Post-Ventricular Premature Contraction Phase Correction Improves the Predictive Value of Average T-Wave Alternans in Ambulatory ECG Recordings

Machine Learning approach for TWA detection relying on ensemble data design

Background and objective

T-wave alternans (TWA) is a fluctuation of the ST-T complex of the surface electrocardiogram (ECG) on an every-other-beat basis. It has been shown to be clinically helpful for sudden cardiac death stratification, though the lack of a gold standard to benchmark detection methods limits its application and impairs the development of alternative techniques. In this work, a novel approach based on machine learning for TWA detection is proposed. Additionally, a complete experimental setup is presented for TWA detection methods benchmarking.

Methods

The proposed experimental setup is based on the use of open-source databases to enable experiment replication and the use of real ECG signals with added TWA episodes. Also, intra-patient overfitting and class imbalance have been carefully avoided. The Spectral Method (SM), the Modified Moving Average Method (MMA), and the Time Domain Method (TM) are used to obtain input features to the Machine Learning (ML) algorithms, namely, K Nearest Neighbor, Decision Trees, Random Forest, Support Vector Machine and Multi-Layer Perceptron.

Results

There were not found large differences in the performance of the different ML algorithms. Decision Trees showed the best overall performance (accuracy 0.88 ± 0.04 , precision 0.89 ± 0.05 , Recall 0.90 ± 0.05 , F1 score 0.89 ± 0.03 ). Compared to the SM (accuracy 0.79, precision 0.93, Recall 0.64, F1 score 0.76) there was an improvement in every metric except for the precision.

Conclusions

In this work, a realistic database to test the presence of TWA using ML algorithms was assembled. The ML algorithms overall outperformed the SM used as a gold standard. Learning from data to identify alternans elicits a substantial detection growth at the expense of a small increment of the false alarm.

An Improved UNet++ Model for Congestive Heart Failure Diagnosis Using Short-Term RR Intervals

Congestive heart failure (CHF), a progressive and complex syndrome caused by ventricular dysfunction, is difficult to detect at an early stage. Heart rate variability (HRV) was proposed as a prognostic indicator for CHF. Inspired by the success of 2-D UNet++ in medical image segmentation, in this paper, we introduce an end-to-end encoder-decoder model to detect CHF using HRV signals. The developed model enhances the UNet++ model with Squeeze-and-Excitation (SE) residual blocks to extract deep features hierarchically and distinguish CHF patients from normal subjects. Two open-source databases are utilized for evaluating the proposed method, and three segment lengths of intervals between successive R-peaks are employed in comparison with state-of-the-art methods. The proposed method achieves an accuracy of 85.64%, 86.65% and 88.79% when 500, 1000 and 2000 RR intervals are utilized, respectively. It demonstrates that HRV evaluation based on deep learning can be an important tool for early detection of CHF, and may assist clinicians in achieving timely and accurate diagnoses.

Pulse Arrival Time and Pulse Interval as Accurate Markers to Detect Mechanical Alternans

Mechanical alternans (MA) is a powerful predictor of adverse prognosis in patients with heart failure and cardiomyopathy, but its use remains limited due to the need of invasive continuous arterial pressure recordings. This study aims to assess novel cardiovascular correlates of MA in the intact human heart to facilitate affordable and non-invasive detection of MA and advance our understanding of the underlying pathophysiology. Arterial pressure, respiration, and ECG were recorded in 12 subjects with healthy ventricles during voluntarily controlled breathing at different respiratory rate, before and after administration of beta-blockers. MA was induced by ventricular pacing. A total of 67 recordings lasting approximately 90 s each were analyzed. Mechanical alternans (MA) was measured in the systolic blood pressure. We studied cardiovascular correlates of MA, including maximum pressure rise during systole (dPdtmax), pulse arrival time (PAT), pulse wave interval (PI), RR interval (RRI), ECG QRS complexes and T-waves. MA was detected in 30% of the analyzed recordings. Beta-blockade significantly reduced MA prevalence (from 50 to 11%, p < 0.05). Binary classification showed that MA was detected by alternans in dPdtmax (100% sens, 96% spec), PAT (100% sens, 81% spec) and PI (80% sens, 81% spec). Alternans in PAT and in PI also showed high degree of temporal synchronization with MA (80 ± 33 and 73 ± 40%, respectively). These data suggest that cardiac contractility is a primary factor in the establishment of MA. Our findings show that MA was highly correlated with invasive measurements of PAT and PI. Since PAT and PI can be estimated using non-invasive technologies, these markers could potentially enable affordable MA detection for risk-prediction.

Quantification of Ventricular Repolarization Variation for Sudden Cardiac Death Risk Stratification in Atrial Fibrillation

OBJECTIVE

Atrial fibrillation (AF) rhythm gives rise to an irregular response in ventricular activity, preventing the use of standard ECG-derived risk markers based on ventricular repolarization heterogeneity under this particular condition. In this study, we proposed new indices to quantify repolarization variations in AF patients, assessing their stratification performance in a chronic heart failure population with AF.

METHODS

We developed a method based on a selective bin averaging technique. Consecutive beats preceded by a similar RR interval were selected, from which the average variation within the ST-T complex for each RR range was computed. We proposed two sets of indices: 1) the 2-beat index of ventricular repolarization variation, ( I_V2), computed from pairs of stable consecutive beats; and 2) the 3-beat indices of ventricular repolarization variation, computed in triplets of stable consecutive beats ( I_V3).

RESULTS

These indices showed a significant association with sudden cardiac death (SCD) outcome in the study population. In addition, risk assessment based on the combination of the proposed indices improved stratification performance compared to their individual potential.

CONCLUSION

Patients with enhanced ventricular repolarization variation computed in terms of the proposed indices were successfully associated to a higher SCD incidence in our study population, evidencing their prognostic value.

SIGNIFICANCE

using a simple ambulatory ECG recording, it is possible to stratify AF patients at risk of SCD, which may help cardiologists in adopting most effective therapeutic strategies, with a positive impact in both the patient and healthcare systems.