A machine learning approach to multi-level ECG signal quality classification

Simulating Motion Artifact Using an Autoregressive Model for Research in Biomedical Signal Quality Analysis

Motion artifact contamination may adversely affect the interpretation of biological signals. The development of algorithms to detect, identify, quantify, and mitigate motion artifact is typically performed using a ground truth signal contaminated with previously recorded motion artifact, or simulated motion artifact. The diversity of available motion artifact recordings is limited, and the rationales for existing models of motion artifact are poorly described. In this paper we developed an autoregressive (AR) model of motion artifact based on data collected from 6 subjects walking at slow, medium, and fast paces. The AR model was evaluated for its ability to generate diverse data that replicated the properties of the experimental data. The simulated motion artifact data was successful at learning key time domain and frequency domain properties, including the mean, variance, and power spectrum of the data, but was ineffective for imitating the morphology and probability distribution of the motion artifact data (kurtosis % error of 100.9-103.6%). More sophisticated models of motion artifact may be necessary to develop simulations of motion artifact.

A Deep Learning Approach for Featureless Robust Quality Assessment of Intermittent Atrial Fibrillation Recordings from Portable and Wearable Devices

Atrial fibrillation (AF) is the most common heart rhythm disturbance in clinical practice. It often starts with asymptomatic and very short episodes, which are extremely difficult to detect without long-term monitoring of the patient’s electrocardiogram (ECG). Although recent portable and wearable devices may become very useful in this context, they often record ECG signals strongly corrupted with noise and artifacts. This impairs automatized ulterior analyses that could only be conducted reliably through a previous stage of automatic identification of high-quality ECG intervals. So far, a variety of techniques for ECG quality assessment have been proposed, but poor performances have been reported on recordings from patients with AF. This work introduces a novel deep learning-based algorithm to robustly identify high-quality ECG segments within the challenging environment of single-lead recordings alternating sinus rhythm, AF episodes and other rhythms. The method is based on the high learning capability of a convolutional neural network, which has been trained with 2-D images obtained when turning ECG signals into wavelet scalograms. For its validation, almost 100,000 ECG segments from three different databases have been analyzed during 500 learning-testing iterations, thus involving more than 320,000 ECGs analyzed in total. The obtained results have revealed a discriminant ability to detect high-quality and discard low-quality ECG excerpts of about 93%, only misclassifying around 5% of clean AF segments as noisy ones. In addition, the method has also been able to deal with raw ECG recordings, without requiring signal preprocessing or feature extraction as previous stages. Consequently, it is particularly suitable for portable and wearable devices embedding, facilitating early detection of AF as well as other automatized diagnostic facilities by reliably providing high-quality ECG excerpts to further processing stages.

Electrohysterogram for ANN-Based Prediction of Imminent Labor in Women with Threatened Preterm Labor Undergoing Tocolytic Therapy

Threatened preterm labor (TPL) is the most common cause of hospitalization in the second half of pregnancy and entails high costs for health systems. Currently, no reliable labor proximity prediction techniques are available for clinical use. Regular checks by uterine electrohysterogram (EHG) for predicting preterm labor have been widely studied. The aim of the present study was to assess the feasibility of predicting labor with a 7- and 14-day time horizon in TPL women, who may be under tocolytic treatment, using EHG and/or obstetric data. Based on 140 EHG recordings, artificial neural networks were used to develop prediction models. Non-linear EHG parameters were found to be more reliable than linear for differentiating labor in under and over 7/14 days. Using EHG and obstetric data, the <7- and <14-day labor prediction models achieved an AUC in the test group of 87.1 ± 4.3% and 76.2 ± 5.8%, respectively. These results suggest that EHG can be reliable for predicting imminent labor in TPL women, regardless of the tocolytic therapy stage. This paves the way for the development of diagnostic tools to help obstetricians make better decisions on treatments, hospital stays and admitting TPL women, and can therefore reduce costs and improve maternal and fetal wellbeing.

A Deep Learning Method to Detect Atrial Fibrillation Based on Continuous Wavelet Transform

Atrial fibrillation (AF) is one of the most common arrhythmias. The automatic AF detection is of great clinical significance but at the same time it remains a big problem to researchers. In this study, a novel deep learning method to detect AF was proposed. For a 10 s length single lead electrocardiogram (ECG) signal, the continuous wavelet transform (CWT) was used to obtain the wavelet coefficient matrix, and then a convolutional neural network (CNN) with a specific architecture was trained to discriminate the rhythm of the signal. The ECG data in multiple databases were divided into 4 classes according to the rhythm annotation: normal sinus rhythm (NSR), atrial fibrillation (AF), other types of arrhythmia except AF (OTHER), and noise signal (NOISE). The method was evaluated using three different wavelet bases. The experiment showed promising results when using a Morlet wavelet, with an overall accuracy of 97.56%, an average sensitivity of 97.56%, an average specificity of 99.19%. Besides, the area under curve (AUC) value is 0.9983, which showed that the proposed method was effective for detecting AF.

Deep Learning-Based Stacked Denoising and Autoencoder for ECG Heartbeat Classification

The electrocardiogram (ECG) is a widely used, noninvasive test for analyzing arrhythmia. However, the ECG signal is prone to contamination by different kinds of noise. Such noise may cause deformation on the ECG heartbeat waveform, leading to cardiologists’ mislabeling or misinterpreting heartbeats due to varying types of artifacts and interference. To address this problem, some previous studies propose a computerized technique based on machine learning (ML) to distinguish between normal and abnormal heartbeats. Unfortunately, ML works on a handcrafted, feature-based approach and lacks feature representation. To overcome such drawbacks, deep learning (DL) is proposed in the pre-training and fine-tuning phases to produce an automated feature representation for multi-class classification of arrhythmia conditions. In the pre-training phase, stacked denoising autoencoders (DAEs) and autoencoders (AEs) are used for feature learning; in the fine-tuning phase, deep neural networks (DNNs) are implemented as a classifier. To the best of our knowledge, this research is the first to implement stacked autoencoders by using DAEs and AEs for feature learning in DL. Physionet’s well-known MIT-BIH Arrhythmia Database, as well as the MIT-BIH Noise Stress Test Database (NSTDB). Only four records are used from the NSTDB dataset: 118 24 dB, 118 −6 dB, 119 24 dB, and 119 −6 dB, with two levels of signal-to-noise ratio (SNRs) at 24 dB and −6 dB. In the validation process, six models are compared to select the best DL model. For all fine-tuned hyperparameters, the best model of ECG heartbeat classification achieves an accuracy, sensitivity, specificity, precision, and F1-score of 99.34%, 93.83%, 99.57%, 89.81%, and 91.44%, respectively. As the results demonstrate, the proposed DL model can extract high-level features not only from the training data but also from unseen data. Such a model has good application prospects in clinical practice.

Personalized Monitoring Model for Electrocardiogram Signals: Diagnostic Accuracy Study

BACKGROUND

Due to the COVID-19 pandemic, the demand for remote electrocardiogram (ECG) monitoring has increased drastically in an attempt to prevent the spread of the virus and keep vulnerable individuals with less severe cases out of hospitals. Enabling clinicians to set up remote patient ECG monitoring easily and determining how to classify the ECG signals accurately so relevant alerts are sent in a timely fashion is an urgent problem to be addressed for remote patient monitoring (RPM) to be adopted widely. Hence, a new technique is required to enable routine and widespread use of RPM, as is needed due to COVID-19.

OBJECTIVE

The primary aim of this research is to create a robust and easy-to-use solution for personalized ECG monitoring in real-world settings that is precise, easily configurable, and understandable by clinicians.

METHODS

In this paper, we propose a Personalized Monitoring Model (PMM) for ECG data based on motif discovery. Motif discovery finds meaningful or frequently recurring patterns in patient ECG readings. The main strategy is to use motif discovery to extract a small sample of personalized motifs for each individual patient and then use these motifs to predict abnormalities in real-time readings of that patient using an artificial logical network configured by a physician.

RESULTS

Our approach was tested on 30 minutes of ECG readings from 32 patients. The average diagnostic accuracy of the PMM was always above 90% and reached 100% for some parameters, compared to 80% accuracy for the Generalized Monitoring Models (GMM). Regardless of parameter settings, PMM training models were generated within 3-4 minutes, compared to 1 hour (or longer, with increasing amounts of training data) for the GMM.

CONCLUSIONS

Our proposed PMM almost eliminates many of the training and small sample issues associated with GMMs. It also addresses accuracy and computational cost issues of the GMM, caused by the uniqueness of heartbeats and training issues. In addition, it addresses the fact that doctors and nurses typically do not have data science training and the skills needed to configure, understand, and even trust existing black box machine learning models.

Artefact detection and quality assessment of ambulatory ECG signals

BACKGROUND AND OBJECTIVES

The presence of noise sources could reduce the diagnostic capability of the ECG signal and result in inappropriate treatment decisions. To mitigate this problem, automated algorithms to detect artefacts and quantify the quality of the recorded signal are needed. In this study we present an automated method for the detection of artefacts and quantification of the signal quality. The suggested methodology extracts descriptive features from the autocorrelation function and feeds these to a RUSBoost classifier. The posterior probability of the clean class is used to create a continuous signal quality assessment index. Firstly, the robustness of the proposed algorithm is investigated and secondly, the novel signal quality assessment index is evaluated.

METHODS

Data were used from three different studies: a Sleep study, the PhysioNet 2017 Challenge and a Stress study. Binary labels, clean or contaminated, were available from different annotators with experience in ECG analysis. Two types of realistic ECG noise from the MIT-BIH Noise Stress Test Database (NSTDB) were added to the Sleep study to test the quality index. Firstly, the model was trained on the Sleep dataset and subsequently tested on a subset of the other two datasets. Secondly, all recording conditions were taken into account by training the model on a subset derived from the three datasets. Lastly, the posterior probabilities of the model for the different levels of agreement between the annotators were compared.

RESULTS

AUC values between 0.988 and 1.000 were obtained when training the model on the Sleep dataset. These results were further improved when training on the three datasets and thus taking all recording conditions into account. A Pearson correlation coefficient of 0.8131 was observed between the score of the clean class and the level of agreement. Additionally, significant quality decreases per noise level for both types of added noise were observed.

CONCLUSIONS

The main novelty of this study is the new approach to ECG signal quality assessment based on the posterior clean class probability of the classifier.

A Cascaded Convolutional Neural Network for Assessing Signal Quality of Dynamic ECG

Motion artifacts and myoelectrical noise are common issues complicating the collection and processing of dynamic electrocardiogram (ECG) signals. Recent signal quality studies have utilized a binary classification metric in which ECG samples are determined to either be clean or noisy. However, the clinical use of dynamic ECGs requires specific noise level classification for varying applications. Conventional signal processing methods, including waveform discrimination, are limited in their ability to remove motion artifacts and myoelectrical noise from dynamic ECGs. As such, a novel cascaded convolutional neural network (CNN) is proposed and demonstrated for application to the five-classification problem (low interference, mild motion artifacts, mild myoelectrical noise, severe motion artifacts, and severe myoelectrical noise). Specifically, this study finally categorizes dynamic ECG signals into three levels (low, mild, and severe) using the proposed CNN to meet clinical requirements. The network includes two components, the first of which was used to distinguish signal interference types, while the second was used to distinguish signal interference levels. This model does not require feature engineering, includes powerful nonlinear mapping capabilities, and is robust to varying noise types. Experimental data are composed of private dataset and public dataset, which were acquired from 90,000 four-second dynamic ECG signals and MIT-BIH Arrhythmia database, respectively. Experimental results produced an overall recognition rate of 92.7% on private dataset and 91.8% on public dataset. These results suggest the proposed technique to be a valuable new tool for dynamic ECG auxiliary diagnosis.

Gastroenterology Meets Machine Learning: Status Quo and Quo Vadis

Machine learning has undergone a transition phase from being a pure statistical tool to being one of the main drivers of modern medicine. In gastroenterology, this technology is motivating a growing number of studies that rely on these innovative methods to deal with critical issues related to this practice. Hence, in the light of the burgeoning research on the use of machine learning in gastroenterology, a systematic review of the literature is timely. In this work, we present the results gleaned through a systematic review of prominent gastroenterology literature using machine learning techniques. Based on the analysis of 88 journal articles, we delimit the scope of application, we discuss current limitations including bias, lack of transparency, accountability, and data availability, and we put forward future avenues.

Data Fusion of Multivariate Time Series: Application to Noisy 12-Lead ECG Signals

Twelve-lead Electrocardiograph (ECG) signals fusion is crucial for further ECG signal processing. In this paper, based on the idea of the local weighted linear prediction algorithm, a novel fusion data algorithm is proposed, which was applied in data fusion of the 12-lead ECG signals. In order to analyze the signal quality comprehensively, the quality characteristics should be adequately retained in the final fused result. In our algorithm, the values for the weighted coefficient of state points were closely related to the final fused result. Thus, two fuzzy inference systems were designed to calculate the weighted coefficients. For the sake of assessing the performance of our method, synthetic ECG signals and realistic ECG signals were applied in the experiments. Experimental results indicate that our method can fuse the 12-lead ECG signals effectively with inherit the quality characteristics of original ECG signals inherited properly.

Real-time Evaluation of ECG Acquisition Systems through Signal Quality Assessment in Horses during Submaximal Treadmill Test

This paper reports on a novel real time index designed to assess the quality of electrocardiographic (ECG) traces recorded in a group of five horses during a submaximal treadmill test procedure. During the experimental protocol two ECG monitoring systems were simultaneously applied to the animals. The first system was equipped with textile electrodes while the second one with standard red-dot electrodes. The procedure comprised four phases with an increased treadmill velocity, specifically, Walk 1, Trot 1, Trot 2 and Gallop. Three signal quality levels have been fixed according to the amount of noise present in the ECG trace: good (G), acceptable (A), and unacceptable (U). Moreover, a statistical comparison between textile and red-dot electrodes has been performed in terms of percentage of signal belonging to each class. Even if preliminary, results showed that in each experimental phase textile electrodes are more robust to movement artifacts with respect to the reddot showing a significant evidence of their better performance. These results enable to design robust wearable monitoring systems suitable to improve the quality of collected ECG, reducing the great amount of motion artifacts due to red-dot electrode application and leading to a more accurate diagnosis of high speed arrhythmias.

Automated recognition of cardiac arrhythmias using sparse decomposition over composite dictionary

BACKGROUND AND OBJECTIVE

Cardiovascular diseases (CVDs) are the leading cause of deaths worldwide. Due to an increase in the rate of global mortalities, biopathological signal processing and evaluation are widely used in the ambulatory situations for healthcare applications. For decades, the processing of pathological electrocardiogram (ECG) signals for arrhythmia detection has been thoroughly studied for diagnosis of various cardiovascular diseases. Apart from these studies, efficient diagnosis of ECG signals remains a challenge in the clinical cardiovascular domain due to its non-stationary nature. The classical signal processing methods are widely employed to analyze the ECG signals, but they exhibit certain limitations and hence, are insufficient to achieve higher accuracy.

METHODS

This study presents a novel technique for an efficient representation of electrocardiogram (ECG) signals using sparse decomposition using composite dictionary (CD). The dictionary consists of the stockwell, sine and cosine analytical functions. The technique decomposes an input ECG signal into stationary and non-stationary components or atoms. For each of these atoms, five features i.e., permutation entropy, energy, RR-interval, standard deviation and kurtosis are extracted to determine the feature sets representing the heartbeats that are classified into different categories using the multi-class least-square twin support vector machines. The artificial bee colony (ABC) technique is used to determine the optimal classifier parameters. The proposed method is evaluated under category and personalized schemes and its validation is performed on MIT-BIH data.

RESULTS

The experimental results reported a higher overall accuracy of 99.21% and 90.08% in category and personalized schemes respectively than the existing techniques reported in the literature. Further a sensitivity, positive predictivity and F-score of 99.21% each in the category based scheme and 90.08% each in the personalized schemes respectively.

CONCLUSIONS

The proposed methodology can be utilized in computerized decision support systems to monitor different classes of cardiac arrhythmias with higher accuracy for early detection and treatment of cardiovascular diseases.

ECG signal classification for the detection of cardiac arrhythmias using a convolutional recurrent neural network

OBJECTIVE

The electrocardiogram (ECG) provides an effective, non-invasive approach for clinical diagnosis in patients with cardiac diseases such as atrial fibrillation (AF). AF is the most common cardiac rhythm disturbance and affects ~2% of the general population in industrialized countries. Automatic AF detection in clinics remains a challenging task due to the high inter-patient variability of ECGs, and unsatisfactory existing approaches for AF diagnosis (e.g. atrial or ventricular activity-based analyses).

APPROACH

We have developed RhythmNet, a 21-layer 1D convolutional recurrent neural network, trained using 8528 single-lead ECG recordings from the 2017 PhysioNet/Computing in Cardiology (CinC) Challenge, to classify ECGs of different rhythms including AF automatically. Our RhythmNet architecture contained 16 convolutions to extract features directly from raw ECG waveforms, followed by three recurrent layers to process ECGs of varying lengths and to detect arrhythmia events in long recordings. Large 15  ×  1 convolutional filters were used to effectively learn the detailed variations of the signal within small time-frames such as the P-waves and QRS complexes. We employed residual connections throughout RhythmNet, along with batch-normalization and rectified linear activation units to improve convergence during training.

MAIN RESULTS

We evaluated our algorithm on 3658 testing data and obtained an F 1 accuracy of 82% for classifying sinus rhythm, AF, and other arrhythmias. RhythmNet was also ranked 5th in the 2017 CinC Challenge.

SIGNIFICANCE

Potentially, our approach could aid AF diagnosis in clinics and be used for patient self-monitoring to improve the early detection and effective treatment of AF.

Capturing Electrocardiogram Signals from Chairs by Multiple Capacitively Coupled Unipolar Electrodes

A prototype of an electrocardiogram (ECG) signal acquisition system with multiple unipolar capacitively coupled electrodes is designed and experimentally tested. Capacitively coupled electrodes made of a standard printed circuit board (PCB) are used as the sensing electrodes. Different from the conventional measurement schematics, where one single lead ECG signal is acquired from a pair of sensing electrodes, the sensing electrodes in our approaches operate in a unipolar mode, i.e., the biopotential signals picked up by each sensing electrodes are amplified and sampled separately. Four unipolar electrodes are mounted on the backrest of a regular chair and therefore four channel of signals containing ECG information are sampled and processed. It is found that the qualities of ECG signal contained in the four channel are different from each other. In order to pick up the ECG signal, an index for quality evaluation, as well as for aggregation of multiple signals, is proposed based on phase space reconstruction. Experimental tests are carried out while subjects sitting on the chair and clothed. The results indicate that the ECG signals can be reliably obtained in such a unipolar way.

Model-Based Assessment of f-Wave Signal Quality in Patients With Atrial Fibrillation

OBJECTIVE

The detection and analysis of atrial fibrillation (AF) in the ECG is greatly influenced by signal quality. The present study proposes and evaluates a model-based f-wave signal quality index (SQI), denoted , for use in the QRST-cancelled ECG signal.

METHODS

is computed using a harmonic f-wave model, allowing for variation in frequency and amplitude. The properties of  are evaluated on both f-waves and P-waves using 378 12-lead ECGs, 1875 single-lead ECGs, and simulated signals.

RESULTS

decreases monotonically when noise is added to f-wave signals, even for noise which overlaps spectrally with f-waves. Moreover, is shown to be closely associated with the accuracy of AF frequency estimation, where implies accurate estimation. When  is used as a measure of f-wave presence, AF detection performance improves: the sensitivity increases from 97.0% to 98.1% and the specificity increases from 97.4% to 97.8% when compared to the reference detector.

CONCLUSION

The proposed SQI represents a novel approach to assessing f-wave signal quality, as well as to determining whether f-waves are present.

SIGNIFICANCE

The use of  improves the detection of AF and benefits the analysis of noisy ECGs.

Computational techniques for ECG analysis and interpretation in light of their contribution to medical advances

Widely developed for clinical screening, electrocardiogram (ECG) recordings capture the cardiac electrical activity from the body surface. ECG analysis can therefore be a crucial first step to help diagnose, understand and predict cardiovascular disorders responsible for 30% of deaths worldwide. Computational techniques, and more specifically machine learning techniques and computational modelling are powerful tools for classification, clustering and simulation, and they have recently been applied to address the analysis of medical data, especially ECG data. This review describes the computational methods in use for ECG analysis, with a focus on machine learning and 3D computer simulations, as well as their accuracy, clinical implications and contributions to medical advances. The first section focuses on heartbeat classification and the techniques developed to extract and classify abnormal from regular beats. The second section focuses on patient diagnosis from whole recordings, applied to different diseases. The third section presents real-time diagnosis and applications to wearable devices. The fourth section highlights the recent field of personalized ECG computer simulations and their interpretation. Finally, the discussion section outlines the challenges of ECG analysis and provides a critical assessment of the methods presented. The computational methods reported in this review are a strong asset for medical discoveries and their translation to the clinical world may lead to promising advances.

Grid mapping: a novel method of signal quality evaluation on a single lead electrocardiogram

Diagnosis of long-term electrocardiogram (ECG) calls for automatic and accurate methods of ECG signal quality estimation, not only to lighten the burden of the doctors but also to avoid misdiagnoses. In this paper, a novel waveform-based method of phase-space reconstruction for signal quality estimation on a single lead ECG was proposed by projecting the amplitude of the ECG and its first order difference into grid cells. The waveform of a single lead ECG was divided into non-overlapping episodes (T_s = 10, 20, 30 s), and the number of grids in both the width and the height of each map are in the range [20, 100] (N_X = N_Y = 20, 30, 40, … 90, 100). The blank pane ratio (BPR) and the entropy were calculated from the distribution of ECG sampling points which were projected into the grid cells. Signal Quality Indices (SQI) bSQI and eSQI were calculated according to the BPR and the entropy, respectively. The MIT-BIH Noise Stress Test Database was used to test the performance of bSQI and eSQI on ECG signal quality estimation. The signal-to-noise ratio (SNR) during the noisy segments of the ECG records in the database is 24, 18, 12, 6, 0 and - 6 dB, respectively. For the SQI quantitative analysis, the records were divided into three groups: good quality group (24, 18 dB), moderate group (12, 6 dB) and bad quality group (0, - 6 dB). The classification among good quality group, moderate quality group and bad quality group were made by linear support-vector machine with the combination of the BPR, the entropy, the bSQI and the eSQI. The classification accuracy was 82.4% and the Cohen's Kappa coefficient was 0.74 on a scale of N_X = 40 and T_s = 20 s. In conclusion, the novel grid mapping offers an intuitive and simple approach to achieving signal quality estimation on a single lead ECG.

Noise Maps for Quantitative and Clinical Severity Towards Long-Term ECG Monitoring

Noise and artifacts are inherent contaminating components and are particularly present in Holter electrocardiogram (ECG) monitoring. The presence of noise is even more significant in long-term monitoring (LTM) recordings, as these are collected for several days in patients following their daily activities; hence, strong artifact components can temporarily impair the clinical measurements from the LTM recordings. Traditionally, the noise presence has been dealt with as a problem of non-desirable component removal by means of several quantitative signal metrics such as the signal-to-noise ratio (SNR), but current systems do not provide any information about the true impact of noise on the ECG clinical evaluation. As a first step towards an alternative to classical approaches, this work assesses the ECG quality under the assumption that an ECG has good quality when it is clinically interpretable. Therefore, our hypotheses are that it is possible (a) to create a clinical severity score for the effect of the noise on the ECG, (b) to characterize its consistency in terms of its temporal and statistical distribution, and (c) to use it for signal quality evaluation in LTM scenarios. For this purpose, a database of external event recorder (EER) signals is assembled and labeled from a clinical point of view for its use as the gold standard of noise severity categorization. These devices are assumed to capture those signal segments more prone to be corrupted with noise during long-term periods. Then, the ECG noise is characterized through the comparison of these clinical severity criteria with conventional quantitative metrics taken from traditional noise-removal approaches, and noise maps are proposed as a novel representation tool to achieve this comparison. Our results showed that neither of the benchmarked quantitative noise measurement criteria represent an accurate enough estimation of the clinical severity of the noise. A case study of long-term ECG is reported, showing the statistical and temporal correspondences and properties with respect to EER signals used to create the gold standard for clinical noise. The proposed noise maps, together with the statistical consistency of the characterization of the noise clinical severity, paves the way towards forthcoming systems providing us with noise maps of the noise clinical severity, allowing the user to process different ECG segments with different techniques and in terms of different measured clinical parameters.

Benchmarking of a T-wave alternans detection method based on empirical mode decomposition

BACKGROUND AND OBJECTIVE

T-wave alternans (TWA) is a fluctuation of the ST-T complex occurring on an every-other-beat basis of the surface electrocardiogram (ECG). It has been shown to be an informative risk stratifier for sudden cardiac death, though the lack of gold standard to benchmark detection methods has promoted the use of synthetic signals. This work proposes a novel signal model to study the performance of a TWA detection. Additionally, the methodological validation of a denoising technique based on empirical mode decomposition (EMD), which is used here along with the spectral method, is also tackled.

METHODS

The proposed test bed system is based on the following guidelines: (1) use of open source databases to enable experimental replication; (2) use of real ECG signals and physiological noise; (3) inclusion of randomized TWA episodes. Both sensitivity (Se) and specificity (Sp) are separately analyzed. Also a nonparametric hypothesis test, based on Bootstrap resampling, is used to determine whether the presence of the EMD block actually improves the performance.

RESULTS

The results show an outstanding specificity when the EMD block is used, even in very noisy conditions (0.96 compared to 0.72 for SNR = 8 dB), being always superior than that of the conventional SM alone. Regarding the sensitivity, using the EMD method also outperforms in noisy conditions (0.57 compared to 0.46 for SNR=8 dB), while it decreases in noiseless conditions.

CONCLUSIONS

The proposed test setting designed to analyze the performance guarantees that the actual physiological variability of the cardiac system is reproduced. The use of the EMD-based block in noisy environment enables the identification of most patients with fatal arrhythmias.

Ventricular Fibrillation and Tachycardia detection from surface ECG using time-frequency representation images as input dataset for machine learning

BACKGROUND AND OBJECTIVE

To safely select the proper therapy for Ventricullar Fibrillation (VF) is essential to distinct it correctly from Ventricular Tachycardia (VT) and other rhythms. Provided that the required therapy would not be the same, an erroneous detection might lead to serious injuries to the patient or even cause Ventricular Fibrillation (VF). The main novelty of this paper is the use of time-frequency (t-f) representation images as the direct input to the classifier. We hypothesize that this method allow to improve classification results as it allows to eliminate the typical feature selection and extraction stage, and its corresponding loss of information.

METHODS

The standard AHA and MIT-BIH databases were used for evaluation and comparison with other authors. Previous to t-f Pseudo Wigner-Ville (PWV) calculation, only a basic preprocessing for denoising and signal alignment is necessary. In order to check the validity of the method independently of the classifier, four different classifiers are used: Logistic Regression with L2 Regularization (L2 RLR), Adaptive Neural Network Classifier (ANNC), Support Vector Machine (SSVM), and Bagging classifier (BAGG).

RESULTS

The main classification results for VF detection (including flutter episodes) are 95.56% sensitivity and 98.8% specificity, 88.80% sensitivity and 99.5% specificity for ventricular tachycardia (VT), 98.98% sensitivity and 97.7% specificity for normal sinus, and 96.87% sensitivity and 99.55% specificity for other rhythms.

CONCLUSION

Results shows that using t-f data representations to feed classifiers provide superior performance values than the feature selection strategies used in previous works. It opens the door to be used in any other detection applications.

Non-invasive Fetal ECG Signal Quality Assessment for Multichannel Heart Rate Estimation

OBJECTIVE

The noninvasive fetal ECG (NI-FECG) from abdominal recordings offers novel prospects for prenatal monitoring. However, NI-FECG signals are corrupted by various nonstationary noise sources, making the processing of abdominal recordings a challenging task. In this paper, we present an online approach that dynamically assess the quality of NI-FECG to improve fetal heart rate (FHR) estimation.

METHODS

Using a naive Bayes classifier, state-of-the-art and novel signal quality indices (SQIs), and an existing adaptive Kalman filter, FHR estimation was improved. For the purpose of training and validating the proposed methods, a large annotated private clinical dataset was used.

RESULTS

The suggested classification scheme demonstrated an accuracy of Krippendorff's alpha in determining the overall quality of NI-FECG signals. The proposed Kalman filter outperformed alternative methods for FHR estimation achieving accuracy.

CONCLUSION

The proposed algorithm was able to reliably reflect changes of signal quality and can be used in improving FHR estimation.

SIGNIFICANCE

NI-ECG signal quality estimation and multichannel information fusion are largely unexplored topics. Based on previous works, multichannel FHR estimation is a field that could strongly benefit from such methods. The developed SQI algorithms as well as resulting classifier were made available under a GNU GPL open-source license and contributed to the FECGSYN toolbox.

Set-Based Discriminative Measure for Electrocardiogram Beat Classification

Computer aided diagnosis systems can help to reduce the high mortality rate among cardiac patients. Automatical classification of electrocardiogram (ECG) beats plays an important role in such systems, but this issue is challenging because of the complexities of ECG signals. In literature, feature designing has been broadly-studied. However, such methodology is inevitably limited by the heuristics of hand-crafting process and the challenge of signals themselves. To address it, we treat the problem of ECG beat classification from the metric and measurement perspective. We propose a novel approach, named “Set-Based Discriminative Measure”, which first learns a discriminative metric space to ensure that intra-class distances are smaller than inter-class distances for ECG features in a global way, and then measures a new set-based dissimilarity in such learned space to cope with the local variation of samples. Experimental results have demonstrated the advantage of this approach in terms of effectiveness, robustness, and flexibility based on ECG beats from the MIT-BIH Arrhythmia Database.

Cardiac arrhythmia beat classification using DOST and PSO tuned SVM

BACKGROUND AND OBJECTIVE

The increase in the number of deaths due to cardiovascular diseases (CVDs) has gained significant attention from the study of electrocardiogram (ECG) signals. These ECG signals are studied by the experienced cardiologist for accurate and proper diagnosis, but it becomes difficult and time-consuming for long-term recordings. Various signal processing techniques are studied to analyze the ECG signal, but they bear limitations due to the non-stationary behavior of ECG signals. Hence, this study aims to improve the classification accuracy rate and provide an automated diagnostic solution for the detection of cardiac arrhythmias.

METHODS

The proposed methodology consists of four stages, i.e. filtering, R-peak detection, feature extraction and classification stages. In this study, Wavelet based approach is used to filter the raw ECG signal, whereas Pan-Tompkins algorithm is used for detecting the R-peak inside the ECG signal. In the feature extraction stage, discrete orthogonal Stockwell transform (DOST) approach is presented for an efficient time-frequency representation (i.e. morphological descriptors) of a time domain signal and retains the absolute phase information to distinguish the various non-stationary behavior ECG signals. Moreover, these morphological descriptors are further reduced in lower dimensional space by using principal component analysis and combined with the dynamic features (i.e based on RR-interval of the ECG signals) of the input signal. This combination of two different kinds of descriptors represents each feature set of an input signal that is utilized for classification into subsequent categories by employing PSO tuned support vector machines (SVM).

RESULTS

The proposed methodology is validated on the baseline MIT-BIH arrhythmia database and evaluated under two assessment schemes, yielding an improved overall accuracy of 99.18% for sixteen classes in the category-based and 89.10% for five classes (mapped according to AAMI standard) in the patient-based assessment scheme respectively to the state-of-art diagnosis. The results reported are further compared to the existing methodologies in literature.

CONCLUSIONS

The proposed feature representation of cardiac signals based on symmetrical features along with PSO based optimization technique for the SVM classifier reported an improved classification accuracy in both the assessment schemes evaluated on the benchmark MIT-BIH arrhythmia database and hence can be utilized for automated computer-aided diagnosis of cardiac arrhythmia beats.

A real-time approach for heart rate monitoring using a Hilbert transform in seismocardiograms

Heart rate monitoring helps in assessing the functionality and condition of the cardiovascular system. We present a new real-time applicable approach for estimating beat-to-beat time intervals and heart rate in seismocardiograms acquired from a tri-axial microelectromechanical accelerometer. Seismocardiography (SCG) is a non-invasive method for heart monitoring which measures the mechanical activity of the heart. Measuring true beat-to-beat time intervals from SCG could be used for monitoring of the heart rhythm, for heart rate variability analysis and for many other clinical applications. In this paper we present the Hilbert adaptive beat identification technique for the detection of heartbeat timings and inter-beat time intervals in SCG from healthy volunteers in three different positions, i.e. supine, left and right recumbent. Our method is electrocardiogram (ECG) independent, as it does not require any ECG fiducial points to estimate the beat-to-beat intervals. The performance of the algorithm was tested against standard ECG measurements. The average true positive rate, positive prediction value and detection error rate for the different positions were, respectively, supine (95.8%, 96.0% and ≃0.6%), left (99.3%, 98.8% and ≃0.001%) and right (99.53%, 99.3% and ≃0.01%). High correlation and agreement was observed between SCG and ECG inter-beat intervals (r  >  0.99) for all positions, which highlights the capability of the algorithm for SCG heart monitoring from different positions. Additionally, we demonstrate the applicability of the proposed method in smartphone based SCG. In conclusion, the proposed algorithm can be used for real-time continuous unobtrusive cardiac monitoring, smartphone cardiography, and in wearable devices aimed at health and well-being applications.

Signal Quality Analysis of Ambulatory Electrocardiograms to Gate False Myocardial Ischemia Alarms

OBJECTIVE

The objective of this study is to propose and validate an alarm gating system for a myocardial ischemia monitoring system that uses ambulatory electrocardiogram. The PeriOperative ISchemic Evaluation study recommended the selective administration of β blockers to patients at risk of cardiac events following noncardiac surgery. Patients at risk are identified by monitoring ST segment deviations in the electrocardiogram (ECG); however, patients are encouraged to ambulate to improve recovery, which deteriorates the signal quality of the ECG leading to false alarms.

METHODS

The proposed alarm gating system computes a signal quality index (SQI) to quantify the ECG signal quality and rejects alarms with a low SQI. The system was validated by artificially contaminating ECG records with motion artifact records obtained from the long-term ST database and MIT-BIH noise stress test database, respectively.

RESULTS

Without alarm gating, the myocardial ischemia monitoring system attained a Precision of 0.31 and a Recall of 0.78. The alarm gating improved the Precision to 0.58 with a reduction of Recall to 0.77.

CONCLUSION

The proposed system successfully gated false alarms with future work exploring the misidentification of fiducial points by myocardial ischemia monitoring systems.

SIGNIFICANCE

The reduction of false alarms due to the proposed system will decrease the incidence of the alarm fatigue condition typically found in clinicians. Alarm fatigue condition was rated as the top patient safety hazard from 2012 to 2015 by the Emergency Care Research Institute.

False alarm reduction in critical care

High false alarm rates in the ICU decrease quality of care by slowing staff response times while increasing patient delirium through noise pollution. The 2015 PhysioNet/Computing in Cardiology Challenge provides a set of 1250 multi-parameter ICU data segments associated with critical arrhythmia alarms, and challenges the general research community to address the issue of false alarm suppression using all available signals. Each data segment was 5 minutes long (for real time analysis), ending at the time of the alarm. For retrospective analysis, we provided a further 30 seconds of data after the alarm was triggered. A total of 750 data segments were made available for training and 500 were held back for testing. Each alarm was reviewed by expert annotators, at least two of whom agreed that the alarm was either true or false. Challenge participants were invited to submit a complete, working algorithm to distinguish true from false alarms, and received a score based on their program's performance on the hidden test set. This score was based on the percentage of alarms correct, but with a penalty that weights the suppression of true alarms five times more heavily than acceptance of false alarms. We provided three example entries based on well-known, open source signal processing algorithms, to serve as a basis for comparison and as a starting point for participants to develop their own code. A total of 38 teams submitted a total of 215 entries in this year's Challenge. This editorial reviews the background issues for this challenge, the design of the challenge itself, the key achievements, and the follow-up research generated as a result of the Challenge, published in the concurrent special issue of Physiological Measurement. Additionally we make some recommendations for future changes in the field of patient monitoring as a result of the Challenge.

Automated integration of wireless biosignal collection devices for patient-centred decision-making in point-of-care systems

The proper acquisition of biosignals data from various biosensor devices and their remote accessibility are still issues that prevent the wide adoption of point-of-care systems in the routine of monitoring chronic patients. This Letter presents an advanced framework for enabling patient monitoring that utilises a cloud computing infrastructure for data management and analysis. The framework introduces also a local mechanism for uniform biosignals collection from wearables and biosignal sensors, and decision support modules, in order to enable prompt and essential decisions. A prototype smartphone application and the related cloud modules have been implemented for demonstrating the value of the proposed framework. Initial results regarding the performance of the system and the effectiveness in data management and decision-making have been quite encouraging.