ECG Analysis Research Articles

Temperature and repeated catecholamine surges modulate regional wall motion abnormalities in a rodent takotsubo syndrome model

Transient regional wall motion abnormalities (RWMA) after stress are a hallmark of experimental and clinical Takotsubo Syndrome (TS). The mechanisms driving RWMA remain incompletely understood. This study investigates these mechanisms in a small-animal model of TS, validating the model’s complication profile, recovery dynamics, and recurrence, while assessing the predictive value of echocardiography. Male, 7-9-week-old rats underwent TS induction via isoprenaline infusion. The effects of body temperature and repeated catecholaminergic surges on RWMA were assessed using high-resolution speckle-tracking echocardiography. Complications were evaluated through ECG, autopsy, and blood gas analysis. The TS phenotype showed reduced longitudinal strain with RWMA characterized by apical end-systolic akinesia/dyskinesia, a contracting base, and a narrow transition zone, resulting in apical ballooning. RWMA originated from the apex and recovered in the opposite direction. Hyperthermia increased RWMA incidence, while hypothermia attenuated RWMA (p = 0.008). A repeated catecholamine surge exacerbated RWMA in the acute phase (p = 0.042) but not during recovery (p = 0.308), with reduced RWMA susceptibility after recovery (RR 0.45, 95%CI; 0.27–0.76). Systolic function at initial echocardiography predicted worse outcomes (odds ratio 0.73; 95%CI 0.57–0.85), and RWMA extent correlated with recovery time (Rho = 0.772). TS rats developed complications similar to patients, including heart failure, arrhythmias, and thrombus formation. This study validates the small-animal TS model by replicating patient findings, and emphasizing the importance of controlling body temperature and limiting adrenergic drug use in managing TS. RWMA analysis offers promise as a diagnostic and prognostic tool in TS, and the delayed adaptive response presents a potential pathway for therapeutic intervention.

Efficacy and Safety of Boswellia serrata and Apium graveolens L. Extract Against Knee Osteoarthritis and Cartilage Degeneration: A Randomized, Double-blind, Multicenter, Placebo-Controlled Clinical Trial.

Osteoarthritis is the prevailing form of inflammatory condition in joints of adults and the aging population, leading to long-term disability and chronic pain. Current therapeutic options have variable therapeutic efficacy and/or several side effects. A randomized, placebo-controlled, double-blind clinical trial was conducted in 62 participants using a nutraceutical [standardized Boswellia serrata Roxb. gum resin (300 mg) and Apium graveolens L. seed extract (250 mg)], to determine its safety and efficacy forsupporting cartilage health and reduction in knee osteoarthritis symptoms. All participants were assessed for physical function and pain with the help of WOMAC, VAS, Physicians' Global Assessment for the six-minute walk test/pain. Knee X-ray, KOOS questionnaire score, and FACIT-F score were assessed. Additionally, inflammatory, cartilage degeneration and regeneration biomarkers in serum and urine were evaluated at baseline and after 90 days of treatment. Oral administration of the nutraceutical resulted in prolonged symptomatic relief with reduced pain, stiffness, and swelling. Inflammatory (serum IL-7, IL-1, IL-6, hs-CRP, TNF-α, ESR) and cartilage degeneration biomarkers (serum CTX-II, COMP, MMP-3 and urinary CTX-II) were decreased in the nutraceutical group compared to baseline and placebo. Furthermore, serum N-propeptide of collagen IIA (PIIANP) and procollagen-type-C propeptide (PIICP) levels were increased in the nutraceutical group, suggesting collagen synthesis contributing to cartilage regeneration. At given doses for 90 days, there were no adverse effects based on the clinical examination, biochemical, hematological, and ECG analysis. Taken together, the combination of Boswellia and celery could be a safe and promising herbal nutraceutical option for managing osteoarthritis and cartilage health effectively.

A comparative analysis of CNNs and LSTMs for ECG-based diagnosis of arrythmia and congestive heart failure

Cardiac arrhythmias are major global health concern and their early detection is critical for diagnosis. This study comprehensively evaluates the effectiveness of CNNs and LSTMs for the classification of cardiac arrhythmias, considering three PhysioNet datasets. ECG records are segmented to accommodate around ∼10s of ECG data. Followed by transformation to scalograms using DWT for training VGG-16; and WTS for feature extraction and dimensionality reduction for training LSTM network. VGG-16 achieved 96.44% test accuracy while LSTM achieved 92%. Results also highlight the effectiveness of VGG-16 for short-duration ECG analysis, while LSTM excels in long-term monitoring on edge devices for personalized healthcare.

Assessment of the Cardiovascular Risk of High-Fat-High-Fructose Diet in Hereditary Hypertriacylglycerolemic Rats and Venlafaxine Effect

Metabolic syndrome (MetS) represents a worldwide health problem, affecting cardiovascular and mental health. People with MetS are often suffering from depression. We used hereditary hypertriacylglycerolemic (HTG) rats as an animal model of MetS, and these were fed a high-fat-high-fructose diet (HFFD) to imitate unhealthy eating habits of people having several MetS risk factors and suffering depression. Male HTG rats were fed a standard diet (HTG-SD) or HFFD for eight weeks (HFFD8). Venlafaxine was administered for the last three weeks of the experiment (HFFD8+VE). Heart function was observed on the level of intact organisms (standard ECG in vivo), isolated hearts (perfusion according to Langendorff ex vivo), and molecular level, using the RT-PCR technique. The function of the isolated perfused heart was monitored under baseline and ischemia/reperfusion conditions. Analysis of ECG showed electrical abnormalities in vivo, such as significant QRS complex prolongation and increased heart rate. Ex vivo venlafaxine significantly reduced QT interval after ischemia/reperfusion injury. Baseline values of contractile abilities of the heart tended to be suppressed by HFFD. A significant reduction of LVDP was present in the HFFD8 group. Molecular analysis of specific genes involved in cardiac electrical (Cacna1c, Scn5a), contractile (Myh6, Myh7), metabolic function (Pgc1α) and calcium handling (Serca2a, Ryr2) supported some of the functional findings in vivo and ex vivo. Based on the present effect of venlafaxine on heart function, further research is needed regarding its cardiometabolic safety in the treatment of patients with MetS suffering from depression.

Heart remodelling affects ECG in rat DOCA/salt model.

Myocardial remodelling involves structural and functional changes in the heart, potentially leading to heart failure. The deoxycorticosterone acetate (DOCA)/salt model is a widely used experimental approach to study hypertension-induced cardiac remodelling. It allows to investigate the mechanisms underlying myocardial fibrosis and hypertrophy, which are key contributors to impaired cardiac function. In this study, myocardial remodelling in rat deoxycorticosterone acetate/salt model was examined over a three-week period. The experiment involved 11 male Sprague-Dawley rats, divided into two groups: fibrosis (n=6) and control (n=5). Myocardial remodelling was induced in the fibrosis group through unilateral nephrectomy, deoxyco-rticosterone acetate administration, and increased salt intake. The results revealed significant structural changes, including increased left ventricular wall thickness, myocardial fractional volume, and development of myocardial fibrosis. Despite these changes, left ventricular ejection fraction was preserved and even increased. ECG analysis showed significant prolongation of the PR interval and widening of the QRS complex in the fibrosis group, indicating disrupted atrioventricular and ventricular conduction, likely due to fibrosis and hypertrophy. Correlation analysis suggested a potential relationship between QRS duration and myocardial hypertrophy, although no significant correlations were found among other ECG parameters and structural changes detected by MRI. The study highlights the advantage of the DOCA/salt model in exploring the impact of myocardial remodelling on electrophysiological properties. Notably, this study is among the first to show that early myocardial remodelling in this model is accompanied by distinct electrophysiological changes, suggesting that advanced methods combined with established animal models can open new opportunities for research in this field. Key words Myocardial fibrosis, Remodelling, Animal model, DOCA-salt, Magnetic resonance imaging.

ECG and locomotion analysis

ECG and locomotion analysis

Machine Learning-Based Approach for ECG Analysis and Cardiac Anomaly Detection to Enhance Early Diagnosis and Treatment

Machine Learning-Based Approach for ECG Analysis and Cardiac Anomaly Detection to Enhance Early Diagnosis and Treatment

Interpretable ECG analysis for myocardial infarction detection through counterfactuals

In the evolving landscape of ECG signal analysis, the challenge of limited transparency in machine learning models remains a significant barrier to their effective integration into clinical practice. This study addresses this issue by investigating the use of counterfactual explanations to improve model interpretability for clinicians, particularly in differentiating healthy subjects from myocardial infarction patients. Utilizing the PTB-XL dataset, we developed a methodology for systematic feature extraction and refinement to prepare for counterfactual analysis. This led to the creation of the Visualizing Counterfactual Clues on Electrocardiograms (VCCE) method, designed to improve the practicality of counterfactual explanations in a clinical setting. The validity of our approach was assessed using custom metrics that reflect the diagnostic relevance of counterfactuals, evaluated with the help of two cardiologists. Our findings suggest that this approach could support future efforts in using ECGs to predict patient outcomes for cardiac conditions, achieving interpretation validity scores of 23.29 ±1.04 and 20.28 ±0.99 out of 25 for high and moderate-quality interpretations, respectively. Clinical alignment scores of 0.83 ±0.12 for high-quality and 0.57 ±0.10 for moderate-quality interpretations underscore the potential clinical applicability of our method. The methodology and findings of this study contribute to the ongoing discussion on enhancing the interpretability of machine learning models in cardiology, offering a concept that bridges the gap between advanced data analysis techniques and clinical decision-making. The source code for this study is available at https://github.com/tanyelai/vcce.

Advancing personalised care in atrial fibrillation and stroke: The potential impact of AI from prevention to rehabilitation.

Atrial fibrillation (AF) is a complex condition caused by various underlying pathophysiological disorders and is the most common heart arrhythmia worldwide, affecting 2 % of the European population. This prevalence increases with age, imposing significant financial, economic, and human burdens. In Europe, stroke is the second leading cause of death and the primary cause of disability, with numbers expected to rise due to ageing and improved survival rates. Functional recovery from AF-related stroke is often unsatisfactory, leading to prolonged hospital stays, severe disability, and high mortality. Despite advances in AF and stroke research, the full pathophysiological and management issues between AF and stroke increasingly need innovative approaches such as artificial intelligence (AI) and machine learning (ML). Current risk assessment tools focus on static risk factors, neglecting the dynamic nature of risk influenced by acute illness, ageing, and comorbidities. Incorporating biomarkers and automated ECG analysis could enhance pathophysiological understanding. This paper highlights the need for personalised, integrative approaches in AF and stroke management, emphasising the potential of AI and ML to improve risk prediction, treatment personalisation, and rehabilitation outcomes. Further research is essential to optimise care and reduce the burden of AF and stroke on patients and healthcare systems.

Optimizing machine learning for enhanced automated ECG analysis in cardiovascular healthcare

The rapid advancement of connected health technology, exemplified by wearable devices like the Apple Watch, has revolutionized healthcare by enhancing the diagnosis, monitoring, and treatment of various conditions, particularly heart-related issues. However, these devices generate vast amounts of ECG data that require interpretation, underscoring the need for reliable automated ECG analysis methods. This study explores the use of machine learning and deep learning algorithms, including Support Vector Classifier (SVC), RandomForest, XGBoost, and LinearSVC, for ECG classification, aiming to improve accuracy and diagnostic capabilities. While traditional methods rely on heuristic features and shallow architectures, this research focuses on leveraging deep learning architectures to automatically extract relevant features from ECG signals. The proposed approach demonstrates promising results in accurately categorizing heartbeats, offering a potential solution to the limitations of current classification methods. By optimizing classification models with metaheuristic algorithms, such as JADE, the study achieves significant performance improvements, highlighting the effectiveness of integrating advanced optimization techniques into ECG analysis processes. Ultimately, the findings underscore the potential of machine learning and deep learning algorithms in advancing automated ECG analysis for improved cardiovascular healthcare.

Abstract Su501: Feasibility and Implications of Electrocardiogram-based Prediction of Incessant Refractory Ventricular Fibrillation

Background: Refractory ventricular fibrillation (R-VF) patients, defined as requiring ≥ 3 shocks during resuscitation of OHCA, have relatively poor outcomes. Prior studies have predicted R-VF using the ECG, suggesting potential to guide early patient-specific treatment for R-VF. However, R-VF comprises multiple phenotypes with distinct defibrillation response profiles: Incessant VF (no shocks terminate VF), recurrent VF (all shocks terminate VF but VF recurs), or mixed VF (shocks variably terminate VF). These phenotypes may have distinct prognoses, and may also respond differently to R-VF treatments such as early antiarrhythmics, modified vasopressor dosing, alternative defibrillation strategies, or expedited invasive interventions. Early prediction of specific R-VF phenotype could potentially improve outcome through preemptive targeting of treatment to individual phenotype. Aim: We sought to predict incessant R-VF using machine learning analysis of the defibrillator ECG. Methods: We performed a retrospective study of adult VF OHCA treated by EMS from 2006–2019 in King County WA. We linked a clinical OHCA registry with defibrillator recordings to evaluate ROSC and survival to hospital discharge by VF type (non-refractory; or incessant, recurrent, or mixed R-VF prior to the third shock). To predict incessant R-VF specifically, we used a series of four 2-s ECG segments from each patient within 60s surrounding the first shock. Patients were randomized 65%/35% for training/test. A support vector machine classifier was trained to predict subsequent incessant R-VF using ECG scalograms. Algorithm performance was assessed by area under the receiver operating characteristic curve (AUC) and by sensitivity at >75% specificity. Results: Of 1111 included VF patients, median age was 62, 241 (22%) were female, 658 (59%) were refractory (≥ 3 shocks), 800 (72%) achieved ROSC, and 520 (47%) survived. The incessant R-VF phenotype accounted for 10% (N=116) and had worse ROSC and survival compared to recurrent R-VF and non-refractory patients (Table). Training and test AUCs to predict incessant R-VF were 0.77 (95% CI: 0.72–0.82, N=723) and 0.76 (95% CI: 0.66–0.82, N=388) respectively. Test sensitivity and specificity were 64% and 76%. Conclusion: Refractory VF comprises distinct phenotypes with disparate outcomes. Incessant R-VF has notably poor prognosis. Machine learning may be able to predict incessant R-VF, potentially guiding preemptive, patient-specific treatments to improve outcome.

Abstract 4136938: Autonomous Personalized Dynamic ECG Analysis Predicts 1-hour Incidence of Sudden Cardiac Arrest

Background: Sudden cardiac arrest (SCA) is a leading cause of death in the USA. Defibrillators prevent SCA by delivering shocks for sustained ventricular tachyarrhythmias (VT). However, current clinical methods for predicting SCA are limited. We have shown that distinct nonlinear fluctuation patterns “hidden” within normal-appearing ECG waveforms have unique prognostic value. Hypothesis: Autonomous personalized dynamic ECG analysis predicts the 1-hour incidence of SCA. Methods: In an established pressure-overload model that exhibits key features of human VT and SCA (e.g., cardiac restitution properties), 30% of 37 freely ambulating animals had spontaneous SCA in 4 weeks. The exposures included a rolling 1-hour window of time, frequency and nonlinear domain measures of heart rate and QT time series during sinus rhythm, and EntropyX QT , a higher-dimensional machine learning-based nonlinear measure of cardiac repolarization dynamics. Results: In multivariate analyses, the major independent predictors of 1-hour incidence of SCA were (1) high EntropyX QT ; (2) high parasympathetic tone, as indexed by the high-frequency component of the heart rate power spectral density; and (3) the degree of coupling between RR and QT intervals. The baseline values of EntropyX QT and parasympathetic tone increased by 3.6-fold and 46.5-fold, respectively, during the hour preceding SCA. The adjusted ROC curve area was > 89%. Conclusion: Continuous autonomous analysis of cardiac repolarization and autonomic dynamics strongly and independently predicts the subacute incidence of SCA. Incorporation of these measures into current strategies of SCA risk stratification has potential to save many lives and warrants further evaluation.

Cardiotoxic and Cardioprotective Effects of Methylene Blue in the Animal Model of Cardiac Ischemia and Reperfusion.

Background/Objectives: Treatment of patients with myocardial ischemic diseases crucially involves cardiac reperfusion (CR). However, oxidative stress and tissue lesions caused by CR may also lead to lethal complications, such as arrythmias and vasoplegic syndrome (VS). Although methylene blue (MB) has long been used to treat VS due to cardiac ischemia and reperfusion (CIR) and/or surgery because of its vascular effects, MB's effects on the heart are unclear. Therefore, we investigated the potential cardioprotective or arrhythmogenic effects of MB in an animal model of CIR. To this end, 12-16-week-old male Wistar rats were divided into four experimental groups: (a) rats subjected to SHAM surgery with no ischemia; (b) rats subjected to CIR and treated with a vehicle (SS + CIR); and (c) rats subjected to CIR and treated with 2 mg/kg i.v. MB before ischemia (MB + ISQ) or (d) after ischemia but before reperfusion (ISQ + MB). An ECG analysis was used to evaluate the incidence of ventricular arrhythmias (VAs), atrioventricular blocks (AVBs), and lethality (LET) resulting from CIR. After CIR, rat hearts were removed for histopathological analysis and lipid hydroperoxide (LH) measurements. Results: The incidence of VA, AVB, and LET was significantly increased in the MB + ISQ group (VA = 100%; AVB = 100%; LET = 100%) but significantly reduced in the ISQ + MB group (VA = 42.8%; AVB = 28.5%; LET = 21.4%) compared with the SS + CIR group (VA = 85.7%; AVB = 71.4%; LET = 64.2%). LH concentration was significantly reduced in both MB-treated groups, but myocardial injuries were increased only in the MB + ISQ group when compared with the SS + CIR group. Conclusions: These results indicate that MB produces a biphasic effect on CIR, with cardiotoxic effects when administered before cardiac ischemia and cardioprotective effects when administered after ischemia but before cardiac reperfusion.

Assessment of age-related features of the condition of heavy-duty truck drivers based on spectral analysis of EEG-ECG and stress tests during mass occupational examinations

The authors studied the physiological characteristics of drivers of heavy-duty coal mine vehicles based on spectral analysis of EEG-ECG and stress tests, which have prognostic value in professional activity. The study aims to assess the functional condition of heavy-duty truck drivers of different age groups based on a comprehensive spectral analysis of EEG-ECG under occupational examinations. During the survey, the authors have studied the spectral parameters of the EEG‒ECG in 300 male drivers aged 25 to 66 years. Specialists have performed EEG registration when placing electrodes in the 10–20 system. The EEG analysis was performed with a preliminary visual analysis followed by the removal of artifacts. The scientists assessed the level of representation of the main alpha rhythm and its spatial distribution. The ECG registration was in the first lead. The duration of EEG-ECG registration is about 15 minutes in a typical sequence of stress tests: baseline, rhythmic photostimulation, hyperventilation with recovery periods. The spectral analysis of the EEG‒ECG was performed automatically on a Neurotravel device (Italy). The authors found lower alpha and delta activity in the initial EEG spectra in older drivers aged 45–66 years, which contributed to the cumulative effect of stress load. Light stress stimulation significantly reduced alpha and delta activity and increased beta EEG activity in both groups, while in the older group the high-frequency range significantly exceeded the low-frequency range. Simultaneously with the changes in the EEG, stress photostimulation increased the energy of the ECG spectra in all frequency ranges. The authors established the appearance of additional stress indicators during light stimulation in older drivers based on the correlation of EEG‒ECG spectra (alpha/delta) and the absence of such changes in the group of young drivers aged 25–35 years. Moreover, the EEG–ECG correlation coefficients and their critical values increased in groups in increments of 6 years — from 35 to 65 years. Elderly drivers, based on neuro-cardiodynamic indicators, were in a state of chronic stress and the resulting risks to health and professional activity, and the identified physiological mechanisms of stress may be at the heart of ways to prevent it. Limitations. The study is limited to the assessment of EEG and ECG indicators in 300 heavy truck drivers. Ethics. The study was performed using non-invasive methods and approved by the Bioethical Committee of the Research Institute of Complex Problems of Hygiene and Occupational Diseases, Protocol No. 5, § 1 dated 12/26/2018, in accordance with the Helsinki Declaration of the World Medical Association "Ethical Principles of conducting medical research with human participation as a subject" as amended in 2013 and the "Rules of Good Clinical Practice" approved by by Order of the Ministry of Health of the Russian Federation No. 200n dated April 1, 2016. All the examined patients signed a voluntary informed consent to participate in the study.

Age-related features of the main risk factors and clinical and laboratory parameters in men aged 30–50 years with atherothrombotic stroke

Introduction. Ischemic stroke in patients under 50 years of age is not uncommon. The prevalence of stroke in young people increases due to the increase in risk factors. Early detection of risk factors and optimization of primary and secondary preventive therapy are mandatory to reduce the burden of stroke in young people and preserve potential years of life.Objective. To identify and evaluate the main risk factors and clinical and laboratory parameters in men aged 31–40 and 41–50 years with atherothrombotic stroke.Material and methods. Before the development of the COVID-19 pandemic, 60 patients were examined (mean age 43.3±3.8 years). All patients underwent neuroimaging of the brain, ultrasound examination of the main vessels of the head, echocardiography, ECG, and laboratory analysis of blood serum.Results. Among all patients, there were 10 patients aged 31–40 years (mean age 37.4±2.6 years) and 50 patients aged 41–50 years (mean age 44.5±2.8 years). All patients had atherosclerotic lesions of the main arteries of the head, with stenosis of more than 70 % only in patients aged 41–50 years. Among patients aged 31–40 years and 41–50 years, the most common risk factors were arterial hypertension (90 and 72 %), smoking (70 and 72 %), dyslipidemia (66.7 and 60 %), and regular alcohol consumption (30 and 36 %). The study showed statistically significant correlations of the main vascular risk factors with serum lipid metabolism indices, hemostasis and endothelial dysfunction markers.Conclusions. The study showed the frequency of occurrence of the main risk factors in patients with atherothrombotic stroke aged 31–40 and 41–50 years, as well as the role of lipid metabolism disorders, hemostasis disorders and endothelial dysfunction in their development.

Deep learning hybrid model ECG classification using AlexNet and parallel dual branch fusion network model.

Cardiovascular diseases are a cause of death making it crucial to accurately diagnose them. Electrocardiography plays a role in detecting heart issues such as heart attacks, bundle branch blocks and irregular heart rhythms. Manual analysis of ECGs is prone to mistakes and time consuming, underscoring the importance of automated methods. This study uses AI models like AlexNet and a dual branch model for categorizing ECG signals from the PTB Diagnostic ECG Database. AlexNet achieved a validation accuracy of 98.64% and a test set accuracy of 99% while the dual branch fusion network model achieved a test set accuracy of 99%. Data preprocessing involved standardizing, balancing and reshaping ECG signals. These models exhibited precision, sensitivity and specificity. In comparison to state of the arts' models such as Hybrid AlexNet SVM and DCNN LSTM our proposed models displayed performance. The high accuracy rates of 99% underscore their potential for ECG classification. These results validate the advantages of incorporating learning models into setups for automated ECG analysis providing adaptable solutions for various healthcare settings including rural areas.

Nursing analysis of holter ECG - a literature gap

Nursing analysis of holter ECG - a literature gap

Chest pain and AI interpretation: The essential role of clinical judgment and quality data

Chest pain is a common and complex condition in emergency departments, often complicating the diagnostic process due to its wide range of potential causes. This case report presents a 72-year-old man with chest pain and ST-segment elevation on ECG, initially suggestive of acute coronary syndrom. AI-assisted ECG analysis reinforced this suspicion; however, angiography revealed patent coronary arteries. When his symptoms persisted, further investigation uncovered a pneumothorax. Additional evaluation identified a digitization error affecting the AI interpretation. This case underscores the importance of a comprehensive clinical approach for chest pain and highlights the need for quality data input and alternative imaging, such as thoracic ultrasound, for accurate and timely differential diagnoses.

Larger QRS but better hemodynamic function after high output His bundle pacing in patients with heart failure and narrow QRS. A paradox or a new concept?

Abstract Background QRS duration is an important factor in determining Cardiac Resynchronisation Therapy (CRT) response and a useful surrogate for electromechanical dyssynchrony. However, we don’t have studies confirming this concept in conduction system pacing. Purpose We performed His Bundle Pacing (HBP) with high output (3.5V/1msec) in HF patients with dilated or ischemic cardiomyopathy and narrow QRS to evaluate hemodynamic benefit independent of the QRS duration or AV delay shortening. Methods 14 patients with dilated (7 patients) or ischemic (7 patients) cardiomyopathy (EF<35%) and narrow QRS (<120ms) were referred for implantation of a defibrillator. We implanted an ICD- HBP system, and we collected clinical, echocardiographic, and electrocardiographic data. Results We obtained HBP in all 14 patients successfully. PQ after implantation was shorter from 172±48 to 143±16 (P=0.048) so as to pace the ventricle. Baseline QRS duration increased after pacing from 110±15 to 123±16ms (P=0.008) as measured by ECG analyzer. Despite this increase in QRS duration basal mean EF increased from 30±4% to 46±10% (P<0.001) after a median follow-up of 17 months. Left ventricular end-diastolic diameter and left ventricular end-diastolic volume decreased from 62±5 to 57±5mm (P=0.013) and from 195±69 to 126±27ml (P<0.001), respectively. Left ventricular end-systolic volume decreased from 139±58 to 69±27 (P<0.001). TAPSE increased from 19±4 to 27±4. NYHA class decreased by one class in every patient after only one month. HBP threshold was 1.05V±0.7/1msec and raised to 1.33±0.93/1msec (due mainly to 2 patients with very high threshold). The impedance was 588±109 and decreased to 484±69 Ohms. R-wave at implant was 3.2±1.8mV. In 2 out of 14 patients, threshold increased but was still less than 3.5V/1msec. Conclusions HBP in HF patients with dilated or ischemic cardiomyopathy and narrow QRS improves hemodynamic function in right and left ventricle and decreases NYHA class. This benefit appears to be produced by a higher stimulation output despite a larger QRS after pacing. We exclude that a very short AV delay could be favorable, by measuring acute EF changes only after increasing the pacing output without AV delay modifications. To our knowledge these are the first cases of beneficial HBP in HF patients with narrow QRS. Interestingly, larger QRS corresponds to higher EF in patients with high output HIs Bundle Pacing.

Validation of an artificial intelligence system for holter ECG analysis

Abstract Introduction Holter ECG is a commonly used tool to identify heart rhythm disorders. As physician-based analysis of such Holter ECG is time consuming, various approaches using computerized interpretation have been developed. However, accuracy of these tools remains inferior to physician-based analysis. Thus, this study assesses the diagnostic performance of an artificial intelligence (AI)-powered Holter ECG service in comparison to results from reference databases as well as with physician-based evaluation. Methods 231 ECG recordings, covering 2,261 hours of ECG with annotations for 10,024,315 heartbeats from the following reference databases were used: MIT-BIH Arrhythmia (MITDB), American Heart Association ECG (AHDB), MIT-BIH Atrial Fibrillation (AFDB), Long Term Atrial Fibrillation (LTAFDB). Those recordings were originally collected from a mixed population of inpatients and outpatients between 1975 and 2008. Analysis of these recordings regarding the detection of QRS complexes, QRS classification and detection of Atrial Fibrillation (AF) was performed using a novel service for AI-based ECG analysis. Results were compared to the diagnosis stated within the database. QRS complex detection accuracy was performed on recordings from MITDB and AHADB by checking if time positions are within 150 ms to reference annotations. The QRS classification in three classes (normal, supraventricular, ventricular) was then compared for each detected heartbeat. Mismatches were adjudicated manually by three experts. Episodes of Atrial Fibrillation were detected in recordings from MITDB, AFDB and LTAFDB and compared to the reference episodes from the databases. The accuracy was evaluated based on the overlap with the reference annotations in the time-domain. Results Sensitivity and Precision of heartbeat detection compared to the reference databases exceeded 99% Figure 1 displays the results of heartbeat classification. Normal heartbeats were classified with a sensitivity of 99.6%, precision of 99.9% and specificity of 99.6%. For ventricular heartbeats, a sensitivity of 99.58 %, precision of 99.88 and specificity of 99.99% were achieved. Premature supraventricular contractions were recognized with a sensitivity of 99.58 %, precision of 100% and specificity of 100%. When comparing the 248,187 QRS complex classifications of MITDB and AHADB to the AI-powered analysis, 128 discrepancies were revealed. 43 were due to a discrepancy of convention as these beats were in sections that were also marked as AF episodes by the databases and the AI-powered analysis marks all beats in AF episodes as normal. Of the 85 remaining discrepancies, more than 87% were adjudicated in favor of the AI-powered analysis by the experts. Figure 2 shows the comparison of AF detection results showing Sensitivity and Precision above 99%. Conclusion AI-powered Holter ECG analysis compared to reference ECG stripes exceed 99% of sensitivity, specificity and precision.QRS complex detectionAF/ AFib detection