Detection Of Fibrillation Research Articles

Detection of Atrial Fibrillation from ECG Signal Using Efficient Feature Selection and Classification

Detection of Atrial Fibrillation from ECG Signal Using Efficient Feature Selection and Classification

Mitigating Data Variability and Overfitting in Deep Learning Models for Atrial Fibrillation Detection Using Single-Lead ECGs

Mitigating Data Variability and Overfitting in Deep Learning Models for Atrial Fibrillation Detection Using Single-Lead ECGs

Diagnostic accuracy of smartwatches for automated atrial fibrillation detection over time: insights from the basel wearable study

Abstract Background The first smart-watch (SW) able to record a single lead ECG (SL-ECG) with automated atrial fibrillation (AF) detection algorithm was introduced 2018. Multiple other manufacturers followed. While manufacturers claim to continuously improve their devices’ algorithm, little is known about the impact on diagnostic accuracy on AF detection over time in a real-world cohort of patients. Purpose To assess if the diagnostic accuracy of algorithm-based AF detection of 5 smart devices has improved over time. Methods In this prospective study, consecutive patients undergoing electrophysiological procedures in a tertiary referral center were included between April 2021 to May 2023. Each participant received a 12-lead ECG and 5 SL-ECGs with different smart devices (AliveCor KardiaMobile, Apple Watch 6, Fitbit Sense, Samsung Galaxy Watch 3 and Withings ScanWatch). All SW were continuously updated for newest software or firmware as recommend by the manufacturers. The SW rhythm classification was compared to manual cardiologist-based interpretation of SL-ECGs. The change in number of inconclusive SL-ECGs by manufacturer’s algorithm were investigated over time via Cochran-Armitage test for trend and illustrated over time via logistic regression. Results 247 participants were included (30% female, mean age 66.3 [IQR: 57-74] years) resulting in a total of 1’235 SL-ECGs. Among these, 202 SL-ECGs (16%) were labeled as inconclusive by at least one smart-device. Inconclusive rates for individual SW ranged from 15% to 19%. When comparing the first and last quintile of patients in the study period, the inconclusive rates for the KardiaMobile were 14% vs. 6%, for the Apple Watch 14% vs. 6%, for the Fitbit Sense 20% vs. 14%, for the Samsung Galaxy 18% vs. 10% and for the Withings ScanWatch 22% vs. 14%, respectively. When assessed continuously over time, we observed a trend towards fewer inconclusive tracings, without reaching statistical significance (Figure 1). Conclusion Over 26 months, inconclusive tracings of 5 SW didn’t significantly decrease despite frequent software updates in a real-world patient cohort. Larger, long-term cohorts are needed to assess if automated AF detection has significantly improved over time.Figure 1

Effectiveness of a stepwise approach for screening of atrial fibrillation after stroke: insights from the SAFAS study

Abstract Background Detection of atrial fibrillation (AF) is critical after ischemic stroke, providing information regarding the mechanism of the event and leading to modification in the antithrombotic strategy. While most guidelines recommend screening patients for AF with 12-lead ECG, telemetry, long-duration Holter monitoring and implantable cardiac monitor (ICM), the optimal timing and combination of such screening tools remain unclear. Objective This study aimed at investigating the suitability of a sequential combination of screening techniques (12-lead ECG, telemetry, in hospital long-lasting Holter monitoring, and ICM in the detection of AF after stroke. Methods Patients without previously known AF admitted to the Dijon University Hospital stroke unit for acute ischemic stroke were prospectively enrolled. After a stepwise screening approach for AF based on admission ECG, telemetry monitoring during the stroke unit stay and long-duration Holter monitoring during hospital stay, cryptogenic stroke patients were implanted of an ICM. Primary endpoint was the presence of AF detected during the 3-year period after stroke based on this sequential screening approach. Results A total of 240 patients were included. Among them, 104 (43.3%) patients had a documented cause of stroke non-related to AF. Among the remaining 136 patients (53.7% male, 70.8±13.7 yo), AF was detected in 82 (60%) patients over the acute screening phase or the 3-year follow-up with ICM. AF was diagnosed using 12-lead ECG, in-hospital telemetry, and in hospital long-lasting Holter monitoring in 17 (13%), 25 (18%), and 18 (13%) patients, respectively. AF was detected after the first 24 hours on the long-lasting Holter monitoring in 66% of patients. Among the 76 (56%) patients classified as cryptogenic after the complete stroke work-up and implanted from an ICM, AF was detected in 22 (29%) patients. AF occurred during the first, second, and third years of implantable monitoring in 14 (18.4%), 5 (6.6%), and 3 (3.9%) patients, respectively (Figure 1). Mean time from ICM implantation to AF diagnosis was 308+/-279 days. Finally, among all AF detected, 72% (60/83) were found during the initial intensive in-hospital screening. Conclusion A stepwise approach for AF screening after ischemic stroke allows the early detection of AF in a substantial number of patients during hospital stay. Even with such proactive initial monitoring strategy, invasive monitoring remains complementary to non-invasive screening tools not to overlook more distant AF episodes. Studies focusing on the relative risk of ischemic stroke recurrence according to AF timing and burden are needed.Figure 1

Performance of a vectorcardiographic deep learning algorithm compared to single-lead and 12-lead ECG for atrial flutter detection: implications for wearable devices

Abstract Background Wearable devices have developed the capacity to offer single-lead ECG analysis and are already widespread in use, particularly for detection of atrial fibrillation (1,2). Vectorcardiography entails the transformation of electrical vectors into X, Y, and Z axes to constitute a vectorcardiogram (VCG) (3). The VCG has diagnostic promise given the capacity to generate it without requiring clinical presentation for standard 12-lead ECG acquisition. Furthermore, a novel credit-card sized device capable of acquiring a VCG has recently been developed – raising the potential for arrhythmia detection and discrimination beyond wearable single-lead ECG systems. Objective To assess the performance of a deep learning algorithm trained on single lead and 12-lead ECGs and VCGs for detecting atrial flutter. Methods A convolutional neural network (AFL-CNN) was developed on a dataset of 4,000 ECGs and VCGs from a single US academic medical institution (80% training and 20% for testing). Validation was performed on a separate publicly available dataset (2223 ECGs). Model performance was assessed on single lead ECG analysis (lead I, obtained by removing the other leads) and standard 12-lead ECG. The gold standard for AFL diagnosis was physician-confirmed 12-lead ECG diagnoses in the dataset. The VCG was generated from the 12-lead ECG via Kors transformation matrix to which AFL-CNN was also applied. Model performance was assessed with sensitivity, specificity, positive and negative predictive values, accuracy, and F1 statistic. Results On single lead ECG analysis, AFL-CNN achieved 71.2% sensitivity and 96.9% specificity with PPV 84.0% and NPV 93.7% and 92.2% accuracy with F1 83.2%. When AFL-CNN was applied to the VCG, performance improved: 91.0% sensitivity and 98.7% specificity, PPV 94.24%, NPV 98.0%, 97.3% accuracy, and F1 95.31%. Lastly, the performance remained high when AFL-CNN was applied to the 12-lead ECG: 95.4% sensitivity, 98.63% specificity, PPV 94.0%, NPV 99.0%, and accuracy of 98.0%, with F1 97.2%. Conclusion A deep learning algorithm applied to vectorcardiography for the detection of atrial flutter showed improved performance over single lead analysis and comparable performance to standard 12-lead applications. The finding of improved model performance beyond single lead ECG tracings underscores both the limitations of current wearable technologies in arrhythmia detection beyond atrial fibrillation and the potential for VCG-based algorithm prediction to fill gaps in healthcare inequality when standard 12-lead acquisition is challenging to obtain.

Continuing evaluation of atrial fibrillation detection after cryptogenic stroke: one-year findings from a multi-center study with confirm Rx ICM

Abstract Background The detection of atrial fibrillation (AF) after a cryptogenic stroke (CS) carries crucial therapeutic implications. While clinical trials have meticulously explored the incidence of AF in CS patients, real-world data on this topic remain relatively sparse. In this study, we aimed to assess the incidence of AF among patients with cryptogenic stroke, who underwent the implantation of an insertable cardiac monitor (ICM). The abstract reports the interim results from the CS sub-study of the Smart Registry. Methods A prospective, single-arm, multi-center registry was conducted to identify AF in patients with cryptogenic stroke using the Confirm R ICM. Inclusion criteria comprised participants aged 40 years or older who had experienced cryptogenic strokes within a 90-day window. Prior to enrollment, comprehensive diagnostic assessments were performed to meticulously rule out all potential sources of embolism, in accordance with the minimal standards of the American and European Guidelines. At each follow-up visit, expert electrophysiologists reviewed and adjudicated ICM detected AF episodes. The primary endpoint was the cumulative incidence of true device-detected AF (lasting more than 30 seconds) at 6 months. Kaplan-Meier curves were employed to assess the cumulative incidence of true device-detected AF through 12 months. Results In our study, 117 patients were enrolled across 19 global sites from September 2021 to August 31, 2023. Kaplan-Meier analysis revealed a 21.3% probability (confidence interval (CI): 14.5%, 30.7%) of true AF episodes lasting more than 30 seconds at 6 months, increasing to 31.0% (CI: 22.0%, 42.5%) at 12 months. Similar results were observed in episodes lasting over 120 seconds at both 6 months (20.7%) and 12 months (30.5%). In average, 14.4 true AF episodes per patient were detected per year. The median time from implantation to AF detection (> 30 seconds) was 31.5 days (interquartile range, 7 to 191). Conclusion Our ongoing study illustrates the effectiveness of the Confirm Rx ICM in detecting atrial fibrillation after cryptogenic stroke. These results confirm the diagnostics value of long-term continuous AF monitoring in cryptogenic stroke patients.

Impact of photoplethysmography-based smartphone rhythm monitoring on atrial fibrillation detection and management after cardiac surgery: a pragmatic randomized trial

Abstract Introduction Postoperative atrial fibrillation (POAF) occurring after cardiac surgery is common and associated with adverse outcomes. Systematic monitoring of POAF beyond discharge is cumbersome. The emergence of photoplethysmography (PPG)-based rhythm monitoring with digital consumer devices could potentially mitigate these hurdles. Selected smartphone applications leveraging this technology have demonstrated excellent usability and accuracy for the detection of atrial fibrillation (AF). However, the impact of PPG-detected POAF on real-life clinical practice remains uncertain. Purpose To determine whether intermittent PPG-based smartphone rhythm monitoring, after being discharged home following cardiac surgery, impacts AF management. Methods The SURGICAL-AF 2 study is a pragmatic, investigator-initiated, open-label, multicenter, randomized clinical trial, conducted in three Belgian centers. The intervention group performed one-minute rhythm checks three times daily with a smartphone-based PPG application after hospitalization for cardiac surgery until the first follow-up visit with a cardiologist, scheduled at 21 – 91 days. Rhythm monitoring was not mandated in the usual care group. The primary endpoint was a composite of initiation of oral anticoagulation (OAC), cardioversion, up-titration or initiation of antiarrhythmic drugs (Vaughan-Williams class I or III) or implantation of a cardiac implantable electronic device (CIED). Secondary endpoints were incidence of POAF and actionable POAF, defined as a detection in patients with CHA2DS2-VASc score ≥2 for women or ≥1 for men who are not treated with OAC. Results Of the 450 patients randomized (238 patients in the intervention group and 212 patients in the usual care group; mean [SD] age, 64.1 [9.2] years; 96 women [21.3%]; 130 patients with AF before inclusion [28.9%]; 103 patients on OAC [22.9%], median [interquartile range] CHA2DS2-VASc score, 2 [1-3]), 98.7% completed the trial. In the intent-to-treat analysis, the primary end point occurred in 24 patients (10.3%) in the intervention group versus 5 patients (2.4%) in the usual care group (odds ratio (OR) 4.7, 95% CI, 1.8 - 12.6; P =.002). POAF was detected in 44 patients (18.8%) in the intervention group and was actionable in 25 patients (10.7%) versus 4 patients (1.9%) in the usual care group and actionable in 2 patients (0.9%). (POAF detection, OR 12.0, 95% CI, 4.2 - 34.5; P < .001; actionable POAF, OR 12.5, 95% CI, 2.9 - 52.6; P < .001) Conclusions In unselected patients discharged home following cardiac surgery, PPG-based smartphone monitoring revealed significantly more POAF which led to changes in AF management (OAC initiation, rhythm control therapy or CIED insertion). Longer follow-up is needed to determine whether these changes will lead to improved outcomes.

A novel method to identify critical isthmus of right atrial reentrant tachycardia; Functional substrate mapping characteristics during sinus rhythm

Abstract Background Right atrial tachycardia (AT) is a frequent rhythm disorder in patients with atrial scar mainly due to surgical incisions or congenital heart diseases. Despite the mounting evidence about AT mechanisms and types, data is scarce regarding the conduction properties as well as functional characteristics of atrial substrate during sinus rhythm that has role in the maintenance of the tachycardia. Purpose We sought to evaluate the relationship between the functional substrate mapping (FSM) characteristics of right atrium and the critical isthmus (CI) of re-entrant ATs in patients with underlying atrial scar. Methods Among patients with the history of right AT who underwent catheter ablation with 3-D mapping were retrospectively enrolled. Voltage map and isochronal late activation mapping (ILAM) were created during sinus / paced rhythm using multielectrode catheters to detect deceleration zones (DZ). Subsequently AT was induced with programmed stimulation and activation mapping was performed to detect CI of the tachycardia. Atrial tachyarrhythmia (ATa) recurrence was defined as detection of atrial fibrillation or AT (≥ 30 s) during the follow-up. Results A total of 24 patients [mean age: 46 ±15, gender: 13 (54 %) female] with right AT were included. A total of 37 ATs were mapped [16 (43.2%); localized reentry, 21 (56.7%), macroreentry]. Atrial low voltage zones comprised 23.3 ± 13.0 % of total right atrium. The mean value of bipolar voltage, EGM duration and conduction velocity during sinus rhythm corresponding to CI of ATs were 0.18 ± 0.10 mV, 121.7 ± 29.4 ms, 0.06 ± 0.04 m/s and, respectively. Total number of DZs per chamber was 1.1 ± 0.3, which were all located in the low-voltage zone (<0.5 mV) detected by high density mapping. All CIs of non- cavo-tricuspid-isthmus (CTI) dependent reentry were colocalized with DZs detected during FSM. The positive predictive value of DZs to detect CI of inducible ATs is 80.8%. During the mean follow-up of 11.7 ± 8.1 months, freedom from ATa was 87.5%. Conclusion Although, CTI-dependent macroreentry is the most common mechanism in patients with RA scar, our findings demonstrated the relevance of FSM to predict non-CTI dependent ATs. Conduction slowing manifested as DZs with continuous-fragmented signal morphology may guide to tailor ablation strategy in case of underlying RA scar.

Cost-effectiveness of atrial fibrillation detection with insertable cardiac monitors in large-artery or small-vessel disease stroke in the United Kingdom

Abstract Background The randomised STROKE-AF trial demonstrated clinical effectiveness of continuous cardiac monitoring to detect atrial fibrillation (AF) with insertable cardiac monitors (ICMs) for patients with ischemic stroke attributed to large-artery or small-vessel disease (LAD/SVD). ICMs were ten-fold more effective at detecting AF compared to the standard of care (SoC). However, cost-effectiveness of this monitoring strategy is unknown. Purpose The aim of this economic evaluation was to assess the cost-effectiveness of long-term continuous cardiac monitoring via ICMs, for detection of AF (and, in event of AF diagnosis, subsequent anticoagulation), in patients with ischemic stroke attributed to large-or-small vessel disease compared to the SoC of conventional follow up in the United Kingdom. Methods We constructed a Markov model with a cycle length of three months and a lifetime horizon to conduct the cost-effectiveness analysis, from the National Health Service (NHS) perspective. Costs and health-related benefits were discounted at an annual rate of 3.5%. We expressed outcomes in terms of quality-adjusted life-years (QALYs). Patient characteristics in the analysis and detection probabilities were based on the STROKE-AF trial. Detection of AF resulted in a switch in treatment from antiplatelet to direct oral anticoagulant (DOAC). Risk of cerebrovascular events were dependent on a patient’s AF detection status and subsequent treatment. Scenario, deterministic, and probabilistic sensitivity analyses were conducted to test the robustness of the base-case incremental cost-effectiveness ratio (ICER) and account for uncertainties in the model. We stratified patients into risk groups based on CHA2DS2-VASc scores and risk of AF detection. Results The base case incremental cost-effectiveness ratio of the implantation of an ICM was £15,485 per QALY, substantially under the NICE willingness-to-pay threshold of £20,000-£30,000 per QALY. ICM implantations resulted in 0.139 QALYs gained per patient at an incremental cost of £2,147. There were 0.054 fewer stroke events per patient in the ICM arm versus SoC. Results were robust under sensitivity analysis. Conclusion ICMs were cost-effective compared to SoC for detection of AF in patients with LAD/SVD stroke in the United Kingdom driven largely by more stroke events avoided. Subgroup analysis showed that ICMs were more cost-effective in patients with higher CHA2DS2-VASc scores and those with a higher individual risk of AF.Markov Model SchematicSummary ICER Results and Events avoided

Comparative analysis of classical artificial intelligence and deep learning models for atrial fibrillation detection in electrocardiograms

Abstract Background/Introduction Early diagnosis of atrial fibrillation (AF) presents a challenging yet critical task for appropriate interventions aimed at reducing disease-related burden. In this context, strategies employing classical artificial intelligence (CAI) and deep learning (DL) have emerged as promising approaches to optimize cardiac disorder screening and detection. Purpose This study aimed to compare a CAI model and a DL model for the detection of AF in patients undergoing electrocardiographic (ECG) examinations in tertiary healthcare centers. Methods Between December 2022 and November 2023, a total of 135,476 ECGs were performed, comprising 5,067 with AF and 130,409 without AF. The ECGs were analyzed using both artificial intelligence models. The obtained results were then compared to the gold standard (cardiologist's report). In the CAI model, signals were extracted from ECG images, analyzing five key parameters: cardiac rhythm, atrial depolarization, atrioventricular conduction, ventricular depolarization, and ventricular repolarization (figure 1A). These parameters were benchmarked against the standard values from the Brazilian Society of Cardiology guidelines for detecting cardiac anomalies. Conversely, the DL model utilized a one-dimensional ResNet-based Convolutional Neural Network (CNN). This model was trained using ADAM optimization and binary cross-entropy loss, enabling the learning of complex patterns in the data (figure 1B). Results The mean age was 54.6 years (71.9 years with AF and 53.9 without AF). In the AF population, 52.2% were male (46% were male in the overall sample). In the analysis conducted, the CAI model showed a sensitivity and specificity of 90% and 62%, respectively, while the DL model had 90% and 69%, respectively. ROC curves were generated for both models, demonstrating the superior performance of the DL model (figure 2A). Conclusions Although the sensitivity remained similar between the models, the DL model distinguished itself with higher specificity. These results suggest that artificial intelligence, particularly the deep learning approach, holds promise as a supportive resource in AF diagnosis. However, further studies are needed to evaluate the models more thoroughly and determine their clinical applicability in a broader context.Artificial intelligence modelsROC curves

Accuracy of the OMRON HEM-7361T blood pressure monitor in the differentiation between atrial fibrillation and sinus rhythm.

We investigated the accuracy of the OMRON HEM-7361T automated oscillometric blood pressure (BP) monitor in the differentiation between atrial fibrillation and sinus rhythm. An approximately equal number of patients with persistent atrial fibrillation and individuals with sinus rhythm were recruited from outpatients and inpatients of Ruijin Hospital, Shanghai, China. BP was measured three times consecutively with a 30-s interval with the OMRON HEM-7361T automatic electronic BP monitor for atrial fibrillation detection. A hand-held single lead electrocardiogram device was used for simultaneous electrocardiogram recordings. The device accurately identified atrial fibrillation in 100 (99.0%) of the 101 patients, with only 1 patient incorrectly classified as non-atrial fibrillation. The device correctly identified 99 (95.2%) of the 104 participants with sinus rhythm as non-atrial fibrillation, with five participants incorrectly classified as atrial fibrillation. The device had a positive predictive value of 95.2%, negative predictive value of 99.0%, and overall accuracy of 97.1%. Among the six misclassified participants, one with atrial fibrillation had a heart rate of 65 beats/min, and four of the five participants with sinus rhythm had cardiac arrhythmias (atrial or ventricular premature beat in one participants, sinus tachycardia in one participant, and both arrhythmias in one participant). The OMRON HEM-7361T BP monitor is accurate in the differentiation between atrial fibrillation and sinus rhythm. Whether the device is sufficiently accurate in the differentiation between atrial fibrillation and other cardiac arrhythmias remains under investigation.

MANAGEMENT OF REMOTE MONITORING FOR IMPLANTABLE DEVICES CONTROLS: SINGLE CENTER EXPERIENCE

Abstract Background Nowadays the benefits of remote monitoring (RM)of cardiac implantable electronic devices (CIEDs) are Known: an improved both device surveillance and patient clinical management and a decrease in office follow up. So RM is used as supplement in clinic follow up for patients with CIED.However despite decades to RM utilization there are challenges in managing patients with RM including large number of trasmissions and lack of trained staff. There are many different model of care to RM management. Methods We describe our model of care to management CIED recipients with RM considering data collected from January 2021 to November 2023. Results Our model of care consists of 4 electrophysiologists Physicians and 4 Tecnicians. RM enrollment is performed at scheduled in office visit after ten days from hospital discharge after device implantation. implantable cardiac device(ICD)recipients with RM are annualy followed in office and biennal for pacemaker (PM) recipients and they send every three months transmissions. Trained tecnicians gives home monitor, take care of patients education and collected and screen transmissions. Tecnicians inform immediately physicians in case of urgent alert (shock or antitachicardia pacing therapy (ATP), sustained ventricular tachycardia(TV), ventricular fibrillation (Fv) indicator of elective replacement (ERI), lead releted alert or noise). Physician contact patients the same day (exept Saturday and Sunday). In addiction Physicians twice a week review less urgent and scheduled transmissions and call the patients by phone for visit in office for patients with low rate of biventricular pacing, detection of atrial fibrillation. We have 1000 patients with RM: 746 ICD recipients (103 Biotronik, 150 Abbott, 200 Boston scientific, 293 medtronic), 50 implantable cardiac monitor (ICM) recipients and 204 PM recipients. We analyze 12721 transmissions.Considering ICD recipient we detect 96 episode of ATP and 31 of shock. Shock of ICD was registered in 52 patients (6,9%)We received 1982 atrial fibrillation (AF) detection, with 35 for new detectable AF.The most of transmisisone were for low bIV pacing. We contacted these patients by phone about clinical status and in some case we called them in office. Conclusions Our model of care allows to control a large number of trasmissions with fast management of urgent alert trasmissions.

Machine Learning Modeling to Predict Atrial Fibrillation Detection in Embolic Stroke of Undetermined Source Patients.

In patients with embolic stroke of undetermined source (ESUS), occult atrial fibrillation (AF) has been implicated as a key source of cardioembolism. However, only a minority acquire implantable cardiac loop recorders (ILRs) to detect occult paroxysmal AF, partly due to financial cost and procedural inconvenience. Without the initiation of appropriate anticoagulation, these patients are at risk of increased ischemic stroke recurrence. Hence, cost-effective and accurate methods of predicting AF in ESUS patients are highly sought after. We aimed to incorporate clinical and echocardiography data into machine learning (ML) algorithms for AF prediction on ILRs in ESUS. This was a single-center cohort study that included 157 consecutive patients diagnosed with ESUS from October 2014 to October 2017 who had ILR evaluation. We developed four ML models, with hyperparameters tuned, to predict AF detection on an ILR. The median age of the cohort was 67 (IQR 59-74) years old and the median monitoring duration was 1051 (IQR 478-1287) days. Of the 157 patients, 32 (20.4%) had occult AF detected on the ILR. Support vector machine predicted for AF with a 95% confidence interval area under the receiver operating characteristic curve (AUC) of 0.736-0.737, multilayer perceptron with an AUC of 0.697-0.708, XGBoost with an AUC of 0.697-0.697, and random forest with an AUC of 0.663-0.674. ML feature importance found that age, HDL-C, and admitting heart rate were important non-echocardiography variables, while peak mitral A-wave velocity and left atrial volume were important echocardiography parameters aiding this prediction. Machine learning modeling incorporating clinical and echocardiographic variables predicted AF in ESUS patients with moderate accuracy.

Cardiac arrhythmias in patients with type 2 diabetes mellitus

Introduction. The contribution of type 2 diabetes mellitus (type 2 diabetes) and cardiovascular diseases (CVD) to the increase in disability and mortality rates worldwide continues to grow steadily. Currently, the main focus is on the problem of atherosclerotic cardiovascular diseases (ACCD) in DM, while the manifestations of cardiac arrhythmias (LDC) in conditions of carbohydrate metabolism disorders have not been sufficiently studied.The purpose of the study. To analyze and compare laboratory, instrumental and objective indicators in patients with type 2 diabetes mellitus (type 2 diabetes mellitus) and without, who have a history of cardiac arrhythmias (LDC).Materials and methods. 103 patients with LDC (52 men and 51 women) were examined. 2 groups were formed: the main group – patients with LDC and type 2 diabetes (n=51) (group 1), the control group – patients with LDC without diabetes (n=52) (group 2). A comparative analysis of laboratory-instrumental and clinical-anamnestic data in the groups was performed. Statistical data processing was performed using the statistical software package "Excel" ("Microsoft"), the program "Statistica 10" ("Statsoft Inc"). The value of p <0.05 was taken as the significance level. The reliability of the differences between the groups was determined using the Mann-Whitney U-test, and the correlation analysis was carried out using the Spearman criterion.Results. The average age of the general group was 67[18;97] years, patients with type 2 diabetes 69 [53;93] years, patients without diabetes 65 [18;97] years. According to the performed regression analysis, it was found that DM in patients was associated with the detection of atrial fibrillation (OR 2.69 (95% CI 1.076;6.71), p=0.034), in particular, its constant form (p=0.025, coefficient. Kramer's V = 0.33), as well as with the presence of supraventricular extrasystole (NJES) (OR 0.235 (95%CI 0.09;0.59), p=0.002, coefficient. Kramer's V = 0.311). At the same time, there was no significant association of DM with the presence of paroxysmal and persistent forms of atrial fibrillation, atrial flutter (TP), BLNPG, AV block, CA block, ventricular extrasystole, supraventricular tachycardia (NVT) and SSSI.When evaluating the functional parameters of the myocardium according to Echo-K G data, it was found that in the main group, compared with the control group, significant (34% vs. 27%) and pronounced (18% vs. 6%) pulmonary hypertension were more common, but the difference was not statistically significant (p>0.05).When assessing the structural and geometric parameters of the myocardium, it was found that concentric myocardial hypertrophy was more common in the main group (22% vs. 15%), and in the control group – concentric myocardial remodeling (14% vs. 25%), (p>0.05).Conclusions. DM is a significant risk factor for a more severe course of LDC, in particular atrial fibrillation and supraventricular ectopias. In this regard, it is of fundamental importance to control carbohydrate metabolism and prevent risk factors for cardiovascular diseases in order to preserve the basic functions of the myocardium and improve the quality of life of patients with diabetes mellitus.

Machine Learning Approaches in Atrial Fibrillation Detection

Atrial fibrillation (AF) characterized by rapid and irregular electrical activity in the atria represents a prevalent form of cardiac arrhythmia that significantly challenges healthcare systems due to its links to heightened mortality and morbidity rates. Early detection of AF is critical for accurate and effective management and treatment. In response to this pressing need, numerous researchers have used machine learning (ML) to enhance the precision and efficiency of AF detection. By analyzing available datasets, signal lengths, preprocessing techniques, and a diverse array of ML approaches, this paper aims to cover methodologies of AF detection using electrocardiogram (ECG) data and ML.

Diagnostic yield of implantable loop recorders: results from the hellenic registry

ObjectiveImplantable loop recorders (ILRs) are increasingly being used for long-term cardiac monitoring in different clinical settings. The aim of this study was to investigate the real-world performance of ILRs—including the time to diagnosis—in unselected patients with different ILR indications. Methods and ResultsIn this multicenter, observational study, 871 patients with an indication of pre-syncope/syncope (61.9%), unexplained palpitations (10.4%), and atrial fibrillation (AF) detection with a history of cryptogenic stroke (CS) (27.7%) underwent ILR implantation. The median follow-up was 28.8 ± 12.9 months. In the presyncope/syncope group, 167 (31%) received a diagnosis established by the device. Kaplan-Meier estimates indicated that 16.9% of patients had a diagnosis at 6 months, and the proportion increased to 22.5% at 1 year. Of 91 patients with palpitations, 20 (22%) received a diagnosis based on the device. The diagnosis was established in 12.2% of patients at 6 months, and the proportion increased to 13.3% at 1 year. Among 241 patients with CS, 47 (19.5%) were diagnosed with AF. The diagnostic yield of the device was 10.4% at 6 months and 12.4% at 1 year. In all cases, oral anticoagulation was initiated. Overall, ILR diagnosis altered the therapeutic strategy in 26.1% of the presyncope/syncope group, 2.2% of the palpitations group, and 3.7% of the CS group in addition to oral anticoagulation initiation. ConclusionIn this real-world patient population, ILR determines diagnosis and initiates new therapeutic management for nearly one-fourth of patients. ILR implantation is valuable in the evaluation of patients with unexplained presyncope/syncope, CS, and palpitations.

Residual-attention deep learning model for atrial fibrillation detection from Holter recordings

Residual-attention deep learning model for atrial fibrillation detection from Holter recordings

Learning From Alarms: A Robust Learning Approach for Accurate Photoplethysmography-Based Atrial Fibrillation Detection Using Eight Million Samples Labeled With Imprecise Arrhythmia Alarms.

Atrial fibrillation (AF) is a common cardiac arrhythmia with serious health consequences if not detected and treated early. Detecting AF using wearable devices with photoplethysmography (PPG) sensors and deep neural networks has demonstrated some success using proprietary algorithms in commercial solutions. However, to improve continuous AF detection in ambulatory settings towards a population-wide screening use case, we face several challenges, one of which is the lack of large-scale labeled training data. To address this challenge, we propose to leverage AF alarms from bedside patient monitors to label concurrent PPG signals, resulting in the largest PPG-AF dataset so far (8.5 M 30-second records from 24,100 patients) and demonstrating a practical approach to build large labeled PPG datasets. Furthermore, we recognize that the AF labels thus obtained contain errors because of false AF alarms generated from imperfect built-in algorithms from bedside monitors. Dealing with label noise with unknown distribution characteristics in this case requires advanced algorithms. We, therefore, introduce and open-source a novel loss design, the cluster membership consistency (CMC) loss, to mitigate label errors. By comparing CMC with state-of-the-art methods selected from a noisy label competition, we demonstrate its superiority in handling label noise in PPG data, resilience to poor-quality signals, and computational efficiency.

Diagnostic accuracy for detecting atrial fibrillation using a novel machine learning algorithm in a blood pressure monitor

BackgroundEarly detection of atrial fibrillation (AF) is key for preventing strokes. Blood pressure monitors (BPMs) with built-in AF screening features have the potential for early detection at home. Recently, 2 BPMs (HEM-7371T1-AZ and HEM-7372T1-AZAZ, Omron Healthcare Co., Ltd.) that share a novel AF screening feature have been developed. Their AF screening feature utilizes an algorithm that incorporates machine learning, with the potential to improve diagnostic accuracy. ObjectiveThe purpose of this study was to evaluate the performance of this AF screening feature in a multicenter, prospective clinical study at 5 sites in the United States. MethodsA total of 559 subjects were enrolled for this study: 267 in AF cohort and 292 in the non-AF cohort. AF screening was performed in all subjects by the 2 Omron BPMs and by 1 Microlife BPM (BP 3MX1-3, WatchBP Home A, Microlife Corp.), and a simultaneous 12-lead electrocardiogram (ECG) was recorded for comparison. All 12-lead ECGs were interpreted by a board-certified cardiologist who was blinded to the BPM results. Sensitivity, specificity, and accuracy for the diagnosis of AF were calculated. ResultsOmron HEM-7371T1-AZ BPM had sensitivity of 95.1% (95% confidence interval [CI] 91.8%–97.4%), specificity 98.6% (95% CI 96.6%–99.7%), and accuracy of 97.0% (95% CI 95.2%–98.2%). Equivalent results were obtained with the Omron HEM-7371T1-AZAZ BPM. This compared favorably to the Microlife BPM (sensitivity 78.5%, 95% CI 73.1%–83.3%; specificity 97.6%, 95% CI 95.1%–99.0%; accuracy 88.4%, 95% CI 85.5%–91.0%). ConclusionThese data support both home and professional use of these novel Omron BPMs for the detection of AF.

Rate of atrial fibrillation by Holter-Stroke Risk Analysis in undetermined TIA/rapidly improving stroke symptoms patients.

Holter-SRA (Stroke Risk Analysis) is an automated analysis of ECG monitoring for Atrial Fibrillation (AF) detection. The aim of this study was to evaluate the rate of AF in undetermined TIA/Rapidly improving stroke symptoms (RISS) patients. Prospective study of undetermined TIA/RISS patients who presented to the emergency department. Early vascular studies (angio CT, transthoracic echocardiography and ECG) were performed in emergency department. The Holter-SRA device was placed for 2 h and the patients were classified into: confirmed AF, high risk of AF or low risk of AF. Prolonged ambulatory monitoring (7 days) was carried out every month for patients with a high-risk pattern. The results were evaluated until definitive detection of AF or low-risk pattern. The endpoints were rate of AF and vascular recurrence at 90 days. Over a period of 24 months, 83 undetermined TIA/RISS patients were enrolled. The mean age was 70 ± 10 years and 61% were men. The median ABCD2 score was 4 points (1-7). After 2 h of monitoring in the emergency department, AF was detected in one patient (1.2%), 51 patients with a low-risk pattern and 31 patients (37.3%) showed a high-risk pattern of AF. During the ambulatory monitoring, of the 31 patients high risk pattern patients, AF was diagnosed to 17 cases and of the 51 patients with a low-risk pattern, one case experienced a recurrent vascular due to undetected AF (1.9% false negative). Three patients (3.6%) suffered a vascular recurrence within the first 90 days, before AF diagnosis. In our study, AF was detected in 22.9% of the 83 patients with indeterminate TIA/RISS. Holter-SRA has allowed us to increase the detection of AF, especially those patients with a high-risk pattern in the first 3 months.