9]  Andrzej Romanowski,et al.  X-ray Imaging Analysis of Silo Flow Parameters Based on Trace Particles Using Targeted Crowdsourcing , 2019, Sensors.

[10]  Jan Werth,et al.  Deep learning approach for ECG-based automatic sleep state classification in preterm infants , 2020, Biomed. Signal Process. Control..

[11]  Sung Bum Pan,et al.  Deep Learning Based on 1-D Ensemble Networks Using ECG for Real-Time User Recognition , 2019, IEEE Transactions on Industrial Informatics.

[12]  Rickey E. Carter,et al.  Screening for cardiac contractile dysfunction using an artificial intelligence–enabled electrocardiogram , 2019, Nature Medicine.

[13]  Thirimachos Bourlai,et al.  A Novel Approach for ECG-Based Human Identification Using Spectral Correlation and Deep Learning , 2020, IEEE Transactions on Biometrics, Behavior, and Identity Science.

[14]  T. Rymarczyk,et al.  Innovative methods of neural reconstruction for tomographic images in maintenance of tank industrial reactors , 2019, Ekspolatacja i Niezawodnosc - Maintenance and Reliability.

[15]  Özal Yildirim,et al.  A novel wavelet sequence based on deep bidirectional LSTM network model for ECG signal classification , 2018, Comput. Biol. Medicine.

[16]  Hongling Zhu,et al.  Automatic multilabel electrocardiogram diagnosis of heart rhythm or conduction abnormalities with deep learning: a cohort study. , 2020, The Lancet. Digital health.

[17]  Xiaodai Dong,et al.  Inter-Patient CNN-LSTM for QRS Complex Detection in Noisy ECG Signals , 2019, IEEE Access.

[18]  Peng Lu,et al.  An Effective LSTM Recurrent Network to Detect Arrhythmia on Imbalanced ECG Dataset , 2019, Journal of healthcare engineering.

[19]  Yuh-Tay Sheen.  3D spectral analysis for vibration signals by wavelet-based demodulation , 2006 .

[20]  Tomasz Rymarczyk,et al.  Increasing the Reliability of Flood Embankments with Neural Imaging Method , 2018, Applied Sciences.

[21]  H. Calkins,et al.  Beat-to-beat QT interval variability: novel evidence for repolarization lability in ischemic and nonischemic dilated cardiomyopathy. , 1997, Circulation.

[22]  Rickey E Carter,et al.  An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction , 2019, The Lancet.

[23]  Matin Hashemi,et al.  LSTM-Based ECG Classification for Continuous Monitoring on Personal Wearable Devices , 2018, IEEE Journal of Biomedical and Health Informatics.

[24]  Jacek Kucharski,et al.  Surface temperature control of a rotating cylinder heated by moving inductors , 2017 .

[25]  Pankoo Kim,et al.  A Deep Bidirectional GRU Network Model for Biometric Electrocardiogram Classification Based on Recurrent Neural Networks , 2019, IEEE Access.

[26]  Antoni Świć,et al.  Maintenance of industrial reactors supported by deep learning driven ultrasound tomography , 2019 .

[27]  U. Rajendra Acharya,et al.  Arrhythmia detection using deep convolutional neural network with long duration ECG signals , 2018, Comput. Biol. Medicine.

[28]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[29]  Arnab Bhattacharjee,et al.  Deep Learning Approach to Cardiovascular Disease Classification Employing Modified ECG Signal from Empirical Mode Decomposition , 2019, Biomed. Signal Process. Control..

[30]  Sang-Hyun Kim,et al.  Prediction of Blood Pressure after Induction of Anesthesia Using Deep Learning: A Feasibility Study , 2019 .