ECG Analysis

Inter-Patient CNN-LSTM for QRS Complex Detection in Noisy ECG Signals

In this paper, a convolutional neural network (CNN) with long short-term memory (LSTM) is designed to detect QRS complexes in noisy electrocardiogram (ECG) signals. The CNN performs feature extraction while the LSTM determines the QRS complex timings. A multi-layer perception (MLP) after the LSTM is added to format the QRS complex detection predictions. With a unique data preparation procedure that includes proper design of training dataset, the proposed CNN-LSTM can achieve superior inter-patient testing performance, which means the testing and training datasets do not share any same patient ECG records. This generalization ability characteristic is critical to automated ECG analysis in an age of big data collected from noisy wearable ECG devices. The MIT-BIH and the European ST-T noise stress test databases are used to validate the effectiveness of the proposed algorithm in terms of sensitivity (recall), positive predictive value (precision), F1 score and timing root mean square error of R peak positions.

 

Detecting Noisy ECG QRS Complexes Using WaveletCNN Autoencoder and ConvLSTM

In this paper, we propose a novel machine learning pipeline to detect QRS complexes in very noisy wearable electrocardiogram (ECG) devices. The machine learning pipeline consists of a Butterworth filter, two wavelet convolutional neural networks (WaveletCNNs) autoencoders, an optional QRS complex inverter, a Monte Carlo k-nearest neighbours (k-NN), and a convolutional long short-term memory (ConvLSTM). WaveletCNN autoencoders filter out electrode contact noise, instrumentation noise, and motion artifact noise by using the advantages of wavelet filters and convolutional neural networks. The QRS complex inverter flips inverted QRS complexes. Monte Carlo k-NN performs automatic gain control on the ECG signals in order to normalize it. The ConvLSTM executes the final QRS complex detection by using the power of a convolutional neural network and a long short-term memory. The MIT-BIH, the European ST-T, and the Long Term ST database Noise Stress Test databases provide the training and testing ECG recordings. The proposed machine learning pipeline performs 3 standard deviations better than the state of the art QRS complex detection algorithms in terms of F 1 score for very noisy environments.