Attia, Zachi I.

Abstract Aims An artificial intelligence-augmented electrocardiogram (AI-ECG) algorithm can identify left ventricular systolic dysfunction (LVSD). We sought to determine whether this AI-ECG algorithm could stratify mortality risk in cardiac intensive care unit (CICU) patients, independent of the presence of LVSD by transthoracic echocardiography (TTE). Methods and results We included 11 266 unique Mayo Clinic CICU patients admitted from 2007 to 2018 who underwent AI-ECG after CICU admission. Left ventricular ejection fraction (LVEF) data were extracted for patients with a TTE during hospitalization. Hospital mortality was analysed using multivariable logistic regression. Mean age was 68 ± 15 years, including 37% females. Higher AI-ECG probability of LVSD remained associated with higher hospital mortality [adjusted odds ratio (OR) 1.05 per 0.1 higher, 95% confidence interval (CI) 1.02–1.08, P = 0.003] after adjustment for LVEF, which itself was inversely related with the risk of hospital mortality (adjusted OR 0.96 per 5% higher, 95% CI 0.93–0.99, P = 0.02). Patients with available LVEF data (n = 8242) were divided based on the presence of predicted (by AI-ECG) vs. observed (by TTE) LVSD (defined as LVEF ≤ 35%), using TTE as the gold standard. A stepwise increase in hospital mortality was observed for patients with a true negative, false positive, false negative, and true positive AI-ECG. Conclusion The AI-ECG prediction of LVSD is associated with hospital mortality in CICU patients, affording risk stratification in addition to that provided by echocardiographic LVEF. Our results emphasize the prognostic value of electrocardiographic patterns reflecting underlying myocardial disease that are recognized by the AI-ECG.

Abstract Background Left ventricular systolic dysfunction (LVSD) is associated with increased morbidity and mortality. Although there are effective treatments for patients with LVSD to prevent mortality, heart failure and to improve symptoms, many patients remain undetected and untreated. We have recently derived a deep learning algorithm to detect LVSD using the electrocardiogram (ECG) which could have an important screening role, particularly in limited resources settings. We evaluated the accuracy of this algorithm for the first time in Africa in a sample of subjects attending a cardiology clinic. Methods We conducted a retrospective study in a general cardiac clinic in Uganda. Consecutive patients ≥18 years who had a digital ECG and echocardiogram done within two days of each other were included. We excluded patients with pacemakers or missing information regarding left ventricular ejection fraction (LVEF). Routine 10-second, twelve-lead surface rest ECG were performed using an Edan PC ECG Model SE-1515, Hamburg, Germany. The probability of LVSD was estimated with the Mayo Clinic artificial intelligence (AI) ECG algorithm. LVEF was calculated by the MMode (Teichholz method) using a Philips Ultrasound system, HD7XE, Bothel, Washington, USA. LVSD was defined as a LVEF≤35%. We assessed the overall diagnostic performance of the algorithm to identify LVSD in this population with the area under the receiver operating curve (AUC), and estimated sensitivity, specificity and accuracy using a pre-specified cut-off based on the probability for LVSD generated by the algorithm. We conducted secondary analyses using different LVEF cutoff values. Results We included 634 subjects, 32% (200) of whom had hypertension and 12% (77) clinical heart failure. Mean age was 57±18.8 years, 58% were women and the overall prevalence of LVSD was 4%. The AI-ECG had an AUC of 0.866 (see figure below), sensitivity 73.08%, specificity 91.10%, negative predictive value 98.75%, positive predictive value 26.03% and an accuracy of 90.96% using the original threshold. Using the optimal cutoff based on the AUCs, the sensitivity was 80.77% and specificity was 81.05% with a negative predictive value of 98.99%. The ROC for the detection of LVEF of 40% or below was 0.821. Conclusion The Mayo AI-ECG algorithm demonstrated good accuracy, sensitivity and specificity to detect LVSD in patients seen in a clinical setting in Uganda. This tool may facilitate the identification of people at a high risk for LVSD in settings with low resources. ROC Funding Acknowledgement Type of funding source: None

Abstract Background We have demonstrated that artificial intelligence interpretation of ECGs (AI-ECG) can estimate an individual's physiologic age and that the gap between AI-ECG and chronologic age (Age-Gap) is associated with increased mortality. We hypothesized that Age-Gap would predict long-term atherosclerotic cardiovascular disease (ASCVD) and that Age-Gap would refine the ACC/AHA Pooled Cohort Equations' (PCE) predictive abilities. Methods Using the Rochester Epidemiology Project (REP) we evaluated a community-based cohort of consecutive patients seeking primary care between 1998–2000 and followed through March 2016. Inclusion criteria were age 40–79 and complete data to calculate PCE. We excluded those with known ASCVD, AF, HF or an event within 30 days of baseline.A neural network, trained, validated, and tested in an independent cohort of ∼ 500,000 independent patients, using 10-second digital samples of raw, 12 lead ECGs. PCE was categorized as low<5%, intermediate 5–9.9%, high 10–19.9%, and very high≥20%. The primary endpoint was ASCVD and included fatal and non-fatal myocardial infarction and ischemic stroke; the secondary endpoint also included coronary revascularization [Percutaneous Coronary Intervention (PCI) or Coronary Artery Bypass Graft (CABG)], TIA and Cardiovascular mortality. Events were validated in duplicate. Follow-up was truncated at 10 years for PCE analysis. The association between Age-Gap with ASCVD and expanded ASCVD was assessed with cox proportional hazard models that adjusted for chronological age, sex and risk factors. Models were stratified by PCE risk categories to evaluate the effect of PCE predicted risk. Results We included 24,793 patients (54% women, 95% Caucasian) with mean follow up of 12.6±5.1 years. 2,366 (9.5%) developed ASCVD events and 3,401 (13.7%) the expanded ASCVD. Mean chronologic age was 53.6±11.6 years and the AI-ECG age was 54.5±10.9 years, R2=0.7865, p<0.0001. The mean Age-Gap was 0.87±7.38 years. After adjusting for age and sex, those considered older by ECG, compared to their chronologic age had a higher risk for ASCVD when compared to those with <−2 SD age gap (considered younger by ECG). (Figure 1A), with similar results when using the expanded definition of ASCVD (data not shown). Furthermore, Age-Gap enhanced predicted capabilities of the PCE among those with low 10-year predicted risk (<5%): Age and sex adjusted HR 4.73, 95% CI 1.42–15.74, p-value=0.01 and among those with high predicted risk (>20%) age and sex adjusted HR 6.90, 95% CI 1.98–24.08, p-value=0.0006, when comparing those older to younger by ECG respectively (Figure 1B). Conclusion The difference between physiologic AI-ECG age and chronologic age is associated with long-term ASCVD, and enhances current risk calculators (PCE) ability to identify high and low risk individuals. This may help identify individuals who should or should not be treated with newer, expensive risk-reducing therapies. Funding Acknowledgement Type of funding source: Foundation. Main funding source(s): Mayo Clinic

Abstract Background An Artificial intelligence (AI)-enabled algorithm can provide an age estimate from a single standard resting ECG. The gap between AI-estimated age (ECG age) and chronological age (Age-Gap) has been associated with total and cardiovascular mortality. We hypothesized that coronary endothelial dysfunction (CED), an early feature of coronary atherosclerosis, is associated with physiological aging, as measured by ECG age. Purpose The aim of this study is to investigate the association of CED, a potential index of cardiac aging, and AI-estimated physiological aging. Methods 1902 patients with signs and/or symptoms of ischemia and with non-obstructive coronary artery disease (<40% angiographic stenosis) who underwent functional coronary angiography with intracoronary (IC) acetylcholine (Ach) infusions and a resting ECG within one year were included in this analysis. CED was defined as coronary artery constriction [%ΔCoronary artery diameter (CAD) <0] in response to IC Ach. Patients were categorized as CED+ or CED−. The ECG age were calculated by a previously-trained AI algorithm that was built using the Keras framework with Tensorflow backend and Python. The association between ECG-age and Age-Gap, with CED was investigated. Results Average chronological age was 50.9±12.5, ECG age was 54.1±11.6, 1261 (66%) were females. Compared to CED−, the CED+ group had more males and hyperlipidaemia (p<0.05 for both). Age-gap was significantly higher in patients with CED+ (Fig. A). Multiple regression analysis showed that CED+ was associated with Age-gap (standardized β [sβ] coefficient = 0.08, P=0.001) even after adjustment for chronological age (sβ coefficient = 0.07, P=0.001), and even after adjusting for other cardiovascular risk factors like gender and obesity (sβ coefficient = 0.07, P<0.0001). Furthermore, there is a significant gradual increase in vasoconstriction, in response to IC Ach, with increasing Age-gap (Fig. B). Conclusion Abnormal coronary endothelial function is associated with an increased AI-estimated age which could indicate a higher physiological age. Funding Acknowledgement Type of funding source: None

Abstract Background Artificial Intelligence (AI) algorithms enabled the detection of patients with paroxysmal atrial fibrillation (PAF) from a single normal sinus rhythm (NSR) ECG. Coronary microvascular dysfunction (CMD) is a precursor for coronary artery disease, which is a known risk factor for AF. Purpose The aim of this study is to examine the probability of PAF, according to AI-enabled algorithm estimation, in patients with CMD. Methods 1858 patients without persistent atrial fibrillation with signs and/or symptoms of ischemia and with non-obstructive CAD (<40% stenosis) who underwent invasive coronary microvascular functional assessment and the ECG closest to the functional assessment were included in this analysis. Patients with coronary flow velocity reserve (CFR) <2 in response to adenosine were labelled as endothelial-independent CMD; % increase in coronary blood flow (%ΔCBF) <50% in response to acetylcholine were labelled as endothelial-dependent CMD. Patients were categorized into 4 groups. G1: Normal (NL) CFR/NL %ΔCBF; G2: Abnormal (ABN) %ΔCBF only; G3: ABN CFR only; G4: ABL CFR & %ΔCBF. The probability of having PAF (%probAF) was calculated by a previously-trained and validated AI algorithm. AF Flag = %probAF >9%; which is a pre-set cut-off found to have the highest accuracy of identifying patients with PAF (Area Under the Curve = 0.87). Results Mean age for patients was 51.2±12.4 and 66.3% were females. 835 (45%) were in G1, 39 (2%) in G2, 911 (49%) in G3, and 73 (4%) in G4. Compared to G1 and G2, G3 and G4 were older, had more diabetes and higher smoking rates (p<0.05 for all). Furthermore, G4 had a significantly higher %probAF compared to other groups (Fig. 1). G4 were also more likely to be flagged by the algorithm as having PAF, even after adjusting for cardiovascular risk factors, with an odds ratio of 1.9 [CI 95% 1.1–3.3; p=0.03]) (Fig. 2). Conclusion Patients with combined CMD have a significantly higher probability of having PAF based on an AI-enabled algorithm. Further research is warranted to know if patients with CMD would benefit from formal AF screening at the time of diagnosis. Funding Acknowledgement Type of funding source: None

Background: An artificial intelligence (AI) algorithm applied to electrocardiography during sinus rhythm has recently been shown to detect concurrent episodic atrial fibrillation (AF). We sought to characterize the value of AI–enabled electrocardiography (AI-ECG) as a predictor of future AF and assess its performance compared with the CHARGE-AF score (Cohorts for Aging and Research in Genomic Epidemiology–AF) in a population-based sample. Methods: We calculated the probability of AF using AI-ECG, among participants in the population-based Mayo Clinic Study of Aging who had no history of AF at the time of the baseline study visit. Cox proportional hazards models were fit to assess the independent prognostic value and interaction between AI-ECG AF model output and CHARGE-AF score. C statistics were calculated for AI-ECG AF model output, CHARGE-AF score, and combined AI-ECG and CHARGE-AF score. Results: A total of 1936 participants with median age 75.8 (interquartile range, 70.4–81.8) years and median CHARGE-AF score 14.0 (IQR, 13.2–14.7) were included in the analysis. Participants with AI-ECG AF model output of >0.5 at the baseline visit had cumulative incidence of AF 21.5% at 2 years and 52.2% at 10 years. When included in the same model, both AI-ECG AF model output (hazard ratio, 1.76 per SD after logit transformation [95% CI, 1.51–2.04]) and CHARGE-AF score (hazard ratio, 1.90 per SD [95% CI, 1.58–2.28]) independently predicted future AF without significant interaction ( P =0.54). C statistics were 0.69 (95% CI, 0.66–0.72) for AI-ECG AF model output, 0.69 (95% CI, 0.66–0.71) for CHARGE-AF, and 0.72 (95% CI, 0.69–0.75) for combined AI-ECG and CHARGE-AF score. Conclusions: In the present study, both the AI-ECG AF model output and CHARGE-AF score independently predicted incident AF. The AI-ECG may offer a means to assess risk with a single test and without requiring manual or automated clinical data abstraction.

Background: Myocardial injury results in release of cardiac troponin (cTn) into the bloodstream, readily detected by high-sensitivity cTn (hs-cTn) assays. A noninvasive, rapid, home-based test to exclude acute myocardial injury at the time of the test and for the next 7 hours would help manage patients with suspected myocardial infarction. Since myocyte injury is associated with ECG changes, we hypothesized an artificial intelligence ECG (AI-ECG) could identify absence of injury. Objective: To train and test an AI-ECG convolutional neural network (CNN) to identify patients using a single ECG who are suspected of myocardial infarction who are low risk, with hs-cTn levels below the 99 th percentile at test time and for the subsequent 7 hours. Methods: A CNN tuned to identify the absence of a hs-cTnT (5 th Gen cTnT Roche Diagnostics) >15ng/L for men and >10ng/L for women was developed. All ECGs were recorded within one hour of the hs-cTnT assay. We used 73,012 ECGs and hs-cTnT pairs from 47,542 unique patients to train the network, 9031 ECGs from 5,811 patients for internal validation to optimize hyperparameters, and 11,904 ECGs with 21,191 hs-cTnT measurements up to 7 hours after the ECG, from 11,904 different patients as a holdout test set. Results: The mean age was 63.9±17.5 years, and 30,348 of the 59,446 patients (51%) were male. 5,852 patients (49.1%) had no elevation of hs-cTnT and 6,052 (50.9%) had an hs-cTnT above the 99 th percentile at baseline or within 7 hours of the test. Of the 11,904 patients in the test set, using a sensitive threshold, the 12 lead AI ECG identified 1037 patients (8.7%) likely to have a low risk for subsequent hs-cTnT increases > 99 th % (AUC 0.86), and the single lead ECG identified 685 patients. Of the 1037 low risk pts, 59 had an hs-cTnT > 99 th percentile within 7 hours. The mean maximum hs-cTnT among the 59 low-risk patients was 53 ng/L ± 92 vs 184 ng/L ± 1474 in the others. None of these low risk patients died within 14 days of the test. Conclusion: Use of the AI ECG to identify patients without cardiac injury at the time of the test and for 7 subsequent hours is feasible. This may permit rapid, point of care, home-based, bloodless, painless testing to exclude cardiac injury, facilitating care and management of patients with suspected ischemia.

Background: High-sensitivity cardiac troponin (hs-cTn) assays quantify cTn in patients at very low concentrations. Myocyte injury due to ischemia or other pathologies cause blood levels to increase, which is prognostic. A noninvasive, rapid, broadly available, home-based test to detect hs-cTn increases would facilitate risk-stratification. Since myocyte injury is associated with ECG changes, we hypothesized an artificial intelligence ECG (AI-ECG) could non-invasively predict current or impending hs-cTnT elevations. Objective: To develop an AI-ECG convolutional neural network (CNN) to detect an abnormal hs-cTnT (5 th Gen cTnT, Roche Diagnostics) concentration using a 12-lead ECG, and a single lead ECG (lead I), which would enable smartphone, home-based detection. Methods: We developed single lead and 12-lead ECG CNNs to detect a) hs-cTnT concentrations that were at or above the 6ng/L limit that can be reported b) above the 99 th percentile upper limits of >15ng/L for men and >10ng/L for women. All ECGs were recorded within one hour of the hs-cTnT measurements. We used 73,012 ECG and hs-cTnT pairs from 47,542 unique patients to train the network, 9031 ECGs from 5,811 patients for internal validation to optimize hyperparameters, and 18,276 ECG and hs-cTnT pairs from 11,904 different patients as a holdout test set to determine the area under the receiver-operator curve (AUC). Results: The mean age was 63.9±17.5 years, and 30,348 of the 59,446 patients (51%) were male. Of the 91,288 hs-cTnT pairs 73,271 (80.2%) were above 6ng/L and 50,799 (55.6%) are above the 99 th percentile. In the test set, the AUC for the detection of a hs-cTnT level higher than 6ng/L was 0.88 using the 12 lead ECG and 0.834 with the single lead. For the detection of hs-cTnT level above of 99 th percentile, the 12 lead ECG AUC was 0.853 and the single lead was 0.806. Conclusion: The AI-ECG permits detection of levels of hs-cTnT consistent with myocardial injury. This may allow a home-based, non-invasive test that would be massively scalable and could further enhance rapid-risk stratification and patient triage with potentially significant cost reductions and enable novel triage strategies at sites without hs-cTn assays.

Background: Clinical factors including previous history of AF, heart failure, hypertension, valvular heart disease, increased age and male gender increase the risk of AF in CLL patients (Shanafelt, Leukemia and Lymphoma 2017). Treatment with Bruton tyrosine kinase inhibitors (BTKi) such as ibrutinib has also been associated with an increased risk of AF in CLL. We evaluated the role of artificial intelligence electrocardiography (AI-ECG) in predicting ibrutinib-induced AF (and, for reference, AF unrelated to ibrutinib) in patients with CLL. Methods: We identified two cohorts of CLL patients using the Mayo Clinic CLL Database. Cohort 1 included patients evaluated within 12 months of CLL diagnosis who did not ever receive ibrutinib. Cohort 2 included patients who were treated with ibrutinib. The electrocardiographic signature of AF in sinus rhythm was detected by an AI-ECG algorithm previously developed using a convolutional neural network (Attia, Lancet 2019). The baseline AI-ECG AF score (positive defined as >0.10 on a scale of 0-1 which offers best balance between sensitivity and specificity per Attia et al.) was computed based on ECGs obtained within 10 years prior to CLL diagnosis (Cohort 1) or 10 years prior to initiation of ibrutinib therapy (Cohort 2). Patients with AF at baseline, missing data, or with ECGs previously used to train the AI algorithm were excluded. Reverse Kaplan Meier diagrams were plotted for both cohorts grouped by AI-ECG positivity. Cox proportional hazards were fitted to assess the predictive ability of AI-ECG in both cohorts. Results: After screening 2,739 patients and applying exclusion criteria (126 patients had baseline AF) a total of 1,149 patients with median 4 (interquartile range [IQR] 2-9) baseline ECGs were included in the analysis (Figure 1A). Cohort 1 included 951 patients with a median follow up of 3.0 (IQR 0.6-7.0) years and positive baseline AI-ECG in 546 (57%) patients. Cohort 2 included 198 patients with a median follow up of 1.6 (IQR 0.7-3.2) years and positive baseline AI-ECG in 91 (46%) patients. In Cohort 1, the median age was 67 years (IQR 58-72), 681 (72%) of patients were men, 68% had low/intermediate risk CLL-International Prognostic Index (IPI), and 32% had high/very high-risk CLL-IPI. In Cohort 2, the median age was 69 years (IQR 62-75), 139 (70%) of patients were men, 13% had low/intermediate risk CLL-IPI, and 87% had high/very high-risk CLL-IPI. AF occurred during follow up in 164 patients (17%) in Cohort 1 and 46 patients (23%) in Cohort 2. In both Cohorts 1 and 2, a positive baseline AI-ECG significantly increased the incidence of AF during follow up (log rank <0.001) (Figure 1B and C). Hazard ratios (for positive vs. negative AI-ECG) were 33.9 (95% confidence interval [CI] 15.0-76.6) for Cohort 1 and 14.8 (95% CI 5.3-41.3) for Cohort 2. Conclusion: The addition of AI to a standard 12-lead ECG obtained during normal sinus rhythm - an inexpensive and ubiquitous test - predicts the occurrence of future AF in patients with CLL. This holds true irrespective of BTKi -based therapy and has important implications for the management of CLL patients. Disclosures Ding: Merck: Membership on an entity's Board of Directors or advisory committees, Research Funding; DTRM: Research Funding; Astra Zeneca: Research Funding; Abbvie: Research Funding; Octapharma: Membership on an entity's Board of Directors or advisory committees; MEI Pharma: Membership on an entity's Board of Directors or advisory committees; alexion: Membership on an entity's Board of Directors or advisory committees; Beigene: Membership on an entity's Board of Directors or advisory committees. Kenderian:BMS: Research Funding; Gilead: Research Funding; Novartis: Patents & Royalties, Research Funding; Mettaforge: Patents & Royalties; Juno: Research Funding; MorphoSys: Research Funding; Lentigen: Research Funding; Sunesis: Research Funding; Tolero: Research Funding; Kite: Research Funding; Humanigen: Consultancy, Patents & Royalties, Research Funding; Torque: Consultancy. Wang:Novartis: Research Funding; Innocare: Research Funding; Incyte: Research Funding. Kay:Juno Theraputics: Membership on an entity's Board of Directors or advisory committees; Oncotracker: Membership on an entity's Board of Directors or advisory committees; Dava Oncology: Membership on an entity's Board of Directors or advisory committees; Rigel: Membership on an entity's Board of Directors or advisory committees; Morpho-sys: Membership on an entity's Board of Directors or advisory committees; Cytomx: Membership on an entity's Board of Directors or advisory committees; Agios Pharma: Membership on an entity's Board of Directors or advisory committees; Astra Zeneca: Membership on an entity's Board of Directors or advisory committees; Sunesis: Research Funding; Abbvie: Research Funding; MEI Pharma: Research Funding; Pharmacyclics: Membership on an entity's Board of Directors or advisory committees, Research Funding; Acerta Pharma: Research Funding; Bristol Meyer Squib: Membership on an entity's Board of Directors or advisory committees, Research Funding; Tolero Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees, Research Funding. Kapoor:Cellectar: Consultancy; Janssen: Research Funding; Sanofi: Consultancy, Research Funding; Amgen: Research Funding; Takeda: Honoraria, Research Funding; Celgene: Honoraria; GlaxoSmithKline: Research Funding. Parikh:MorphoSys: Research Funding; Janssen: Honoraria, Research Funding; AstraZeneca: Honoraria, Research Funding; Merck: Research Funding; AbbVie: Honoraria, Research Funding; Ascentage Pharma: Research Funding; Genentech: Honoraria; Verastem Oncology: Honoraria; GlaxoSmithKline: Honoraria; TG Therapeutics: Research Funding; Pharmacyclics: Honoraria, Research Funding.

The field of cardiac electrophysiology has been on the cutting edge of advanced digital technologies for many years. More recently, medical device development through traditional clinical trials has been supplemented by direct to consumer products with advancement of wearables and health care apps. The rapid growth of innovation along with the mega-data generated has created challenges and opportunities. This review summarizes the regulatory landscape, applications to clinical practice, opportunities for virtual clinical trials, the use of artificial intelligence to streamline and interpret data, and integration into the electronic medical records and medical practice. Preparation of the new generation of physicians, guidance and promotion by professional societies, and advancement of research in the interpretation and application of big data and the impact of digital technologies on health outcomes will help to advance the adoption and the future of digital health care.