LVSP and LBBP Result in Similar or Improved LV Synchrony and Hemodynamics Compared to BVP.

BACKGROUND: The effect of left ventricular septal myocardial pacing (LVSP) and left bundle branch pacing (LBBP) on ventricular synchrony and left ventricular (LV) hemodynamic status is poorly understood. OBJECTIVES: The aim of this study was to investigate the impact of LVSP and LBBP vs biventricular pacing (BVP) on ventricular electrical synchrony and hemodynamic status in cardiac resynchronization therapy patients. METHODS: In cardiac resynchronization therapy candidates with LV conduction disease, ventricular synchrony was assessed by measuring QRS duration (QRSd) and using ultra-high-frequency electrocardiography. LV electrical dyssynchrony was assessed as the difference between the first activation in leads V1 to V8 to the last from leads V4 to V8. LV hemodynamic status was estimated using invasive systolic blood pressure measurement during multiple transitions between LBBP, LVSP, and BVP. RESULTS: A total of 35 patients with a mean LV ejection fraction of 29% and a mean QRSd of 168 ± 24 ms were included. Thirteen had ischemic cardiomyopathy. QRSd during BVP, LVSP, and LBBP was the same, but LBBP provided shorter LV electrical dyssynchrony than BVP (-10 ms; 95% CI: -16 to -4 ms; P = 0.001); the difference between LVSP and BVP was not significant (-5 ms; 95% CI: -12 to 1 ms; P = 0.10). LBBP was associated with higher systolic blood pressure than BVP (4%; 95% CI: 2% to 5%; P < 0.001), whereas LVSP was not (1%; 95% CI: 0% to 2%; P = 0.10). Hemodynamic differences during LBBP and LVSP vs BVP were more pronounced in nonischemic than ischemic patients. CONCLUSIONS: Ultra-high-frequency electrocardiography allowed the documentation of differences in LV synchrony between LBBP, LVSP, and BVP, which were not observed by measuring QRSd. LVSP provided the same LV synchrony and hemodynamic status as BVP, while LBBP was better than BVP in both.

Ultra-high-frequency ECG volumetric and negative derivative epicardial ventricular electrical activation pattern.

From precordial ECG leads, the conventional determination of the negative derivative of the QRS complex (ND-ECG) assesses epicardial activation. Recently we showed that ultra-high-frequency electrocardiography (UHF-ECG) determines the activation of a larger volume of the ventricular wall. We aimed to combine these two methods to investigate the potential of volumetric and epicardial ventricular activation assessment and thereby determine the transmural activation sequence. We retrospectively analyzed 390 ECG records divided into three groups-healthy subjects with normal ECG, left bundle branch block (LBBB), and right bundle branch block (RBBB) patients. Then we created UHF-ECG and ND-ECG-derived depolarization maps and computed interventricular electrical dyssynchrony. Characteristic spatio-temporal differences were found between the volumetric UHF-ECG activation patterns and epicardial ND-ECG in the Normal, LBBB, and RBBB groups, despite the overall high correlations between both methods. Interventricular electrical dyssynchrony values assessed by the ND-ECG were consistently larger than values computed by the UHF-ECG method. Noninvasively obtained UHF-ECG and ND-ECG analyses describe different ventricular dyssynchrony and the general course of ventricular depolarization. Combining both methods based on standard 12-lead ECG electrode positions allows for a more detailed analysis of volumetric and epicardial ventricular electrical activation, including the assessment of the depolarization wave direction propagation in ventricles.

Bipolar anodal septal pacing with direct LBB capture preserves physiological ventricular activation better than unipolar left bundle branch pacing.

Background: Left bundle branch pacing (LBBP) produces delayed, unphysiological activation of the right ventricle. Using ultra-high-frequency electrocardiography (UHF-ECG), we explored how bipolar anodal septal pacing with direct LBB capture (aLBBP) affects the resultant ventricular depolarization pattern. Methods: In patients with bradycardia, His bundle pacing (HBP), unipolar nonselective LBBP (nsLBBP), aLBBP, and right ventricular septal pacing (RVSP) were performed. Timing of local ventricular activation, in leads V1-V8, was displayed using UHF-ECG, and electrical dyssynchrony (e-DYS) was calculated as the difference between the first and last activation. Durations of local depolarizations were determined as the width of the UHF-QRS complex at 50% of its amplitude. Results: aLBBP was feasible in 63 of 75 consecutive patients with successful nsLBBP. aLBBP significantly improved ventricular dyssynchrony (mean -9 ms; 95% CI (-12;-6) vs. -24 ms (-27;-21), ), p < 0.001) and shortened local depolarization durations in V1-V4 (mean differences -7 ms to -5 ms (-11;-1), p < 0.05) compared to nsLBBP. aLBBP resulted in e-DYS -9 ms (-12; -6) vs. e-DYS 10 ms (7;14), p < 0.001 during HBP. Local depolarization durations in V1-V2 during aLBBP were longer than HBP (differences 5-9 ms (1;14), p < 0.05, with local depolarization duration in V1 during aLBBP being the same as during RVSP (difference 2 ms (-2;6), p = 0.52). Conclusion: Although aLBBP improved ventricular synchrony and depolarization duration of the septum and RV compared to unipolar nsLBBP, the resultant ventricular depolarization was still less physiological than during HBP.

QRS detection and classification in Holter ECG data in one inference step.

While various QRS detection and classification methods were developed in the past, the Holter ECG data acquired during daily activities by wearable devices represent new challenges such as increased noise and artefacts due to patient movements. Here, we present a deep-learning model to detect and classify QRS complexes in single-lead Holter ECG. We introduce a novel approach, delivering QRS detection and classification in one inference step. We used a private dataset (12,111 Holter ECG recordings, length of 30 s) for training, validation, and testing the method. Twelve public databases were used to further test method performance. We built a software tool to rapidly annotate QRS complexes in a private dataset, and we annotated 619,681 QRS complexes. The standardised and down-sampled ECG signal forms a 30-s long input for the deep-learning model. The model consists of five ResNet blocks and a gated recurrent unit layer. The model's output is a 30-s long 4-channel probability vector (no-QRS, normal QRS, premature ventricular contraction, premature atrial contraction). Output probabilities are post-processed to receive predicted QRS annotation marks. For the QRS detection task, the proposed method achieved the F1 score of 0.99 on the private test set. An overall mean F1 cross-database score through twelve external public databases was 0.96 ± 0.06. In terms of QRS classification, the presented method showed micro and macro F1 scores of 0.96 and 0.74 on the private test set, respectively. Cross-database results using four external public datasets showed micro and macro F1 scores of 0.95 ± 0.03 and 0.73 ± 0.06, respectively. Presented results showed that QRS detection and classification could be reliably computed in one inference step. The cross-database tests showed higher overall QRS detection performance than any of compared methods.

Classification of ECG using ensemble of residual CNNs with or without attention mechanism.

Objective. This paper introduces a winning solution (team ISIBrno-AIMT) to the official round of PhysioNet Challenge 2021. The main goal of the challenge was a classification of ECG recordings into 26 multi-label pathological classes with a variable number of leads (e.g. 12, 6, 4, 3, 2). The main objective of this study is to verify whether the multi-head-attention mechanism influences the model performance.Approach. We introduced an ECG classification method based on the ResNet architecture with a multi-head attention mechanism for the official round of the challenge. However, empirical findings collected during model development suggested that the multi-head attention layer might not significantly impact the final classification performance. For this reason, during the follow-up round, we removed a multi-head attention layer to test the influence on model performance. Like the official round, the model is optimized using a mixture of loss functions, i.e. binary cross-entropy, custom challenge score loss function, and custom sparsity loss function. Probability thresholds for each classification class are estimated using the evolutionary optimization method. The final architecture consists of three submodels forming a majority voting classification ensemble.Main results. The modified model without the multi-head attention layer increased the overall challenge score to 0.59 compared to the 0.58 from the official round.Significance. Our findings from the follow-up submission support the fact that the multi-head attention layer in the proposed architecture does not significantly affect the classification performance.

Arterial Aging Best Reflected in Pulse Wave Velocity Measured from Neck to Lower Limbs: A Whole-Body Multichannel Bioimpedance Study.

Pulse wave velocity is a commonly used parameter for evaluating arterial stiffness and the overall condition of the cardiovascular system. The main goal of this study was to establish a methodology to test and validate multichannel bioimpedance as a suitable method for whole-body evaluations of pulse waves. We set the proximal location over the left carotid artery and eight distal locations on both the upper and lower limbs. In this way, it was possible to simultaneously evaluate pulse wave velocity (PWV) in the upper and lower limbs and in the limbs via four extra PWV measurements. Data were acquired from a statistical group of 220 healthy subjects who were divided into three age groups. The data were then analysed. We found a significant dependency of aortic PWV on age in those values measured using the left carotid as the proximal. PWV values in the upper and lower limbs were found to have no significant dependency on age. In addition, the PWV in the left femoral artery shows comparable values to published already carotid-femoral values. Those findings prove the reliability of whole-body multichannel bioimpedance for pulse wave velocity evaluation and provide reference values for whole-body PWV measurement.

Left Ventricular Myocardial Septal Pacing in Close Proximity to LBB Does Not Prolong the Duration of the Left Ventricular Lateral Wall Depolarization Compared to LBB Pacing.

Background: Three different ventricular capture types are observed during left bundle branch pacing (LBBp). They are selective LBB pacing (sLBBp), non-selective LBB pacing (nsLBBp), and myocardial left septal pacing transiting from nsLBBp while decreasing the pacing output (LVSP). Study aimed to compare differences in ventricular depolarization between these captures using ultra-high-frequency electrocardiography (UHF-ECG). Methods: Using decremental pacing voltage output, we identified and studied nsLBBp, sLBBp, and LVSP in patients with bradycardia. Timing of ventricular activations in precordial leads was displayed using UHF-ECGs, and electrical dyssynchrony (e-DYS) was calculated as the difference between the first and last activation. The durations of local depolarizations (Vd) were determined as the width of the UHF-QRS complex at 50% of its amplitude. Results: In 57 consecutive patients, data were collected during nsLBBp (n = 57), LVSP (n = 34), and sLBBp (n = 23). Interventricular dyssynchrony (e-DYS) was significantly lower during LVSP -16 ms (-21; -11), than nsLBBp -24 ms (-28; -20) and sLBBp -31 ms (-36; -25). LVSP had the same V1d-V8d as nsLBBp and sLBBp except for V3d, which during LVSP was shorter than sLBBp; the mean difference -9 ms (-16; -1), p = 0.01. LVSP caused less interventricular dyssynchrony and the same or better local depolarization durations than nsLBBp and sLBBp irrespective of QRS morphology during spontaneous rhythm or paced QRS axis. Conclusions: In patients with bradycardia, LVSP in close proximity to LBB resulted in better interventricular synchrony than nsLBBp and sLBBp and did not significantly prolong depolarization of the left ventricular lateral wall.

3-Dimensional ventricular electrical activation pattern assessed from a novel high-frequency electrocardiographic imaging technique: principles and clinical importance.

The study introduces and validates a novel high-frequency (100-400 Hz bandwidth, 2 kHz sampling frequency) electrocardiographic imaging (HFECGI) technique that measures intramural ventricular electrical activation. Ex-vivo experiments and clinical measurements were employed. Ex-vivo, two pig hearts were suspended in a human-torso shaped tank using surface tank electrodes, epicardial electrode sock, and plunge electrodes. We compared conventional epicardial electrocardiographic imaging (ECGI) with intramural activation by HFECGI and verified with sock and plunge electrodes. Clinical importance of HFECGI measurements was performed on 14 patients with variable conduction abnormalities. From 3 × 4 needle and 108 sock electrodes, 256 torso or 184 body surface electrodes records, transmural activation times, sock epicardial activation times, ECGI-derived activation times, and high-frequency activation times were computed. The ex-vivo transmural measurements showed that HFECGI measures intramural electrical activation, and ECGI-HFECGI activation times differences indicate endo-to-epi or epi-to-endo conduction direction. HFECGI-derived volumetric dyssynchrony was significantly lower than epicardial ECGI dyssynchrony. HFECGI dyssynchrony was able to distinguish between intraventricular conduction disturbance and bundle branch block patients.

Left bundle branch pacing compared to left ventricular septal myocardial pacing increases interventricular dyssynchrony but accelerates left ventricular lateral wall depolarization.

BACKGROUND: Nonselective His-bundle pacing (nsHBp), nonselective left bundle branch pacing (nsLBBp), and left ventricular septal myocardial pacing (LVSP) are recognized as physiological pacing techniques. OBJECTIVE: The purpose of this study was to compare differences in ventricular depolarization between these techniques using ultra-high-frequency electrocardiography (UHF-ECG). METHODS: In patients with bradycardia, nsHBp, nsLBBp (confirmed concomitant left bundle branch [LBB] and myocardial capture), and LVSP (pacing in left ventricular [LV] septal position without proven LBB capture) were performed. Timings of ventricular activations in precordial leads were displayed using UHF-ECG, and electrical dyssynchrony (e-DYS) was calculated as the difference between the first and last activation. Duration of local depolarization (Vd) was determined as width of the UHF-QRS complex at 50% of its amplitude. RESULTS: In 68 patients, data were collected during nsLBBp (35), LVSP (96), and nsHBp (55). nsLBBp resulted in larger e-DYS than did LVSP and nsHBp [- 24 ms (-28;-19) vs -12 ms (-16;-9) vs 10 ms (7;14), respectively; P <.001]. nsLBBp produced similar values of Vd in leads V5-V8 (36-43 ms vs 38-43 ms; P = NS in all leads) but longer Vd in leads V1-V4 (47-59 ms vs 41-44 ms; P <.05) as nsHBp. LVSP caused prolonged Vd in leads V1-V8 compared to nsHBp and longer Vd in leads V5-V8 compared to nsLBBp (44-51 ms vs 36-43 ms; P <.05) regardless of R-wave peak time in lead V5 or QRS morphology in lead V1 present during LVSP. CONCLUSION: nslbbp preserves physiological LV depolarization but increases interventricular electrical dyssynchrony. LV lateral wall depolarization during LVSP is prolonged, but interventricular synchrony is preserved.

Ventricular activation pattern assessment during right ventricular pacing: Ultra-high-frequency ECG study.

BACKGROUND: Right ventricular (RV) pacing causes delayed activation of remote ventricular segments. We used the ultra-high-frequency ECG (UHF-ECG) to describe ventricular depolarization when pacing different RV locations. METHODS: In 51 patients, temporary pacing was performed at the RV septum (mSp); further subclassified as right ventricular inflow tract (RVIT) and right ventricular outflow tract (RVOT) for septal inflow and outflow positions (below or above the plane of His bundle in right anterior oblique), apex, anterior lateral wall, and at the basal RV septum with nonselective His bundle or RBB capture (nsHBorRBBp). The timings of UHF-ECG electrical activations were quantified as left ventricular lateral wall delay (LVLWd; V8 activation delay) and RV lateral wall delay (RVLWd; V1 activation delay). RESULTS: The LVLWd was shortest for nsHBorRBBp (11 ms [95% confidence interval = 5-17]), followed by the RVIT (19 ms [11-26]) and the RVOT (33 ms [27-40]; p < .01 between all of them), although the QRSd for the latter two were the same (153 ms (148-158) vs. 153 ms (148-158); p = .99). RV apical capture not only had a longer LVLWd (34 ms (26-43) compared to mSp (27 ms (20-34), p < .05), but its RVLWd (17 ms (9-25) was also the longest compared to other RV pacing sites (mean values for nsHBorRBBp, mSp, anterior and lateral wall captures being below 6 ms), p < .001 compared to each of them. CONCLUSION: RVIT pacing produces better ventricular synchrony compared to other RV pacing locations with myocardial capture. However, UHF-ECG ventricular dysynchrony seen during RVIT pacing is increased compared to concomitant capture of basal septal myocytes and His bundle or proximal right bundle branch.

Multicenter intracranial EEG dataset for classification of graphoelements and artifactual signals.

EEG signal processing is a fundamental method for neurophysiology research and clinical neurology practice. Historically the classification of EEG into physiological, pathological, or artifacts has been performed by expert visual review of the recordings. However, the size of EEG data recordings is rapidly increasing with a trend for higher channel counts, greater sampling frequency, and longer recording duration and complete reliance on visual data review is not sustainable. In this study, we publicly share annotated intracranial EEG data clips from two institutions: Mayo Clinic, MN, USA and St. Anne's University Hospital Brno, Czech Republic. The dataset contains intracranial EEG that are labeled into three groups: physiological activity, pathological/epileptic activity, and artifactual signals. The dataset published here should support and facilitate training of generalized machine learning and digital signal processing methods for intracranial EEG and promote research reproducibility. Along with the data, we also propose a statistical method that is recommended for comparison of candidate classifier performance utilizing out-of-institution/out-of-patient testing.

Both selective and nonselective His bundle, but not myocardial, pacing preserve ventricular electrical synchrony assessed by ultra-high-frequency ECG.

BACKGROUND: Right ventricular myocardial pacing leads to nonphysiological activation of heart ventricles. Contrary to this, His bundle pacing preserves their fast activation. Ultra-high-frequency electrocardiography (UHF-ECG) is a novel tool for ventricular depolarization assessment. OBJECTIVE: The purpose of this study was to describe UHF-ECG depolarization patterns during myocardial and His bundle pacing. METHODS: Forty-six patients undergoing His bundle pacing to treat bradycardia and spontaneous QRS complexes without bundle branch block were included. UHF-ECG recordings were performed during spontaneous rhythm, pure myocardial para-Hisian capture, and His bundle capture. QRS duration, QRS area, depolarization time in specific leads, and the UHF-ECG-derived ventricular dyssynchrony index were calculated. RESULTS: One hundred thirty-three UHF-ECG recordings were performed in 46 patients (44 spontaneous rhythm, 28 selective His bundle, 43 nonselective His bundle, and 18 myocardial capture). The mean QRS duration was 117 ms for spontaneous rhythm, 118 ms for selective, 135 ms for nonselective, and 166 ms for myocardial capture (P < .001 for nonselective and myocardial capture compared to each of the other types of ventricular activation). The calculated dyssynchrony index was shortest during spontaneous rhythm (12 ms; P = .02 compared to selective and P = .09 compared to nonselective), and it did not differ between selective and nonselective His bundle capture (16 vs 15 ms; P > .99) and was longest during myocardial capture of the para-Hisian area (37 ms; P < .001 compared to each of the other types of ventricular activation). CONCLUSION: In patients without bundle branch block, both types of His bundle, but not myocardial, capture preserve ventricular electrical synchrony as measured using UHF-ECG.

Novel ultra-high-frequency electrocardiogram tool for the description of the ventricular depolarization pattern before and during cardiac resynchronization.

INTRODUCTION: The present study introduces a new ultra-high-frequency 14-lead electrocardiogram technique (UHF-ECG) for mapping ventricular depolarization patterns and calculation of novel dyssynchrony parameters that may improve the selection of patients and application of cardiac resynchronization therapy (CRT). METHODS: Components of the ECG in sixteen frequency bands within the 150 to 1000 Hz range were used to create ventricular depolarization maps. The maximum time difference between the UHF QRS complex centers of mass of leads V1 to V8 was defined as ventricular electrical dyssynchrony (e-DYS), and the duration at 50% of peak voltage amplitude in each lead was defined as the duration of local depolarization (Vd). Proof of principle measurements was performed in seven patients with left (left bundle branch block) and four patients with right bundle branch block (right bundle branch block) before and during CRT using biventricular and His-bundle pacing. RESULTS: The acquired activation maps reflect the activation sequence under the tested conditions. e-DYS decreased considerably more than QRS duration, during both biventricular pacing (-50% vs -8%) and His-bundle pacing (-77% vs -13%). While biventricular pacing slightly increased Vd, His-bundle pacing reduced Vd significantly (+11% vs -36%), indicating the contribution of the fast conduction system. Optimization of biventricular pacing by adjusting VV-interval showed a decrease of e-DYS from 102 to 36 ms with only a small Vd increase and QRS duration decrease. CONCLUSIONS: The UHF-ECG technique provides novel information about electrical activation of the ventricles from a standard ECG electrode setup, potentially improving the selection of patients for CRT and application of CRT.

The relationship between ECG predictors of cardiac resynchronization therapy benefit.

INTRODUCTION: Cardiac resynchronization therapy (CRT) is an effective treatment that reduces mortality and improves cardiac function in patients with left bundle branch block (LBBB). However, about 30% of patients passing the current criteria do not benefit or benefit only a little from CRT. Three predictors of benefit based on different ECG properties were compared: 1) "strict" left bundle branch block classification (SLBBB); 2) QRS area; 3) ventricular electrical delay (VED) which defines the septal-lateral conduction delay. These predictors have never been analyzed concurrently. We analyzed the relationship between them on a subset of 602 records from the MADIT-CRT trial. METHODS & RESULTS: SLBBB classification was performed by two experts; QRS area and VED were computed fully automatically. High-frequency QRS (HFQRS) maps were used to inspect conduction abnormalities. The correlation between SLBBB and other predictors was R = 0.613, 0.523 and 0.390 for VED, QRS area in Z lead, and QRS duration, respectively. Scatter plots were used to pick up disagreement between the predictors. The majority of SLBBB subjects- 295 of 330 (89%)-are supposed to respond positively to CRT according to the VED and QRS area, though 93 of 272 (34%) non-SLBBB should also benefit from CRT according to the VED and QRS area. CONCLUSION: SLBBB classification is limited by the proper setting of cut-off values. In addition, it is too "strict" and excludes patients that may benefit from CRT therapy. QRS area and VED are clearly defined parameters. They may be used to optimize biventricular stimulation. Detailed analysis of conduction irregularities with CRT optimization should be based on HFQRS maps.

Comparison of noninvasive pulse transit time determined from Doppler aortic flow and multichannel bioimpedance plethysmography.

The main goal of this study was to make a comparison of aortic flow timing obtained by PW Doppler in four aortic sections with timing of - dZ/dt max obtained by bioimpedance measurement in nine locations on the thorax and neck. This knowledge is essential for determination of which bioimpedance channel could be used as a proximal for evaluation of pulse wave velocity (PWV) from the beginning of the ascending aorta or another aortic section. Time intervals between the Doppler flow and bioimpedance information (- dZ/dt max) were compared. It was found that the channel located on the left part of the neck is the most suitable as a proximal bioimpedance channel which corresponds to the aortic arch. This match is obtained with regard to the value of the time difference as well as inter-subject stability. This channel can be used as a proximal for evaluation of pulse wave velocity from the aortic arch to the desired distal target place in the body when distance between measured parts is known. The data from 35 volunteers with adequate signal quality were analyzed. Graphical abstract ᅟ.

Intracerebral EEG Artifact Identification Using Convolutional Neural Networks.

Manual and semi-automatic identification of artifacts and unwanted physiological signals in large intracerebral electroencephalographic (iEEG) recordings is time consuming and inaccurate. To date, unsupervised methods to accurately detect iEEG artifacts are not available. This study introduces a novel machine-learning approach for detection of artifacts in iEEG signals in clinically controlled conditions using convolutional neural networks (CNN) and benchmarks the method's performance against expert annotations. The method was trained and tested on data obtained from St Anne's University Hospital (Brno, Czech Republic) and validated on data from Mayo Clinic (Rochester, Minnesota, U.S.A). We show that the proposed technique can be used as a generalized model for iEEG artifact detection. Moreover, a transfer learning process might be used for retraining of the generalized version to form a data-specific model. The generalized model can be efficiently retrained for use with different EEG acquisition systems and noise environments. The generalized and specialized model F1 scores on the testing dataset were 0.81 and 0.96, respectively. The CNN model provides faster, more objective, and more reproducible iEEG artifact detection compared to manual approaches.

Parallel use of a convolutional neural network and bagged tree ensemble for the classification of Holter ECG.

The automated detection of arrhythmia in a Holter ECG signal is a challenging task due to its complex clinical content and data quantity. It is also challenging due to the fact that Holter ECG is usually affected by noise. Such noise may be the result of the regular activity of patients using the Holter ECG-partially unplugged electrodes, short-time disconnections due to movement, or disturbances caused by electric devices or infrastructure. Furthermore, regular patient activities such as movement also affect the ECG signals and, in connection with artificial noise, may render the ECG non-readable or may lead to misinterpretation of the ECG. OBJECTIVE: In accordance with the PhysioNet/CinC Challenge 2017, we propose a method for automated classification of 1-lead Holter ECG recordings. APPROACH: The proposed method classifies a tested record into one of four classes-'normal', 'atrial fibrillation', 'other arrhythmia' or 'too noisy to classify'. It uses two machine learning methods in parallel. The first-a bagged tree ensemble (BTE)-processes a set of 43 features based on QRS detection and PQRS morphology. The second-a convolutional neural network connected to a shallow neural network (CNN/NN)-uses ECG filtered by nine different filters (8× envelograms, 1× band-pass). If the output of CNN/NN reaches a specific level of certainty, its output is used. Otherwise, the BTE output is preferred. MAIN RESULTS: The proposed method was trained using a reduced version of the public PhysioNet/CinC Challenge 2017 dataset (8183 records) and remotely tested on the hidden dataset on PhysioNet servers (3658 records). The method achieved F1 test scores of 0.92, 0.82 and 0.74 for normal recordings, atrial fibrillation and recordings containing other arrhythmias, respectively. The overall F1 score measured on the hidden test-set was 0.83. SIGNIFICANCE: This F1 score led to shared rank #2 in the follow-up PhysioNet/CinC Challenge 2017 ranking.

Ventricular Electrical Delay Measured From Body Surface ECGs Is Associated With Cardiac Resynchronization Therapy Response in Left Bundle Branch Block Patients From the MADIT-CRT Trial (Multicenter Automatic Defibrillator Implantation-Cardiac Resynchronization Therapy).

BACKGROUND: Although cardiac resynchronization therapy (CRT) is beneficial in heart failure patients with left bundle branch block, 30% of these patients do not respond to the therapy. Identifying these patients before implantation of the device is one of the current challenges in clinical cardiology. METHODS: We verified the diagnostic contribution and an optimized computerized approach to measuring ventricular electrical activation delay (VED) from body surface 12-lead ECGs. We applied the method to ECGs acquired before implantation (baseline) in the MADIT-CRT trial (Multicenter Automatic Defibrillator Implantation-Cardiac Resynchronization Therapy). VED values were dichotomized using its quartiles, and we tested the association of VED values with the MADIT-CRT primary end point of heart failure or death. Multivariate Cox proportional models were used to estimate the risk of study end points. In addition, the association between VED values and hemodynamic changes after CRT-D implantation was examined using 1-year follow-up echocardiograms. RESULTS: Our results showed that left bundle branch block patients with baseline VED <31.2 ms had a 35% risk of MADIT-CRT end points, whereas patients with VED ≥31.2 ms had a 14% risk (P<0.001). The hazard ratio for predicting primary end points in patients with low VED was 2.34 (95% confidence interval, 1.53-3.57; P<0.001). Higher VED values were also associated with beneficial hemodynamic changes. These strong VED associations were not found in the right bundle branch block and intraventricular conduction delay cohorts of the MADIT-CRT trial. CONCLUSIONS: Left bundle branch block patients with a high baseline VED value benefited most from CRT, whereas left bundle branch block patients with low VED did not show CRT benefits.

Ventricular dyssynchrony assessment using ultra-high frequency ECG technique.

PURPOSE: The aim of this proof-of-concept study is to introduce new high-dynamic ECG technique with potential to detect temporal-spatial distribution of ventricular electrical depolarization and to assess the level of ventricular dyssynchrony. METHODS: 5-kHz 12-lead ECG data was collected. The amplitude envelopes of the QRS were computed in an ultra-high frequency band of 500-1000 Hz and were averaged (UHFQRS). UHFQRS V lead maps were compiled, and numerical descriptor identifying ventricular dyssynchrony (UHFDYS) was detected. RESULTS: An electrical UHFQRS maps describe the ventricular dyssynchrony distribution in resolution of milliseconds and correlate with strain rate results obtained by speckle tracking echocardiography. The effect of biventricular stimulation is demonstrated by the UHFQRS morphology and by the UHFDYS descriptor in selected examples. CONCLUSIONS: UHFQRS offers a new and simple technique for assessing electrical activation patterns in ventricular dyssynchrony with a temporal-spatial resolution that cannot be obtained by processing standard surface ECG. The main clinical potential of UHFQRS lies in the identification of differences in electrical activation among CRT candidates and detection of improvements in electrical synchrony in patients with biventricular pacing.