Electrocardiographic Waveforms Research Articles

A deep learning approach for generating intracranial pressure waveforms from extracranial signals routinely measured in the intensive care unit

Intracranial pressure (ICP) is commonly monitored to guide treatment in patients with serious brain disorders such as traumatic brain injury and stroke. Established methods to assess ICP are resource intensive and highly invasive. We hypothesized that ICP waveforms can be computed noninvasively from three extracranial physiological waveforms routinely acquired in the Intensive Care Unit (ICU): arterial blood pressure (ABP), photoplethysmography (PPG), and electrocardiography (ECG). We evaluated over 600 h of high-frequency (125 Hz) simultaneously acquired ICP, ABP, ECG, and PPG waveform data in 10 patients admitted to the ICU with critical brain disorders. The data were segmented in non-overlapping 10-s windows, and ABP, ECG, and PPG waveforms were used to train deep learning (DL) models to re-create concurrent ICP. The predictive performance of six different DL models was evaluated in single- and multi-patient iterations. The mean average error (MAE) ± SD of the best-performing models was 1.34 ± 0.59 mmHg in the single-patient and 5.10 ± 0.11 mmHg in the multi-patient analysis. Ablation analysis was conducted to compare contributions from single physiologic sources and demonstrated statistically indistinguishable performances across the top DL models for each waveform (MAE±SD 6.33 ± 0.73, 6.65 ± 0.96, and 7.30 ± 1.28 mmHg, respectively, for ECG, PPG, and ABP; p = 0.42). Results support the preliminary feasibility and accuracy of DL-enabled continuous noninvasive ICP waveform computation using extracranial physiological waveforms. With refinement and further validation, this method could represent a safer and more accessible alternative to invasive ICP, enabling assessment and treatment in low-resource settings.

ECG Interpretation and Clinical Application in Cardiology Nursing Practice

This study is dedicated to elucidating the significance of electrocardiography (ECG) in the management of Acute Coronary Syndrome (ACS), particularly in diagnosing and making treatment decisions for patients with ST-Segment Elevation Myocardial Infarction (STEMI) and Non-ST-Segment Elevation Myocardial Infarction (NSTEMI). Through a detailed analysis of ECG waveform changes, we demonstrate how to quickly identify signs of myocardial infarction, and based on these findings, decide whether coronary angiography or emergency revascularization surgery is necessary.

Classifying electrocardiograph waveforms using trained deep learning neural network based on wavelet representation

<span lang="EN-US">Due to the rise in cardiac patients, an automated system that can identify different heart disorders has been created to lighten and distribute the duty of physicians. This research uses three different electrocardiograph (ECG) signals as indicators of a person's cardiac problems: Normal sinus rhythm (NSR), arrhythmia (ARR), and congestive heart failure (CHF). The continuous wavelet transform (CWT) provides the mechanism for classifying the 190 individual cases of ECG data into a 2-dimensional time-frequency representation. In this paper, the modified GoogLeNet is used for ECG data classification. Using a transfer learning approach and adjustments to parts of the output layers, ECG classification was conducted and the effectiveness of convolutional neural network (CNN) designs was tested. By comparing the results that the optimized neural network and GoogLeNet both had classification accuracy about of 80% and 100%, respectively. The GoogLeNet provide the best result in term of accuracy and training time.</span>

Predicting Individual Survival Distributions Using ECG: A Deep Learning Approach Utilizing Features Extracted by a Learned Diagnostic Model

In the field of healthcare, individual survival prediction is important for personalized treatment planning. This study presents machine learning algorithms for predicting Individual Survival Distributions (ISD) using electrocardiography (ECG) data in two different formats. The models, which predict time until death, are developed and evaluated on a large, population-based cohort from Alberta, Canada. Our results demonstrate that models trained on raw ECG waveforms significantly outperform those trained on traditional ECG measurements in several metrics, including concordance index, hinge L1 loss, margin L1 loss, and margin truncated L1 loss. Additionally, the integration of predicted probabilities from wide-range diagnostic tasks not only enhances our ISD models' performance but also makes them significantly superior to other models across all evaluation metrics in individual survival prediction tasks. This innovative approach highlights the potential to leverage insights from diagnostic models for prognostic tasks, such as individual survival prediction. These findings could have far-reaching implications for the development of personalized treatment plans and open new avenues for future research in survival prediction using ECGs.

Morbid obesity influences the nocturnal electrocardiogram wave and interval durations among suspected sleep apnea patients.

Obesity is a global issue with a major impact on cardiovascular health. This study explores how obesity influences nocturnal cardiac electrophysiology in suspected obstructive sleep apnea (OSA) patients. We randomly selected 12 patients from each of the five World Health Organization body mass index (BMI) classifications groups (ntotal = 60) while keeping the group's age and sex matched. We evaluated 1965 nocturnal electrocardiography (ECG) samples (10 s) using modified lead II recorded during normal saturation conditions. R-wave peaks were detected and confirmed using dedicated software, with the exclusion of ventricular extrasystoles and artifacts. The duration of waves and intervals was manually marked. The average electric potential graphs were computed for each segment. Thresholds for abnormal ECG waveforms were P-wave > 120 ms, PQ interval > 200 ms, QRS complex > 120 ms for, and QTc > 440 ms. Obesity was significantly (p < .05) associated with prolonged conduction times. Compared to the normal weight (18.5 ≤ BMI < 25) group, the morbidly obese patients (BMI ≥ 40) had a significantly longer P-wave duration (101.7 vs. 117.2 ms), PQ interval (175.8 vs. 198.0 ms), QRS interval (89.9 vs. 97.7 ms), and QTc interval (402.8 vs. 421.2 ms). We further examined ECG waveform prolongations related to BMI. Compared to other patient groups, the morbidly obese patients had the highest number of ECG segments with PQ interval (44% of the ECG samples), QRS duration (14%), and QTc duration (20%) above the normal limits. Morbid obesity predisposes patients to prolongation of cardiac conduction times. This might increase the risk of arrhythmias, stroke, and even sudden cardiac death.

Artificial Intelligence Enabled ECG Interpretation Informs Outcomes after Autologous Transplantation for Multiple Myeloma

High dose chemotherapy followed by autologous hematopoietic cell transplantation (auto HCT) is a standard therapy for patients with plasma cell disorders (PCD). A barrier to improving candidate selection for auto HCT is appraisal of risk for transplant related complications. Electrocardiography (ECG) is a standard pre-HCT test to evaluate the potential for arrhythmia and structural heart disease. We hypothesized that an artificial intelligence (AI) based interpretation of ECG waveform data (AI-ECG) could enhance risk prediction for cardiac-specific and other outcomes following auto HCT for patients with PCD. To test this hypothesis, we used several AI-ECG tools developed using a deep-learning based approach that were trained to predict comorbidity (Attia et al, Lancet 2019 &amp; Grogan et al, Mayo Clin Proc 2021). Model outputs include a scaled prediction value indicating increasing risk for atrial fibrillation/flutter (AF), reduced ejection fraction (EF), or cardiac amyloidosis. We evaluated this tool in 9,225 ECGs obtained from 849 patients (median 6 ECGs/patient, inter-quartile range: 3-12) who underwent first or subsequent auto HCT at Memorial Sloan Kettering Cancer Center between 2016-2022 and who had a diagnosis of multiple myeloma (N = 810), amyloidosis (N = 37), or other PCD (N = 2). The endpoints were to determine whether baseline AI-ECG prediction of AF risk correlated with peri-transplant AF events, and to explore whether model outputs correlated with survival and other cardiac clinical outcomes. The mean age of patients was 61.0 years (range 23.7 - 81.2) and median follow up was 2.7 years post auto HCT. Contributing cardiovascular risk factors included a prior history of AF (N = 101, 11.9%), coronary or cerebrovascular disease (N = 95, 11.2%), and diabetes mellitus (N = 98, 11.5%). Reduced dose melphalan (&amp;lt; 200 mg/m 2) was administered in 26%. The HCT-comorbidity index (HCT-CI) was high risk (≥3) in 52.7%. The one-year overall survival and progression-free survival in the whole cohort was 94% (95% confidence interval (95% CI): 92 - 95%) and 86% (84 - 89%), respectively. Among patients not in AF at pre HCT evaluation, the 90-day cumulative incidence of AF was 5.6% (4.2 - 7.3%). In univariable competing risk regression, baseline pre HCT AI-ECG appraisal of AF risk was strongly correlated with incidence AF in patients who were in sinus rhythm before HCT (hazard ratio (HR): 25.1, 95% CI: 11.7 - 53.7, p &amp;lt; 0.001). In a multivariable model adjusting for patient age and HCT-CI, AI-ECG AF predicted risk remained associated with incidence of post HCT AF (HR: 13.4, 95% CI: 5.9 - 30.4, p &amp;lt; 0.001). Considering a decision threshold identified during original model training, AI-ECG predicted AF events with an accuracy of 75% (Mcnemar's p &amp;lt; 2 x 10 -16). The precision, recall, and F1-score were 0.94, 0.77, and 0.85, respectively. AI-ECG predicted age correlated highly with actual patient age (Spearman rho = 0.59, p &amp;lt; 2 x 10 -16). Amyloid risk scores were highly correlated with a known clinical history of amyloidosis (median 0.45 versus 0.05, Wilcoxon rank-sum p &amp;lt; 0.001). Patients with EF &amp;lt; 50% had greater prediction scores for cardiomyopathy compared to patients with normal EF (median 0.23 versus 0.01, Wilcoxon rank-sum p &amp;lt;0.001). In univariable analysis, increasing AI-ECG predicted age (HR 1.02, CI: 1.003 - 1.04, p = 0.02) and predicted AF risk score (HR: 3.2, CI: 1.4 - 7.2, p = 0.006) were associated with risk for death, whereas predicted amyloid score and predicted low EF were not associated with decreased survival. In multivariable analysis of all patients that considered patient age, hematopoietic comorbidity index, and remission status at HCT, increasing predicted AF score trended towards association with increased risk for death (HR: 2.3, 95%CI: 1.0 - 5.5, p = 0.059). Predicted AF score correlated highly with the HCT-CI and patient age. Finally, we examined whether risk prediction for low EF correlated with peri-HCT fluid overload (FO) defined by absolute weight gain. The incidence of ≥ 10% weight gain (grade 2 FO) and ≥ 20% weight gain (grade 3 FO) at 180 days post HCT was 7.4% (95% CI: 5.8%, 9.3%) and 1.0% (95% CI: 0.5%, 1.8%), respectively. Predicted low EF was not associated with development of either grade 2 or grade 3 FO in this population. In summary, AI-ECG predicts episodes of AF in patients undergoing auto HCT for PCD and may augment transplant risk appraisal when considered in conjunction with other measures of comorbidity.

Noninvasive measurement of blood calcium concentration using electrocardiography in peripartum Jersey cows.

We previously developed a noninvasive method for measuring blood calcium concentration (Ca) in Holstein cows on site using electrocardiographic (ECG) variables and calving number, based on a high positive correlation between Ca. Jersey cows easily develop peripartum hypocalcemia compared with other dairy cows. The early detection and treatment of hypocalcemia are particularly important for Jersey cows because delayed treatment can result in various complications. In this study, to establish a simple, noninvasive, on-site diagnosis of hypocalcemia in perinatal Jersey cows, we attempted to create an equation for estimating Ca using ECG waveforms. Overall, 112 Jersey cows 0-2 days postpartum were used. The ECG findings of these cows were measured using the base-apex lead for 30 s and the corrected ST interval (STc = ST peak interval/SS peak interval0.5) was calculated. Simultaneously, blood was collected from the tail vein, and the serum total Ca (tCa) and serum ionized Ca (iCa) were measured. Several items considered related to Ca were investigated. A strong positive correlation was observed between the tCa and iCa (r = 0.96). A positive correlation was observed between the tCa and STc-1 (r = 0.83). Furthermore, significant correlations were observed between skin temperature, calving number, vigor level, rumen movement, and auricle temperature (p < 0.05). Of these, multiple regression analysis was performed to calculate the tCa estimation formula with the STc and calving number (categorized into primipara, second parity, and third or more parity) as explanatory variables and the tCa as the objective variable (r = 0.85, p < 0.01). Of 15 postpartum Jersey cows, the estimation formula could mostly distinguish between cows with hypocalcemia, those with subclinical hypocalcemia, and normal cows. Blood Ca in peripartum Jersey cows can be noninvasively estimated using ECG variables and calving number.

Discriminating cocaine use from other sympathomimetics using wearable electrocardiographic (ECG) sensors

BackgroundOur group has established the feasibility of using on-body electrocardiographic (ECG) sensors to detect cocaine use in the human laboratory. The purpose of the current study was to test whether ECG sensors and features are capable of discriminating cocaine use from other non-cocaine sympathomimetics. MethodsEleven subjects with cocaine use disorder wore the Zephyr BioHarness™ 3 chest band under six experimental (drug and non-drug) conditions, including 1) laboratory, intravenous cocaine self-administration, 2) after a single oral dose of methylphenidate, 3) during aerobic exercise, 4) during tobacco use (N=7 who smoked tobacco), and 5) during routine activities of daily inpatient living (unit activity). Three ECG-derived feature sets served as primary outcome measures, including 1) the RR interval (i.e., heart rate), 2) a group of ECG interval proxies (i.e., PR, QS, QT and QTc intervals), and 3) the full ECG waveform. Discriminatory power between cocaine and non-cocaine conditions for each of the three outcomes measures was expressed as the area under the receiver operating characteristics (AUROC) curve. ResultsAll three outcomes successfully discriminated cocaine use from unit activity, exercise, tobacco, and methylphenidate conditions with a mean AUROC values ranging from 0.66 to 0.99 and with least squares means values all statistically different/higher than 0.5 among all subjects [F(3, 99) = 3.38, p =0.02] and among those with tobacco use [F(4, 84) = 5.39, p = 0.0007]. ConclusionsThese preliminary results support discriminatory power of wearable ECG sensors for detecting cocaine use.

Towards practical evaluation and development of AI for ECG interpretation

Obtaining a right-chest electrocardiogram is essential for diagnosing concomitant right ventricular infarction in patients with inferior wall acute myocardial infarctions. A software program to synthesize right-chest electrocardiographic waveforms from 12-lead electrocardiographic waveforms is available in Japan. However, its reliability has not been fully investigated. Accordingly, the aim of this study was to examine the reliability of ST-segment shifts in the synthesized V3R to V5R leads. ST-segment shifts in actual and synthesized V3R to V5R leads were compared during the last 10 seconds of 131 balloon inflations while performing elective percutaneous coronary intervention in 56 patients with coronary artery disease. ST-segment shifts in the actual and synthesized V3R, V4R, and V5R leads were correlated (r = 0.96, p <0.001, r = 0.94, p <0.001, and r = 0.91, p <0.001, respectively). A Bland-Altman analysis showed that the bias between ST-segment shifts in the actual and synthesized V3R to V5R leads was −3.1, −5.4, and −4.2 μV, respectively, while the limits of agreement between the ST-segment shifts in the actual and synthesized V3R to V5R leads were −59.2 to 52.9, −61.9 to 51.1, and −59.7 to 51.3 μV, respectively. The κ coefficients for ST-segment elevation of ≥50 and ≥100 μV in the actual and synthesized V3R, V4R, and V5R leads were 0.83 and 0.81, 0.66 and 0.83, and 0.57 and 0.80, respectively. In conclusion, these results indicate that ST-segment shifts in the synthesized V3R to V5R leads have acceptable reliability, suggesting that synthesized right-chest electrocardiography can be used to diagnose concomitant right ventricular infarction in patients with inferior wall acute myocardial infarctions.

Electrocardiographic Waveform Abnormalities in Patients With Structural Heart Disease

Scenario: This 12-lead electrocardiogram (ECG) was obtained in a 72-year-old man arriving in the emergency department with shortness of breath and dizziness. Prior medical history includes myocardial infarction (MI), heart failure (HF), chronic obstructive pulmonary disease (COPD), diabetes (type 2), and paroxysmal atrial fibrillation. Current prescribed medications include a β-blocker, aspirin, angiotensin-converting enzyme inhibitor, furosemide, and diltiazem. His blood pressure was 98/50 mm Hg, respiratory rate 24/min, and oxygen saturation as measured by pulse oximetry (Spo2) was 92% on 3 L of oxygen. He was most comfortable sitting up with his arms on the bedside table, and said he felt better after 40 mg of furosemide was administered intravenously.Sinus rhythm with premature atrial complexes (PACs), possible right and left atrial enlargement, prior inferior wall MI (q waves in leads III and aVF), incomplete left intraventricular conduction defect (fQRS waves), left ventricular hypertrophy, and ST-segment and T-wave abnormalities may suggest anterolateral ischemia.Although the underlying rhythm is irregular, suggesting atrial fibrillation, P waves are visible before every QRS complex; thus, this is sinus rhythm with frequent PACs. PACs are not uncommon in patients with structural heart disease and COPD, like this one. The pause after the PACs is a normal physiologic response and represents a compensatory pause associated with PACs. The ECG criterion for right atrial enlargement is a P-wave amplitude greater than 2.5 mm in leads II, III, and aVF (so-called P pulmonale) and can be seen with COPD. Left atrial enlargement is due to pressure or volume overload of the left atrium and can be assessed in V1 (biphasic P waves with negative terminal portion >40 ms long and >1 mm deep). Biatrial enlargement is consistent with this patient’s medical history (MI, HF, and left ventricular hypertrophy); however, ECG is not the reference standard method, so the diagnosis should be confirmed by echocardiography. The T-wave inversion and ST-segment depression most likely represent a “strain” pattern seen with left ventricular hypertrophy, but acute myocardial ischemia should be ruled out given the patient’s prior MI.Immediate management includes addressing the acute respiratory distress and low blood pressure, with HF exacerbation a likely source. This patient’s symptoms were somewhat relieved after the intravenous furosemide, which supports this diagnosis. Checking a venous blood sample for brain natriuretic peptide (BNP) to assess for HF, troponin to rule out myocardial ischemia, and electrolyte levels would help. A chest radiograph would be useful, as might arterial blood gas analysis, given his elevated rate of breathing and the 92% Spo2 with 3 L of oxygen. Findings on physical examination, especially lung and heart sounds, peripheral edema, and daily weight should be evaluated. Ultimately, this patient’s BNP level was 3400 ng/mL, which corroborates HF, and his troponin was negative for myocardial ischemia. He was discharged home 3 days later.

An Open-Source, Interoperable Architecture for Generating Real-Time Surgical Team Cognitive Alerts from Heart-Rate Variability Monitoring.

Clinical alarm and decision support systems that lack clinical context may create non-actionable nuisance alarms that are not clinically relevant and can cause distractions during the most difficult moments of a surgery. We present a novel, interoperable, real-time system for adding contextual awareness to clinical systems by monitoring the heart-rate variability (HRV) of clinical team members. We designed an architecture for real-time capture, analysis, and presentation of HRV data from multiple clinicians and implemented this architecture as an application and device interfaces on the open-source OpenICE interoperability platform. In this work, we extend OpenICE with new capabilities to support the needs of the context-aware OR including a modularized data pipeline for simultaneously processing real-time electrocardiographic (ECG) waveforms from multiple clinicians to create estimates of their individual cognitive load. The system is built with standardized interfaces that allow for free interchange of software and hardware components including sensor devices, ECG filtering and beat detection algorithms, HRV metric calculations, and individual and team alerts based on changes in metrics. By integrating contextual cues and team member state into a unified process model, we believe future clinical applications will be able to emulate some of these behaviors to provide context-aware information to improve the safety and quality of surgical interventions.

Practical R-R Interval Editing for Heart Rate Variability Analysis Using Single-Channel Wearable ECG Devices

Recent innovations in wearable devices have expanded the usage opportunities of single-channel electrocardiography (ECG) recordings in a daily life environment and enabled a variety of indirect daily activity monitoring based on heart rate variability (HRV). In general, wearable ECGs rarely undergo visual inspection by medical experts and therefore may contain noise or artifacts. Although noise/artifact-induced changes in ECG waveforms are known to cause misdetection of the QRS complex (i.e., the most distinguishable ECG components comprised of Q wave, R wave, and S wave), its complete suppression might be technically impossible. Since misdetection occurs in the QRS complex unit, we propose reframing the traditional HRV analysis flow by subdividing the R-R interval (RRI) editing into four steps in accordance with the processing detail (i.e., identification and editing) and its target unit (i.e., QRS complex or RRI). In addition, as a dubious QRS complex identification method for practical use, we utilize the amplitude at the detected point assuming the use of a single-channel wearable ECG without a reference. Initial evaluations using pseudo/real ECG datasets including ECGs with noise and artifacts show that the proposed processing/unit-based subdivision is theoretically effective for improving HRV calculation accuracy, and that the dubious QRS complex identification method for practical use also maintains this effect. Our study starting from practical HRV analysis using single-channel wearable ECG devices encourages reexamining each step in HRV analysis through the interdisciplinary research of clinical medicine and engineering/informatics that reveals the relationship of every two adjacent steps from the perspective of theory and practice.

The electrocardiographic changes generated by centrally applied arachidonic acid in rats

Arachidonic acid (AA) and its metabolites have multifunctional regulatory effects on the central nervous system. Our previous reports disclosed that centrally injected AA organized the cardiovascular system in normal or hypotensive conditions by regulating the central and peripheral mechanism. In the light of the knowledge of the potential cardiovascular effects of AA, the current study aimed to investigate the effects of intracerebroventricular (ICV) injected AA on the electrocardiography (ECG) of the anesthetized rats. The adult Sprague Dawley rats were anesthetized with ketamine and xylazine mixture (50 mg/kg and 20 mg/kg; i.m., respectively). Under the anesthesia, the guide cannula was inserted into the left lateral ventricle of the rats. The ECG traces obtained from the lead II were written by placing electrodes on the limbs of the rats. Centrally injected AA (150 μg; ICV) statistically significantly (p&amp;lt;0.05) caused to the lengthening of the ECG waves and intervals, resulting in a decrease in the heart rate of the rats without changing the ECG waveforms, the amplitude, and also the isoelectric line. The obtained results clearly show that centrally injection of AA caused the deceleration in the heart electrical activity. The deceleration in the electrical activity of the heart caused to show bradycardia in the rats by extending the duration of the ECG waves and intervals.

Electrocardiograph Signals Diagnosis Using Adaptive Neuro-Fuzzy Inference System

Electrocardiograph (ECG) is a bioelectrical signal that is obtained by non-invasive method to register the electrical activities of the heart. This paper provides an attempt to develop computerized system for ECG signal filtering and classification. The proposed system encompass: pre-processing of the signal, extraction of pattern features through independent component analysis (ICA), power spectrum, and RR interval calculation. These processes provide an input feature vector to the Adaptive Neuro Fuzzy Inference System (ANFIS) that acts as a signal classifier. All of the classification process steps are implemented in MATLAB environment. This paper aslo provides a graphical User Interface (GUI) that makes classification process easier. Three cases of ECG waveforms that are selected from MIT-BIH database are considered for the system test; they are Normal (N), Ventricle Fibrillation (VF), and Ventricular tachycardia (VTachy). An accuracy of 96.66% has been achieved by the proposed system.

Heart Rate Variability in Children with Moderate and Severe Traumatic Brain Injury: A Prospective Observational Study

AbstractThe aim of this study was to assess the feasibility of continuous monitoring of heart rate variability (HRV) in children with traumatic brain injury (TBI) hospitalized in a pediatric intensive care unit (PICU) and collect preliminary data on the association between HRV, neurological outcome, and complications. This is a prospective observational cohort study in a tertiary academic PICU. Children admitted to the PICU ≤24 hours after moderate or severe TBI were included in the study. Children suspected of being brain dead at PICU entry or with a pacemaker were excluded. Children underwent continuous monitoring of electrocardiographic (ECG) waveforms over 7 days post-TBI. HRV analysis was performed retrospectively, using a standardized, validated HRV analysis software (CIMVA). The occurrence of medical complications (“event”: intracranial hypertension, cerebral hypoperfusion, seizure, and cardiac arrest) was prospectively documented. Outcome of children 6 months post-TBI was assessed using the Glasgow Outcome Scale – Extended Pediatric (GOS-E Peds). Fifteen patients were included over a 20-month period. Thirteen patients had ECG recordings available and 4 had &gt;20% of missing ECG data. When ECG was available, HRV calculation was feasible (average 88%; range 70–97%). Significant decrease in overall HRV coefficient of variation and Poincaré SD2 (p &lt; 0.05) at 6 hours post–PICU admission was associated with an unfavorable outcome (defined as GOS-E Peds ≥ 3, or a deterioration of ≥2 points over baseline score). Several HRV metrics exhibited significant and nonsignificant variation in HRV during event. This study demonstrates that it is feasible to monitor HRV in the PICU provided ECG data are available; however, missing ECG data are not uncommon. These preliminary data suggest that altered HRV is associated with unfavorable neurological outcome and in-hospital medical complications. Larger prospective studies are needed to confirm these findings and to explore if HRV offers reliable and clinically useful prediction data that may help clinical decision making.

59 THE SHARK-FIN THAT EMERGES AND IMMERGES: A CASE REPORT OF TRANSIENT AND RIPETITIVE TRIANGULAR QRS-ST-T.

Abstract Introduction Shark fin sign, also known as triangular QRS-ST-T waveform (TW), is characterized by a giant R wave (amplitude ≥ 1 mV) resulting from the fusion of the QRS complex, the ST-segment and the T wave and showing triangular morphology. TW pattern is quite uncommon and reveals a life-threatening condition that needs aggressive management. Early recognition is important to prevent acute complications such as ventricular arrhythmias and cardiogenic shock. Case presentation A 45-years old black man with history of arterial hypertension presented with out-of-hospital cardiac arrest (OHCA) at his workplace. Extracorporeal cardiopulmonary resuscitation started immediately and return of spontaneous circulation occurred after 11 minutes and 2 DC shocks. Then he was transferred to our emergency department (ED). At arrival he was hemodynamically stable, ECG showed atrial fibrillation without any repolarization abnormalities and echocardiogram revealed left ventricular hypertrophy (IVS 12mm, PW 12mm) and normal biventricular function. Normal electrolytes and high-sensitivity troponin level of 14 (n.v. 0.01-19.8 ng/L) were found at the blood test. Sedation in the ED was needed due to psychomotor agitation. He was then transferred to the CCU where spontaneous cardioversion to sinus rhythm occurred, and no additional arrhythmias were noted at continuous ECG monitoring. After ten hours, while the patient was still under sedation, ECG monitoring revealed a progressive QRS-ST segment elevation in inferior leads up to a triangular QRS-ST-T giant waveform (“Shark fin sign”), without hemodynamic deterioration. After approximately 3 minutes the ECG showed a progressive and complete regression to a normal waveform. An emergent beside echocardiogram demonstrated inferior wall hypokinesia, but normal pump function. The episode repeated 5 more times in the following hour, waiting for cath lab personnel. Coronary angiography revealed a sub-occlusive mid RCA stenosis without any thrombotic lesion. Percutaneous coronary intervention (PCI) with the implantation of one drug eluting stent (DES) was performed. After the procedure the patient was hemodynamically and electrically stable. ECG did not show any other abnormality and no episodes of recurrent ST-elevation occurred. Sedation was suspended and no focal neurological deficits were noted. Discussion Shark fin sign is the expression of abrupt total deprivation of blood flow in large areas of myocardium that cause massive slowing of transmural electrical conduction. These electrical differences may be at the base of the triangular R waves and predispose to VF induction and cardiogenic shock. This sign is very rare, with an annual incidence of 0.7% in patients admitted for STEMI, and in most of the cases is described in the occlusion of the LAD. In our case we assisted to transient but repetitive inferior triangular ST-elevation with complete regression to normal ECG. The coronary angiogram showed non-obstructive lesion of the RCA. This phenomenon was probably the expression of RCA's plaque coronary vasospasm that caused the ECG shark fin sign and cardiac arrest. Conclusion Shark fin electrocardiographic waveform is an uncommon pattern of STEMI and may be the expression of thrombotic coronary occlusion as well as massive coronary vasospasm. This sign correlates with high risk of cardiogenic shock, electrical instability, and cardiac arrest, thus requiring prompt and aggressive management.

Abstract 12575: Subgroup Comparison of Deep-Learning Electrocardiogram Model Performance for Predicting New-Onset 1-Year Atrial Fibrillation Incidence

Background: Surveillance of high-risk individuals as determined by an atrial fibrillation (AF) prediction model would be a good strategy to prevent stroke and other complications from AF. Recently, several deep learning (DL) models to predict new onset AF using 12-lead electrocardiography (ECG) have been introduced. However, their performance has not been compared among different patient groups. Methods : Data of 99,281 patients who underwent 12-lead ECG at Korea University Anam Hospital from January 2017 to December 2019, excluding patients already diagnosed with AF, were studied. The data was split into a 60% training set, 20% validation set and 20% test set. Using age, sex, and raw ECG waveforms as inputs, DL models of ResNet and a multi-layered perceptron module were developed to predict 1-year new onset AF incidence. Results: The DL model using both ECG waves and simple demographic information (age and sex) as input (ECG-AS model) showed better performance compared to the DL model using only ECG waves (ECG model) for AF prediction in the test dataset (area under the receiver operating characteristics curves [AUROCs] 0.79 for ECG model and 0.89 for ECG-AS model, respectively). In subgroup analysis divided by clinical features such as CHA2DS2-VASC score, there was little difference in the performance of the DL model. Interestingly, The DL model performed worse in patients with normal sinus rhythm (NSR) compared to patients with non-NSR (AUROC 0.70 vs 0.81 for ECG model, 0.82 vs 0.90 for ECG-AS model, Table 1). Conclusions: ECG of NSR may not have sufficient information to predict AF. In applying the AF prediction DL model using 12-lead ECG to real-world clinical practice, it may be helpful to add clinical information to the input along with appropriate patient selection.

Abstract 11190: Subgroup Comparison of Electrocardiogram Deep-Learning Model Performance for Estimating Coronary Artery Calcium Score

Introduction: Despite the clinical usefulness of Coronary artery calcium score (CACS), it is difficult to use it as a routine practice due to its high cost and radiation hazard. Recently, several deep learning (DL) models to estimate CACS using 12-lead electrocardiography (ECG) have been introduced. However, their performance has not been compared among different patient groups. This study compared the performance of DL models to estimate CACS in different patient groups. Methods: The data of 5,427 patients who underwent both CT scan and 12-lead ECG at Korea University Anam Hospital from November 2012 to December 2021 was split into 80% training set and 20% test set. CNN-based DL model for predicting CACS ≥ 1 and CACS ≥ 100 using the raw ECG waveforms and several clinical factors as input was developed. Results: The DL model showed comparable performance for CACS estimation in the test dataset (area under the receiver operating characteristics curves [AUROCs] 0.77 for CACS ≥ 1 and 0.82 for CACS ≥ 100). The DL model showed better performance in patients with cardiovascular risks less than 3 compared to patients with cardiovascular risks more than 3 (AUROC 0.75 vs 0.71 for CACS ≥ 1, 0.77 vs 0.71 for CACS ≥ 100, Table 1). Among subgroups, the elderly (age &gt; 65 year-old) showed the lowest performance (AUROC 0.72 for both CACS ≥ 1 and ≥ 100). Interestingly, the model performance predicting CACS ≥ 1 is better in patients with normal sinus rhythm, whereas the model performance predicting CACS ≥ 100 is better in patients with non-normal sinus rhythm. This suggests that normal ECG can better predict healthy condition and abnormal ECG can better predict pathological condition. Conclusions: The performance of CACS estimation DL model could depend on the patient's cardiovascular risk factors. When applying the DL model to actual clinical practice, it is necessary to consider the performance of the DL model for the patient group to be applied and to selectively apply the appropriate DL model.

Real time system on chip based wearable cardiac activity monitoring sensor

In this research, a real time low power wearable cardiac activity monitoring device is introduced to display real time electrocardiography (ECG) signal and heart rate (HR). ECG records heart’s electrical activity and heart rate measures the contraction of the heart. The developed device consumes very less power as SoC used in this research has power consumption of 85 μW at 1.1V supply voltage. To measure heart rate and ECG waveform, the pre-gelled ECG electrodes are attached on the RA, LA and LL positon of the human’s body. These pre-gelled ECG electrodes attached to the ECG lead which is further connected to the ECG device. The ECG system is validated and the measured average HR of the subject was 71 to 82 beats per minute (BPM) and ECG graph was also plotted.

A comparative study on neural networks for paroxysmal atrial fibrillation events detection from electrocardiography

ObjectiveThis work conducts a comparative study on the effect of neural networks of different architectures on the detection of paroxysmal atrial fibrillation (PAF) events from dynamic electrocardiography (ECG) recordings, a problem raised in the 4th China Physiological Signal Challenge 2021 (CPSC2021). ApproachWe proposed 3 neural network models and an auxiliary one for QRS detection to tackle the problem. A convolutional recurrent neural network (CRNN) model and a U-Net model that accepts ECG waveform input make sample-wise predictions. This regards the PAF events detection as a segmentation task. A stacked bidirectional long short-term memory (LSTM) model takes the sequence of RR intervals, which is derived from the output of the QRS detection model and makes beat-wise predictions. The QRS detection model also has a CRNN architecture, which is slightly different from the model for the AF segmentation task. Final predictions are merged by outputs from models making sample-wise predictions and making beat-wise predictions. Finally, the locations of QRS complexes are used to filter out segments (both normal and AF) shorter than 5 beats.In order to make the neural network models more sensitive to the critical sample points (onsets and offsets) of the AF events, we proposed a novel masked binary cross-entropy (MaskedBCE) loss function for training the models. This loss function is the conventional BCE loss multiplied by a mask, whose values in a neighbourhood of critical sample points are significantly larger than elsewhere. Main resultsOur method received a score of 1.9972 on the first part of the hidden test set of CPSC2021 and a score of 3.0907 on the second part. The average score was 2.5440, ranked 5th out of 17 teams with successful official entries. SignificanceThis work proposed an effective solution to the problem of the detection of PAF events from dynamic ECGs and validated the efficacy of several neural network architectures on this problem.