Automatic recognition of cardiac arrhythmias based on the geometric patterns of Poincaré plots

Use of machine learning and Poincaré density grid in the diagnosis of sinus node dysfunction caused by sinoatrial conduction block in dogs

BACKGROUND

Sinus node dysfunction because of abnormal impulse generation or sinoatrial conduction block causes bradycardia that can be difficult to differentiate from high parasympathetic/low sympathetic modulation (HP/LSM).

HYPOTHESIS

Beat-to-beat relationships of sinus node dysfunction are quantifiably distinguishable by Poincaré plots, machine learning, and 3-dimensional density grid analysis. Moreover, computer modeling establishes sinoatrial conduction block as a mechanism.

ANIMALS

Three groups of dogs were studied with a diagnosis of: (1) balanced autonomic modulation (n = 26), (2) HP/LSM (n = 26), and (3) sinus node dysfunction (n = 21).

METHODS

Heart rate parameters and Poincaré plot data were determined [median (25%-75%)]. Recordings were randomly assigned to training or testing. Supervised machine learning of the training data was evaluated with the testing data. The computer model included impulse rate, exit block probability, and HP/LSM.

RESULTS

Confusion matrices illustrated the effectiveness in diagnosing by both machine learning and Poincaré density grid. Sinus pauses >2 s differentiated (P < .0001) HP/LSM (2340; 583-3947 s) from sinus node dysfunction (8503; 7078-10 050 s), but average heart rate did not. The shortest linear intervals were longer with sinus node dysfunction (315; 278-323 ms) vs HP/LSM (260; 251-292 ms; P = .008), but the longest linear intervals were shorter with sinus node dysfunction (620; 565-698 ms) vs HP/LSM (843; 799-888 ms; P < .0001).

CONCLUSIONS

Number and duration of pauses, not heart rate, differentiated sinus node dysfunction from HP/LSM. Machine learning and Poincaré density grid can accurately identify sinus node dysfunction. Computer modeling supports sinoatrial conduction block as a mechanism of sinus node dysfunction.

A training pipeline of an arrhythmia classifier for atrial fibrillation detection using Photoplethysmography signal

Photoplethysmography (PPG) signal is potentially suitable in atrial fibrillation (AF) detection for its convenience in use and similarity in physiological origin to electrocardiogram (ECG). There are a few preceding studies that have shown the possibility of using the peak-to-peak interval of the PPG signal (PPIp) in AF detection. However, as a generalized model, the accuracy of an AF detector should be pursued on the one hand; on the other hand, its generalizability should be paid attention to in view of the individual differences in PPG manifestation of even the same arrhythmia and the existence of sub-types. Moreover, a binary classifier for atrial fibrillation and normal sinus rhythm is not convincing enough for the similarity between AF and ectopic beats. In this study, we project the atrial fibrillation detection as a multiple-class classification and try to propose a training pipeline that is advantageous both to the accuracy and generalizability of the classifier by designing and determining the configurable options of the pipeline, in terms of input format, deep learning model (with hyperparameter optimization), and scheme of transfer learning. With a rigorous comparison of the possible combinations of the configurable components in the pipeline, we confirmed that first-order difference of heartbeat sequence as the input format, a 2-layer CNN-1-layer Transformer hybrid^R model as the learning model and the whole model fine-tuning as the implementing scheme of transfer learning is the best combination for the pipeline (F1 value: 0.80, overall accuracy: 0.87)^R.

Heart Rate Variability Can Detect Blunt Traumatic Brain Injury Within the First Hour

INTRODUCTION

In patients with multi-organ system trauma, the diagnosis of coinciding traumatic brain injury can be difficult due to injuries from the hemorrhagic shock that confound clinical and radiographic signs of traumatic brain injury. In this study, a novel technique using heart rate variability was developed in a porcine model to detect traumatic brain injury early in the setting of hemorrhagic shock without the need for radiographic imaging or clinical exam.

METHODS

A porcine model of hemorrhagic shock was used with an arm of swine receiving hemorrhagic shock alone and hemorrhagic shock with traumatic brain injury. High-resolution heart rate frequencies were collected at different time intervals using waveforms based on voltage delivered from the heart rate monitor. Waveforms were analyzed to assess statistically significant differences between heart rate variability parameters in those with hemorrhagic shock and traumatic brain injury versus those with only hemorrhagic shock. Stochastic analysis was used to assess the validity of results and create a model by machine learning to better assess the presence of traumatic brain injury.

RESULTS

Significant differences were found in several heart rate variability parameters between the two groups. Additionally, significant differences in heart rate variability parameters were found in swine within 1 hour of inducing hemorrhage in those with traumatic brain injury versus those without. These results were confirmed with stochastic analysis and machine learning was used to generate a model which determined the presence of traumatic brain injury in the setting of hemorrhage shock with 91.6% accuracy.

CONCLUSIONS

 Heart rate variability represents a promising diagnostic tool to aid in the diagnosis of traumatic brain injury within 1 hour of injury.

Predictive performance and impact of algorithms in remote monitoring of chronic conditions: A systematic review and meta-analysis

BACKGROUND

The use of telehealth interventions, such as the remote monitoring of patient clinical data (e.g. blood pressure, blood glucose, heart rate, medication use), has been proposed as a strategy to better manage chronic conditions and to reduce the impact on patients and healthcare systems. The use of algorithms for data acquisition, analysis, transmission, communication and visualisation are now common in remote patient monitoring. However, their use and impact on chronic disease management has not been systematically investigated.

OBJECTIVES

To investigate the use, impact, and performance of remote monitoring algorithms across various types of chronic conditions.

METHODS

A literature search of MEDLINE complete, CINHAL complete, and EMBASE was performed using search terms relating to the concepts of remote monitoring, chronic conditions, and data processing algorithms. Comparable outcomes from studies describing the impact on process measures and clinical and patient-reported outcomes were pooled for a summary effect and meta-analyses. A comparison of studies reporting the predictive performance of algorithms was also conducted using the Youden Index.

RESULTS

A total of 89 articles were included in the review. There was no evidence of a positive impact on healthcare utilisation [OR 1.09 (0.90 to 1.31); P = .35] and mortality [OR 0.83 (0.63 to 1.10); P = .208], but there was a positive effect on generic health status [SDM 0.2912 (0.06 to 0.51); P = .010] and diabetes control [SDM -0.53 (-0.74 to -0.33); P < .001; I² = 15.71] (with two of the three diabetes studies being identified as having a high risk of bias). While the majority of impact studies made use of heuristic threshold-based algorithms (n = 27,87%), most performance studies (n = 36, 62%) analysed non-sequential machine learning methods. There was considerable variance in the quality, sample size and performance amongst these studies. Overall, algorithms involved in diagnosis (n = 22, 47%) had superior performance to those involved in predicting a future event (n = 25, 53%). Detection of arrythmia and ischaemia utilising ECG data showed particularly promising results.

CONCLUSION

The performance of data processing algorithms for the diagnosis of a current condition, particularly those related to the detection of arrythmia and ischaemia, is promising. However, there appears to exist minimal testing in experimental studies, with only two included impact studies citing a performance study as support for the intervention algorithm used. Because of the disconnect between performance and impact studies, there is currently limited evidence of the effect of integrating advanced inference algorithms in remote monitoring interventions. If the field of remote patient monitoring is to progress, future impact studies should address this disconnect by evaluating high performance validated algorithms in robust clinical trials.

Single channel photoplethysmography-based obstructive sleep apnea detection and arrhythmia classification

BACKGROUND

Simplified and easy-to-use monitoring approaches are crucial for the early diagnosis and prevention of obstructive sleep apnea (OSA) and its complications.

OBJECTIVE

In this study, the OSA detection and arrhythmia classification algorithms based on single-channel photoplethysmography (PPG) are proposed for the early screening of OSA.

METHODS

Thirty clinically diagnosed OSA patients participated in this study. Fourteen features were extracted from the PPG signals. The relationship between the number of features as inputs of the support vector machine (SVM) and performance of apnea events detection was evaluated. Also, a multi-classification algorithm based on the modified Hausdorff distance was proposed to recognize sinus rhythm and four arrhythmias highly related with SA.

RESULTS

The feature set composed of meanPP, SDPP, RMSSD, meanAm, and meank1 could provide a satisfactory balance between the performance and complexity of the algorithm for OSA detection. Also, the arrhythmia classification algorithm achieves the average sensitivity, specificity and accuracy of 83.79%, 95.91% and 93.47%, respectively in the classification of all four types of arrhythmia and regular rhythm.

CONCLUSION

Single channel PPG-based OSA detection and arrhythmia classification in this study can provide a feasible and promising approach for the early screening and diagnosis of OSA and OSA-related arrhythmias.

Poincaré Plot Image and Rhythm-Specific Atlas for Atrial Bigeminy and Atrial Fibrillation Detection

A detector based only on RR intervals capable of classifying other tachyarrhythmias in addition to atrial fibrillation (AF) could improve cardiac monitoring. In this paper a new classification method based in a 2D non-linear RRI dynamics representation is presented. For this aim, the concepts of Poincaré Images and Atlases are introduced. Three cardiac rhythms were targeted: Normal sinus rhythm (NSR), AF and atrial bigeminy (AB). Three Physionet open source databases were used. Poincaré Images were generated for all signals using different Poincaré plot configurations: RR, dRR and RRdRR. The study was computed for different time window lengths and bin sizes. For each rhythm, the Poincaré Images of the 80% of that rhythm's patients were used to create a reference image, a Poincaré Atlas. The remaining 20% were used as test set and classified into one of the three rhythms using normalized mutual information and 2D correlation. The process was iterated in a tenfold cross-validation and patient-wise dataset division. Sensitivity results obtained for RRdRR configuration and bin size 40 ms, for a 60 s time window were 94.35% ±3.68, 82.07% ±9.18 and 88.86% ±12.79 with a specificity of 85.52% ±7.46, 95.91% ±3.14, 96.10% ±2.25 for AF, NSR and AB respectively. Results suggest that a rhythms general RRI pattern may be captured using Poincaré Atlases and that these can be used to classify other signal segments using Poincaré Images. In contrast with other studies, the former method could be generalized to more cardiac rhythms and does not depend on rhythm-specific thresholds.

Heart rate variability analysis in horses for the diagnosis of arrhythmias

Heart rate variability (HRV) analysis has been performed on ECG-derived data sets for more than 170 years but is currently undergoing a rapid evolution, thanks to the expansion of the human and veterinary medical technology sector. Traditional HRV analysis was initially performed to identify changes in vago-sympathetic balance, while the most recent focus has expanded to include the use of complex computer algorithms, neural networks and machine learning technology to identify cardiac arrhythmias, particularly atrial fibrillation (AF). Some of these techniques have recently been translated for use in the field of equine cardiology, with particular focus on improving the diagnosis of arrhythmias both at rest and during exercise. This review focuses on understanding the basic HRV variables and important factors to consider when collecting data for use in HRV analysis. In addition, the use of HRV analysis for the diagnosis of arrhythmias is discussed from human, small animal and equine perspectives. Finally, the future of HRV analysis is briefly introduced, including an overview of future developments in this rapidly expanding and exciting field.

Visual representations of canine cardiac arrhythmias with Lorenz (Poincaré) plots

OBJECTIVE

To characterize the patterns associated with Lorenz plots (LPs) or Poincaré plots derived from the Holter recordings of dogs with various cardiac rhythms.

ANIMALS

77 dogs with 24-hour Holter recordings.

PROCEDURES

A 1-hour period from the Holter recordings from each of 20 dogs without arrhythmias and from each of 57 dogs with arrhythmias (10 each with supraventricular premature complexes, complex supraventricular ectopy, ventricular premature complexes, complex ventricular ectopy, and atrial fibrillation, and 7 with high-grade second-degree atrioventricular block) were used to generate the LPs. Patterns depicted in the LPs were described.

RESULTS

Arrhythmia-free Holter recordings yielded LPs with a Y-shaped pattern and variable silent zones. Recordings with single premature complexes yielded LPs with double side and triple side lobes. Complex ectopy was denoted by dots clustered in the lower left corner of the LPs. The LPs of recordings with atrial fibrillation had fan patterns consistent with a nonlinear relationship between atrial electrical impulses and atrioventricular nodal conduction. The recordings with atrioventricular block yielded LPs with island patterns consistent with variable atrioventricular nodal conduction.

CONCLUSIONS AND CLINICAL RELEVANCE

Distinct LP patterns were identified for common cardiac rhythms of dogs, supportive of nonrandom mechanisms as the cause of most rhythms. Visual interpretation of an LP generated from a Holter recording may aid in determining the arrhythmia type and understanding the arrhythmia's mechanism in dogs and other species.

Beat-to-Beat Patterning of Sinus Rhythm Reveals Non-linear Rhythm in the Dog Compared to the Human

The human and dog have sinus arrhythmia; however, the beat-to-beat interval changes were hypothesized to be different. Geometric analyses (R–R interval tachograms, dynamic Poincaré plots) to examine rate changes on a beat-to-beat basis were analyzed along with time and frequency domain heart rate variability from 40 human and 130 canine 24-h electrocardiographic recordings. Humans had bell-shaped beat-interval distributions, narrow interval bands across time with continuous interval change and linear changes in rate. In contrast, dogs had skewed non-singular beat distributions, wide interval bands {despite faster average heart rate of dogs [mean (range); 81 (64–119)] bpm compared to humans [74.5 (59–103) p = 0.005]} with regions displaying a paucity of intervals (zone of avoidance) and linear plus non-linear rate changes. In the dog, dynamic Poincaré plots showed linear rate changes as intervals prolonged until a point of divergence from the line of identity at a mean interval of 598.5 (95% CI: 583.5–613.5) ms (bifurcation interval). The dog had bimodal beat distribution during sleep with slower rates and greater variability than during active hours that showed singular interval distributions, higher rates and less variability. During sleep, Poincaré plots of the dog had clustered or branched patterns of intervals. A slower rate supported greater parasympathetic modulation with a branched compared to the clustered distribution. Treatment with atropine eliminated the non-linear patterns, while hydromorphone shifted the bifurcated branching and beat clustering to longer intervals. These results demonstrate the unique non-linear nature of beat-to-beat variability in the dog compared to humans with increases in interval duration (decrease heart rate). These results provoke the possibility that changes are linear with a dominant sympathetic modulation and non-linear with a dominant parasympathetic modulation. The abrupt bifurcation, zone of avoidance and beat-to-beat patterning are concordant with other studies demonstrating the development of exit block from the sinus node with parasympathetic modulation influencing not only the oscillation of the pacing cells, but conduction to the atria. Studies are required to associate the in vivo sinus node beat patterns identified in this study to the mapping of sinus impulse origin and exit from the sinus node.

Automated Method for Discrimination of Arrhythmias Using Time, Frequency, and Nonlinear Features of Electrocardiogram Signals

We developed an automated approach to differentiate between different types of arrhythmic episodes in electrocardiogram (ECG) signals, because, in real-life scenarios, a software application does not know in advance the type of arrhythmia a patient experiences. Our approach has four main stages: (1) Classification of ventricular fibrillation (VF) versus non-VF segments—including atrial fibrillation (AF), ventricular tachycardia (VT), normal sinus rhythm (NSR), and sinus arrhythmias, such as bigeminy, trigeminy, quadrigeminy, couplet, triplet—using four image-based phase plot features, one frequency domain feature, and the Shannon entropy index. (2) Classification of AF versus non-AF segments. (3) Premature ventricular contraction (PVC) detection on every non-AF segment, using a time domain feature, a frequency domain feature, and two features that characterize the nonlinearity of the data. (4) Determination of the PVC patterns, if present, to categorize distinct types of sinus arrhythmias and NSR. We used the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database, Creighton University’s VT arrhythmia database, the MIT-BIH atrial fibrillation database, and the MIT-BIH malignant ventricular arrhythmia database to test our algorithm. Binary decision tree (BDT) and support vector machine (SVM) classifiers were used in both stage 1 and stage 3. We also compared our proposed algorithm’s performance to other published algorithms. Our VF detection algorithm was accurate, as in balanced datasets (and unbalanced, in parentheses) it provided an accuracy of 95.1% (97.1%), sensitivity of 94.5% (91.1%), and specificity of 94.2% (98.2%). The AF detection was accurate, as the sensitivity and specificity in balanced datasets (and unbalanced, in parentheses) were found to be 97.8% (98.6%) and 97.21% (97.1%), respectively. Our PVC detection algorithm was also robust, as the accuracy, sensitivity, and specificity were found to be 99% (98.1%), 98.0% (96.2%), and 98.4% (99.4%), respectively, for balanced and (unbalanced) datasets.

Use of self-gated radial cardiovascular magnetic resonance to detect and classify arrhythmias (atrial fibrillation and premature ventricular contraction)

BACKGROUND

Arrhythmia can significantly alter the image quality of cardiovascular magnetic resonance (CMR); automatic detection and sorting of the most frequent types of arrhythmias during the CMR acquisition could potentially improve image quality. New CMR techniques, such as non-Cartesian CMR, can allow self-gating: from cardiac motion-related signal changes, we can detect cardiac cycles without an electrocardiogram. We can further use this data to obtain a surrogate for RR intervals (valley intervals: VV). Our purpose was to evaluate the feasibility of an automated method for classification of non-arrhythmic (NA) (regular cycles) and arrhythmic patients (A) (irregular cycles), and for sorting of common arrhythmia patterns between atrial fibrillation (AF) and premature ventricular contraction (PVC), using the cardiac motion-related signal obtained during self-gated free-breathing radial cardiac cine CMR with compressed sensing reconstruction (XD-GRASP).

METHODS

One hundred eleven patients underwent cardiac XD-GRASP CMR between October 2015 and February 2016; 33 were included for retrospective analysis with the proposed method (6 AF, 8 PVC, 19 NA; by recent ECG). We analyzed the VV, using pooled statistics (histograms) and sequential analysis (Poincaré plots), including the median (medVV), the weighted mean (meanVV), the total number of VV values (VVval), and the total range (VVTR) and half range (VVHR) of the cumulative frequency distribution of VV, including the median to half range (medVV/VVHR) and the half range to total range (VVHR/VVTR) ratios. We designed a simple algorithm for using the VV results to differentiate A from NA, and AF from PVC.

RESULTS

Between NA and A, meanVV, VVval, VVTR, VVHR, medVV/VVHR and VVHR/VVTR ratios were significantly different (p values = 0.00014, 0.0027, 0.000028, 5×10-9, 0.002, respectively). Between AF and PVC, meanVV, VVval and medVV/VVHR ratio were significantly different (p values = 0.018, 0.007, 0.044, respectively). Using our algorithm, sensitivity, specificity, and accuracy were 93 %, 95 % and 94 % to discriminate between NA and A, and 83 %, 71 %, and 77 % to discriminate between AF and PVC, respectively; areas under the ROC curve were 0.93 and 0.89.

CONCLUSIONS

Our study shows we can reliably detect arrhythmias and differentiate AF from PVC, using self-gated cardiac cine XD-GRASP CMR.

ECG feature extraction based on the bandwidth properties of variational mode decomposition

It is a difficult process to detect abnormal heart beats, known as arrhythmia, in long-term ECG recording. Thus, computer-aided diagnosis systems have become a supportive tool for helping physicians improve the diagnostic accuracy of heartbeat detection. This paper explores the bandwidth properties of the modes obtained using variational mode decomposition (VMD) to classify arrhythmia electrocardiogram (ECG) beats. VMD is an enhanced version of the empirical mode decomposition (EMD) algorithm for analyzing non-linear and non-stationary signals. It decomposes the signal into a set of band-limited oscillations called modes. ECG signals from the MIT-BIH arrhythmia database are decomposed using VMD, and the amplitude modulation bandwidth B AM, the frequency modulation bandwidth B FM and the total bandwidth B of the modes are used as feature vectors to detect heartbeats such as normal (N), premature ventricular contraction (V), left bundle branch block (L), right bundle branch block (R), paced beat (P) and atrial premature beat (A). Bandwidth estimations based on the instantaneous frequency (IF) and amplitude (IA) spectra of the modes indicate that the proposed VMD-based features have sufficient class discrimination capability regarding ECG beats. Moreover, the extracted features using the bandwidths (B AM, B FM and B) of four modes are used to evaluate the diagnostic accuracy rates of several classifiers such as the k-nearest neighbor classifier (k-NN), the decision tree (DT), the artificial neural network (ANN), the bagged decision tree (BDT), the AdaBoost decision tree (ABDT) and random sub-spaced k-NN (RSNN) for N, R, L, V, P, and A beats. The performance of the proposed VMD-based feature extraction with a BDT classifier has accuracy rates of 99.06%, 99.00%, 99.40%, 99.51%, 98.72%, 98.71%, and 99.02% for overall, N-, R-, L-, V-, P-, and A-type ECG beats, respectively.

Generalized Poincaré Plots-A New Method for Evaluation of Regimes in Cardiac Neural Control in Atrial Fibrillation and Healthy Subjects

Classical Poincaré plot is a standard way to measure nonlinear regulation of cardiovascular control. In our work we propose a generalized form of Poincaré plot where we track correlation between the duration of j preceding and k next RR intervals. The investigation was done in healthy subjects and patients with atrial fibrillation, by varying j,k ≤ 100. In cases where j = k, in healthy subjects the typical pattern was observed by “paths” that were substituting scatterplots and that were initiated and ended by loops of Poincaré plot points. This was not the case for atrial fibrillation patients where Poincaré plot had a simple scattered form. More, a typical matrix of Pearson's correlation coefficients, r(j,k), showed different positions of local maxima, depending on the subject's health condition. In both groups, local maxima were grouped into four clusters which probably determined specific regulatory mechanisms according to correlations between the duration of symmetric and asymmetric observed RR intervals. We quantified matrices' degrees of asymmetry and found that they were significantly different: distributed around zero in healthy, while being negative in atrial fibrillation. Also, Pearson's coefficients were higher in healthy than in atrial fibrillation or in signals with reshuffled intervals. Our hypothesis is that by this novel method we can observe heart rate regimes typical for baseline conditions and “defense reaction” in healthy subjects. These data indicate that neural control mechanisms of heart rate are operating in healthy subjects in contrast with atrial fibrillation, identifying it as the state of risk for stress-dependent pathologies. Regulatory regimes of heart rate can be further quantified and explored by the proposed novel method.