Similarity-Based Fuzzy Classification of ECG and Capnogram Signals

A method for ECG and capnogram signals classification is proposed based on fuzzy similarity evaluation, where shape exchange algorithm and fuzzy inference are combined. It aims to be applied to quasi-periodic biomedical signals and has low computational cost. On the experiments for atrial fibrillation (AF) classification using two databases: MIT-BIH AF and MITBIH Normal Sinus Rhythm, values of 100%, 94.4%, and 97.6% for sensitivity, specificity, and accuracy respectively, and execution time of 0.6 s are obtained. The proposal is capable of been extended to classify different diseases, from ECG and capnogram signals, such as: Brugada syndrome, AV block, hypoventilation, and asthma among others to be implemented in low computational resources devices.

Download Full-text

Classification of cardiac arrhythmias using deep learning

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i2.33.14195 ◽

2018 ◽

Vol 7 (3.3) ◽

pp. 401

Author(s):

Jeong Hwan Kim ◽

Jeong Whan Lee ◽

Kyeong Seop Kim

Keyword(s):

Deep Learning ◽

Normal Sinus Rhythm ◽

Premature Ventricular Contraction ◽

Heart Rhythm ◽

Future Research ◽

Normal Sinus ◽

Input Layer ◽

Hidden Layer ◽

Electrocardiogram Ecg

Background/Objectives: The main objective of this research is to design Deep Learning (DL) architecture to classify an electrocardiogram (ECG) signal into normal sinus rhythm (NSR), premature ventricular contraction (PVC), atrial premature contraction (APC) or right/left bundle branch block (RBBB/LBBB) arrhythmia by empirically optimizing the numbers of hidden layers, the number of neurons in each hidden layer and the number of neurons in input layer in DL model.Methods/Statistical analysis: For our experimental simulations, PhysioBank-MIT/BIH annotated ECG database was considered to classify heart beats into abnormal rhythms (PVC, APC, RBBB, LBBB) or normal sinus. The performance of classifying ECG beats by the proposed DL architecture was evaluated by computing the overall accuracy of classifying NSR or four different arrhythmias.Findings: Base on testing MIT/BIH arrhythmia database, the proposed DL model can classify the heart rhythm into one of NSR, PVC, APC, RBBB or LBBB beat with the mean accuracy of 95.5% by implementing DL architecture with 200 neurons in input layer, 100 neurons in the first and second hidden layer, respectively and 80 neurons in the 3rd hidden layer.Improvements/Applications: Our experimental results show that the proposed DL model might not be quite accurate for detecting APC beats due to its morphological resemblance of NSR. Therefore, we might need to design more sophisticated DL architecture by including more temporal characteristics of APC to increase the classification accuracy of APC arrhythmia in the future research efforts.

Download Full-text

Machine learning issues and opportunities in ultrafast particle classification for label-free microflow cytometry

Scientific Reports ◽

10.1038/s41598-020-77765-w ◽

2020 ◽

Vol 10 (1) ◽

Author(s):

Alessio Lugnan ◽

Emmanuel Gooskens ◽

Jeremy Vatin ◽

Joni Dambre ◽

Peter Bienstman

Keyword(s):

Machine Learning ◽

Computational Cost ◽

Particle Analysis ◽

Label Free ◽

Machine Learning Approach ◽

Microflow Cytometer ◽

Learning Machine ◽

Learning Issues ◽

Low Computational Cost

AbstractMachine learning offers promising solutions for high-throughput single-particle analysis in label-free imaging microflow cytomtery. However, the throughput of online operations such as cell sorting is often limited by the large computational cost of the image analysis while offline operations may require the storage of an exceedingly large amount of data. Moreover, the training of machine learning systems can be easily biased by slight drifts of the measurement conditions, giving rise to a significant but difficult to detect degradation of the learned operations. We propose a simple and versatile machine learning approach to perform microparticle classification at an extremely low computational cost, showing good generalization over large variations in particle position. We present proof-of-principle classification of interference patterns projected by flowing transparent PMMA microbeads with diameters of $${15.2}\,\upmu \text {m}$$ 15.2 μ m and $${18.6}\,\upmu \text {m}$$ 18.6 μ m . To this end, a simple, cheap and compact label-free microflow cytometer is employed. We also discuss in detail the detection and prevention of machine learning bias in training and testing due to slight drifts of the measurement conditions. Moreover, we investigate the implications of modifying the projected particle pattern by means of a diffraction grating, in the context of optical extreme learning machine implementations.

Download Full-text

ATRIAL FIBRILLATION DETECTION ON ELECTROCARDIOGRAMS WITH CONVOLUTIONAL NEURAL NETWORKS

Informatyka Automatyka Pomiary w Gospodarce i Ochronie Środowiska ◽

10.35784/iapgos.116 ◽

2019 ◽

Vol 9 (4) ◽

pp. 69-73

Author(s):

Viktor Kifer ◽

Natalia Zagorodna ◽

Olena Hevko

Keyword(s):

Neural Network ◽

Atrial Fibrillation ◽

Sinus Rhythm ◽

Deep Neural Network ◽

Normal Sinus Rhythm ◽

Normal Sinus ◽

Noisy Signals ◽

Network Usage ◽

Model Training

In this paper, we present our research which confirms the suitability of the convolutional neural network usage for the classification of single-lead ECG recordings. The proposed method was designed for classifying normal sinus rhythm, atrial fibrillation (AF), non-AF related other abnormal heart rhythms and noisy signals. The method combines manually selected features with the features learned by the deep neural network. The Physionet Challenge 2017 dataset of over 8500 ECG recordings was used for the model training and validation. The trained model reaches an average F1-score 0.71 in classifying normal sinus rhythm, AF and other rhythms respectively.

Download Full-text

Epileptic Seizure Data Classification Using RBAs and Linear SVM

Biomedical & Pharmacology Journal ◽

10.13005/bpj/1674 ◽

2019 ◽

Vol 12 (2) ◽

pp. 549-562

Author(s):

Alpika Tripathi ◽

Geetika Srivastava ◽

K.K. Singh ◽

P.K. Maurya

Keyword(s):

Computational Cost ◽

High Sensitivity ◽

Support Vector ◽

Medical Sciences ◽

Hilbert Huang Transform ◽

Relief Algorithm ◽

Eeg Data ◽

Linear Svm ◽

Sensitivity Specificity

The objective of this paper is to make a distinction between EEG data of normal and epileptic subjects. Methods: The dataset is taken from 20-30 years healthy male/female subjects from EEG lab of Dept. of Neurology, Dr. RML Institute of Medical Sciences, Lucknow (India). The feature extraction has been done using the Hilbert Huang Transform (HHT) method. The experimental EEG signals have been decomposed till 5th level of Intrinsic Mode Function (IMF) followed by calculation of high order statistical values of each IMF. Relief algorithm (RBAs) is used for feature selection and classification is performed using Linear Support Vector Machine (Linear SVM). This paper gives an independent approach of classifying Epileptic EEG data with reduced computational cost and high accuracy. Our classification result shows sensitivity, specificity, selectivity and accuracy of 96.4%, 79.16%, 84.3% and 88.5% respectively. The proposed method has been analyzed to be very effective in accurate classification of epileptic EEG data with high sensitivity.

Download Full-text

ECG Classification using Machine Learning

International Journal of Recent Technology and Engineering - 2 ◽

10.35940/ijrte.d6989.118419 ◽

2019 ◽

Vol 8 (4) ◽

pp. 2492-2494

Keyword(s):

Heart Failure ◽

Cardiovascular Disease ◽

Congestive Heart Failure ◽

Sinus Rhythm ◽

Normal Sinus Rhythm ◽

Original Data ◽

Ecg Signals ◽

Normal Sinus ◽

Electrocardiogram Signals

Recently, the obvious increasing number of cardiovascular disease, the automatic classification research of Electrocardiogram signals (ECG) has been playing a important part in the clinical diagnosis of cardiovascular disease. Convolution neural network (CNN) based method is proposed to classify ECG signals. The proposed CNN model consists of five layers in addition to the input layer and the output layer, i.e., two convolution layers, two down sampling layers and one full connection layer, extracting the effective features from the original data and classifying the features using wavelet .The classification of ARR (Arrhythmia), CHF (Congestive Heart Failure), and NSR (Normal Sinus Rhythm) signals. The experimental results contains on ARR signals from the MIT-BIH arrhythmia,CHF signals from the BIDMC Congestive Heart Failure and NSR signals from the MIT-BIH Normal Sinus Rhythm Databases show that the proposed method achieves a promising classification accuracy of 90.63%, significantly outperforming several typical ECG classification methods.

Download Full-text

Automatic detection and classification of cardiac arrhythmia using neural network

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i3.14084 ◽

2018 ◽

Vol 7 (3) ◽

pp. 1482

Author(s):

N N. S. V Rama Raju ◽

V Malleswara Rao ◽

I Srinivasa Rao

Keyword(s):

Neural Network ◽

Wavelet Transform ◽

Normal Sinus Rhythm ◽

Automatic Identification ◽

Supraventricular Arrhythmia ◽

Qrs Complex ◽

Complex Wavelet Transform ◽

Normal Sinus ◽

Ecg Database

This paper proposes a Neural Network classifier model for the automatic identification of the ventricular and supraventricular arrhythmias cardiac arrhythmias. The wavelet transform (DWT) and dual tree complex wavelet transform (DTCWT) is applied for QRS complex detec-tion. After segmentation both feature of DWT and DTCWT is combined for feature extraction, statistical feature has been calculated to re-duce the overhead of classifier. An adaptive filtering with the soft computed wavelet thersholding to the signals before the extraction is done in pre-processing. Different ECG database is considered to evaluate the propose work with MIT-BIH database Normal Sinus Rhythm Da-tabase (NSRD) , and MIT-BIH Supraventricular Arrhythmia Database (svdb) .The evaluated outcomes of ECG classification claims 98 -99 % of accuracy under different training and testing situation.

Download Full-text

Classification of Atrial Arrhythmias using Neural Networks

IAES International Journal of Artificial Intelligence (IJ-AI) ◽

10.11591/ijai.v7.i2.pp90-94 ◽

2018 ◽

Vol 7 (2) ◽

pp. 90 ◽

Cited By ~ 1

Author(s):

Jai Utkarsh ◽

Raju Kumar Pandey ◽

Shrey Kumar Dubey ◽

Shubham Sinha ◽

S. S. Sahu

Keyword(s):

Normal Sinus Rhythm ◽

High Accuracy ◽

Ecg Signal ◽

Atrial Arrhythmias ◽

Autoregressive Modelling ◽

Test Set ◽

Normal Sinus ◽

The Neural Network ◽

Electrocardiogram Ecg

Electrocardiogram (ECG) is an important tool used by clinicians for successful diagnosis and detection of Arrhythmias, like Atrial Fibrillation (AF) and Atrial Flutter (AFL). In this manuscript, an efficient technique of classifying atrial arrhythmias from Normal Sinus Rhythm (NSR) has been presented. Autoregressive Modelling has been used to capture the features of the ECG signal, which are then fed as inputs to the neural network for classification. The standard database available at Physionet Bank repository has been used for training, validation and testing of the model. Exhaustive experimental study has been carried out by extracting ECG samples of duration of 5 seconds, 10 seconds and 20 seconds. It provides an accuracy of 99% and 94.3% on training and test set respectively for 5 sec recordings. In 10 sec and 20 sec samples it shows 100% accuracy. Thus, the proposed method can be used to detect the arrhythmias in a small duration recordings with a fairly high accuracy.

Download Full-text