Medina-Inojosa, Betsy

Abstract Introduction Atrial fibrillation (AF) is a major contributor to stroke risk, yet many cases remain undiagnosed due to its silent nature. Deep learning-based electrocardiogram (DL-ECG) analysis can streamline AF risk stratification, providing a more reliable and accessible method to stratify risk using only a standard resting ECG. However, this tool has been validated only on AF detected clinically or with short ECG event monitors. Purpose This study aimed to validate the predictive accuracy of a DL-ECG model for identifying AF in normal sinus rhythm (NSR) ECGs among patients undergoing continuous cardiac monitoring with implantable loop recorders (ILRs). Methods We conducted a historical cohort study across three major U.S. academic medical centers, enrolling patients with an ILR and no prior diagnosis or documented ECG evidence of AF. Pre-implantation 10-second, 12-lead ECGs in NSR were analyzed using the DL-ECG model. Performance was evaluated using the area under the receiver operating characteristic curve (AUC) and compared with the CHARGE-AF score. Kaplan-Meier survival analysis and Cox proportional hazards models assessed the association between DL-ECG risk categories and incident AF, defined as any episode lasting ≥6 minutes. Results We included 2,523 patients (mean age 66 ± 17 years, 52% male). AF was detected in 658 (26%) over a median follow-up of 19 months (IQR 8-37), Table 1. The DL-ECG model classified 1,634 patients (65%) as at low-risk and 889 (35%) as at high-risk. AF developed in 242 (14.8%) of the low-risk group and 416 (46.8%) of the high-risk group, yielding an AUC of 0.73 (95% CI 0.70-0.75), surpassing the CHARGE-AF AUC of 0.69 (95% CI 0.66-0.71). A combined DL-ECG and CHARGE-AF model yielded an AUC of 0.74 (95% CI 0.72-0.76), Figure 1 presents the AUC, Net Reclassification Index, and DeLong test comparisons among the three models. High-risk patients had a significantly higher incidence of AF, which remained significant after adjusting for CHARGE-AF variables, Figure 2. AF incidence increased as DL-ECG quartiles increased, with 61% of patients at the top quartile developing AF during follow-up, Figure 3. False negatives (patients initially classified as low-risk but in whom later AF was detected) had a longer median time to AF detection (127 days [IQR 28-425] vs. 101 days [IQR 20-318] for true positives, p = 0.01), and where found to have at least one positive ECG during follow-up in 36%, compared to 32% of true negatives, indicating eventual risk reclassification. Conclusions Among patients undergoing ILR monitoring, the DL-ECG model demonstrated superior predictive performance compared to CHARGE-AF, offering a promising and accessible alternative for early detection of silent AF. This approach could aid early AF detection, particularly in resource-limited settings where access to specialized rhythm monitoring services is limited.Table 1 and Figure 1 Figure 2 and 3

Background: The gap between artificial intelligence-derived age (ECG-Age) and chronological age has been validated as a strong predictor of increased cardiovascular morbidity and mortality, independent of age, and is a potential biomarker of biological aging. Acute stress may cause real-time, transient changes in autonomic function via electrophysiological signals that, when repeated over time, may contribute to biological aging and represent mechanisms linking stress and Major Adverse Cardiovascular Events (MACE). We hypothesized that acute mental stress induced in a laboratory is associated with increased ECG and that this increase is associated with a higher risk of MACE among individuals with stable coronary heart disease (CHD). Methods: We pooled two prospective cohorts of participants with stable CHD from a university-based hospital network in Atlanta. Stress was induced using a standardized mental stress test (public speaking task) and measured 12-lead ECGs during rest, mental stress, and recovery periods. To measure ECG-age, we utilized a published algorithm using a neural network on >500,000 independent patients. The study endpoint was MACE including nonfatal myocardial infarction, hospitalization for heart failure, or cardiovascular death. The difference in ECG-Age between resting and acute mental stress was used to define the stress Age-Gap. Cox proportional hazard regression modeled the association of stress Age-Gap operationalized as tertiles adjusting for chronological age, sex, race, cardiovascular risk factors, and baseline Cohen perceived stress scale. Results: 442 participants in the total sample pool (38% women). Over a 5-year median follow-up, 95 participants (21%) experienced MACE. At rest, the mean chronologic age was 55.7±9.9 years, and ECG-Age was 57.1 ±10.7 years, with rest Age-Gap of 0. 9±9.6 (p=0.16). During stress, the stress Age-Gap of 1.95±9.8 (p<0.0001), while recovery Age-Gap was 1.1±10.3 (p=0.13). Those that demonstrated the largest ECG-Age positive change between rest and stress by ECG (top tertile) had a higher risk for MACE when compared to those with the lowest or negative ECG age change (bottom tertile), see Figure. Conclusion: Among patients with stable CHD, acute mental stress caused changes in biological age as measured by AI-ECG consistent with accelerated biological aging. Such changes were associated with MACE. These results uncover accelerated biological aging as a mechanism linking acute stress to CVD outcomes.

Methods: From a sample cohort of 274,539 consecutive adults who sought care at Mayo Clinic between 2018-2022, we ascertained demographic and clinical information, including clinically indicated TTE within 15 days and up to a year after the index clinical visit. The Area Deprivation Index (ADI), a tool that captures neighborhood-level disadvantages that are linked to cardiovascular morbidity and outcomes, was extracted from the Neighborhood Atlas and ranked accordingly as state percentiles stratified in quintiles, with the highest values representing “most disadvantaged” areas. We performed non-parametric multivariate regression models that adjusted for cardiovascular risk factors expressed in odds ratio (OR) and 95% confidence interval (CI). Results: A total of 45,823 (16.9%) subjects had a TTE performed, with a mean age of 63±16 years, 48% were women, and 86% were non-Hispanic white. According to the TTE, 4% of subjects had LVSD, while 11% had a high probability of LVSD by AI-ECG. Median state ADI was 4 IQR (2-7). In the whole cohort, subjects residing in the most disadvantaged areas were less likely to have a TTE [OR 0.94 CI (0.91- 0.97)] see Fig. 1-A ; conversely, they were more likely to have LVSD by TTE [OR 1.80 CI (1.52-2.13)] see Fig. 1-B , and higher probability of LVSD by AI-ECG [OR 1.43 CI (1.36-1.51)] when compared to those with the least disadvantage see Fig. 1-C , independent of age, sex, and cardiovascular risk factors, p for trend<0.001. Conclusion: Individuals living in socially disadvantaged areas were slightly less likely to have an echocardiogram despite a significantly higher likelihood of having an LVSD (suspected by AI-ECG and confirmed by TTE), independent of cardiovascular risk factors. Highlighting an important disparity in the diagnosis of LVSD.

Background: We previously developed deep-learning algorithms that identify coronary artery disease (CAD) risk based on (i) coronary artery calcium (CAC), (ii) obstructive CAD by angiography, and (iii) left ventricular akinesis in ≥1 segment by echocardiogram, using a 12-lead electrocardiogram (CAD ECG-AI). We tested the hypothesis that those with increased probability of CAD by CAD ECG-AI algorithms will have an increased risk of atherosclerotic cardiovascular disease (ASCVD) events and would refine the AHA Predicting Risk of Cardiovascular Disease Events (PREVENT TM ) equations’ predictive capabilities. Methods: We assessed a group of consecutive patients who sought primary care in Olmsted County, MN, between 1997 and 2003. Passive follow-up was conducted using the Rochester Epidemiology Project's record linkage system. Patients included met the same criteria of the original PREVENT TM equations. The probability output of each CAD ECG-AI algorithm was used to predict ASCVD (fatal and non-fatal myocardial infarction and ischemic stroke) and ASCVD-Plus [further including percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG), and all-cause mortality]. Events were validated in duplicate. Cox proportional hazard models adjusted for age, sex, and risk factors while modeling each CAD ECG-AI algorithm as an individual and additive predictor (see footnote). PREVENT TM risk categories stratified analysis to evaluate the effect of the CAD ECG-AI models on predicted ASCVD risk. Results: We included 21,193 patients, with mean ± SD age 51.6 ± 12.0, 54% women and 95% white. After 14.2±12.4 years follow-up, 1,351 (6.4%) developed ASCVD, and 2,808 (13.2%) had ASCVD-Plus. The risk of ASCVD and ASCVD-Plus increased with positivity for the CAD ECG-AI algorithms after adjustments for age, sex, and PREVENT TM equations factors, all p for trend <0.001 (data not shown). In additive analyses, the risk of ASCVD and ASCVD-Plus increased with increased CAD ECG-AI factors, all p for trend <0.001 ( Fig. A-B ). Furthermore, the CAD ECG-AI algorithm enhanced the predictive capabilities of PREVENT TM equations across most risk subgroups ( Fig. C-E ). Conclusions: The deep-learning CAD ECG-AI algorithms displayed an independent and additive association with long-term ASCVD and ASCVD-Plus events in the community and improved the PREVENT TM predicted ASCVD risk. The CAD ECG-AI algorithms could help identify individuals at risk in primary prevention of cardiovascular events.

Background: Sodium-glucose cotransporter-2 inhibitors (SGLT-2i) and glucagon-like peptide-1 receptor agonists (GLP1-RA) have known cardiovascular benefits. We aimed to assess the efficacy of GLP1-RA or SGLT2i initiation in reducing mortality or cardiovascular events in a community-based cohort of adults with ischemic stroke. Methods: Patients ≥18 years old admitted for acute ischemic stroke between Jan 2000 and June 2022 were identified using the Rochester Epidemiology Project. Subjects were considered exposed to SGLT-2i or GLP1-RA if they were treated with either medication after the index stroke and before the primary (all-cause mortality) or secondary outcomes (MI, recurrent ischemic stroke or composite). Kaplan-Meier curves and Cox models are shown in the Figures . Results: We included 7044 subjects, mean age 72±14 yr, 52% male, 94% white, over a median follow-up of 3 [1-7] years, 416 subjects (6%) had a second stroke, 426 (6%) had an incident MI and 3738 (53%) died. Those on SGLT2i/GLP1-RA had a decreased risk for mortality, incident MI, and composite in the univariate analysis. These associations remained significant after multivariable adjustment and were not affected by minimum time of exposure Figure 1&2 . SGLT2i/GLP1-RA use was associated with recurrent ischemic stroke in the multivariate analysis Figure 2C . All associations remained significant in most sub-analyses of individual medications except for the association between GLP1-RA and recurrent stroke Figure 3 Conclusions: SGLT2i or GLP-1 RA use was associated with decreased mortality and risk of myocardial infarction and recurrent stroke in patients with ischemic stroke.

Background: Our team recently developed an Artificial Intelligence model that enables the identification of increased left ventricular filling pressures using a single-lead electrocardiogram (AI-ECG) with excellent performance (AUC 0.91). However, its ability to predict incident heart failure (HF) in the community is yet to be evaluated. Hypothesis: We tested the hypothesis that the output from the AI-ECG will predict the long-term incidence of heart failure (HF) and HF with preserved ejection fraction (HFpEF), and would refine the HF predictive capabilities of the AHA Predicting Risk of Cardiovascular Disease EVENTs (PREVENT TM ) Equations. Methods: Leveraging the resources of the Rochester Epidemiology Project (REP) we analyzed data from patients who sought primary care in Olmsted County, MN, between 1997-2003 with passive follow-up until April 2024. We included those with a digital ECG at baseline and excluded those who were part of the original algorithm development or history of heart failure. The diagnosis of heart failure was ascertained by using ICD codes (ICD9-428,ICD10-I.50) and a word search for the terms related to heart failure. HFpEF was ascertained using the Ejection Fraction (EF) data closest to the heart failure diagnosis date, using EF more than 40% as the criterion. We tested the association between AI-ECG predicted probabilities and new-onset HF and HFpEF using Cox proportional hazard models. These models were also adjusted for cardiovascular risk factors and were stratified to evaluate the effect of the AI-ECG output on PREVENT TM -predicted HF at 10 years. Results: We included 20,276 subjects, mean age 55±14.03, 52% women, 92% non-Hispanic white. After 18.2±8.5 years of follow-up, 4,008 (19.76%) were diagnosed with HF and 3,812 (18.8%) HFpEF. Risks of HF and HFpEF significantly increased in patients with increased left ventricular filling pressures by AI-ECG, independent of risk factors, all p for trend <0.001, see Fig A-B , respectively. AI-ECG enhanced the prediction capabilities of the PREVENT TM score across all risk groups, see Fig-C. Conclusions: AI-ECG-Derived increased left ventricular filling pressure is associated with new-onset HF and HFpEF in the community and was associated with incidental heart failure at any PREVENT TM risk category. Those results suggest a potential role of AI ECG in predicting heart failure.

Abstract Background Cardiovascular disease remains the leading cause of death and burden globally, warranting the need to explore novel predictors of major adverse cardiovascular events (MACE) using new technologies. Artificial Intelligence (AI)-enhanced ECG (AI-ECG) algorithms can provide information on cardiovascular health of individuals, independent of conventional risk factors. Purpose The current study aimed to understand whether a previously described AI-ECG sex estimation algorithm could be applied for risk stratification of major adverse cardiovascular events (MACE), with a focus on a population without previous history of MACE. Methods Patients having 12-lead ECGs between 2018-2019 were included and followed up until August 2023. Exclusion criterion was a prior history of MACE. AI-ECG sex estimation was determined using a previously described algorithm predicting sex based on 12-lead ECGs. The algorithm generates outputs ranging from 0 to 1, representing the estimated probability of being male. Outputs above 0.5 were considered as AI-ECG-estimated male, and below as AI-ECG-estimated female. This was compared with biological sex to determine whether the two matched. Inconsistency between AI-ECG estimated sex and biological sex was defined as discordance. MACE was defined as myocardial infarction, coronary revascularization, non-fatal stroke, and all-cause death. Results Among 80,578 subjects (mean age 59.1 ± 16.1 years, 52.4% female), 9.7% exhibited sex discordance in AI-ECG estimation. Discordance was associated with a 56% higher risk of MACE (HR, 1.56, 95% CI 1.46-1.65, p &lt;0.001), and remained significant in multivariable analysis with traditional risk factors (HR 1.35, 95% CI 1.27-1.43, p &lt;0.001) (Figure 1). Sex-specific analysis showed increased MACE risk in both discordant males and females (Figure 2). Conclusion In patients with no history of MACE, discordance in AI-ECG sex estimation is associated with increased risk of MACE. These findings suggest the potential of AI-ECG sex estimation algorithm for risk assessment, especially in patients with no history of MACE.Figure 1.Cox proportional hazard modelFigure 2.Kaplan-Meier of MACE by sex

Abstract Background The concept of ideal cardiovascular health (CVH) was developed by the American Heart Association (AHA) as a composite of seven health factors and behaviors labeled as the Life’s Simple 7 (LS7). Ideal CVH has been associated with complex measures of cardiovascular aging. We hypothesized that physiologic aging, as determined by a simple, routinely used tool such as the electrocardiogram (ECG) enabled by an artificial intelligence algorithm to predict an individual’s physiologic age (ECG age), will be associated with poor CVH as defined by an adapted version of the AHA LS7. Methods We conducted an observational retrospective study of consecutive patients aged ≥18 years who sought care between June 2019 and March 2022, self-reported behavioral history through a validated intake questionnaire in clinical outpatient visits, had a clinically indicated ECG within one year, and provided research authorization. We applied a previously developed AI-ECG convolutional neural network to derive the ECG-Age. We excluded patients with missing data, those with body mass index (BMI) &lt;18.5 kg/m2, previous history of left-bundle branch block, atherosclerotic cardiovascular disease events, atrial fibrillation, heart failure, and non-skin cancer. An adapted version of the LS7 was used to define CVH, which includes anthropometrics (e.g., BMI), behavioral history (e.g., smoking, physical activity, and diet), vital signs (e.g., blood pressure), and laboratory values (e.g., total cholesterol, and fasting glucose). CVH was stratified as ideal, intermediate, and poor if the LS7 score had values of 9-14, 7-8, and 0-6, respectively, constructed by adding 2,1 or 0 points to each metric that met each category criteria, respectively. Accelerated aging was defined as having a Delta-Age ≥1SD above the mean, whereas Delta-age was defined as ECG minus chronological age. Logistic regression models tested the association between CVH categories and positive Delta-Age while adjusting for chronological age and sex. Results We included 114,214 individuals (53.3% females, 86.9% Caucasian). Mean chronological age was 57.7±14 years and mean AI-ECG- Age was 58 ± 12.44 years (R2 0.80, p-value &lt;0.001). Mean Delta-Age was 0.28 ±8.8 years. LS7 average score was 8.3 ± 2.23; 46.8% had ideal, 31.5% intermediate, and 21.7% had poor CVH (Figure A). Compared to those with ideal CVH, those with poor CVH were more than 50% more likely to have Delta-Age ≥1SD (OR 1.61, CI 95% 1.54-1.69, p-value: &lt;0.001), independent of chronological age and sex. (Figure B). Conclusion Poor CVH, defined by the AHA LS7, is linked with biological aging as determined by AI-enabled ECG, highlighting the importance of promoting CVH in adults as a strategy to decelerate biological aging and reduce its associated mortality.

Background: Our team has previously developed convolutional neural networks (CNNs) to estimate age and sex from a 10-second, 12-lead ECG as indicators of patient wellness. Here, we develop an additional wellness network to estimate body mass index (BMI) from the ECG input signal. Aims: To evaluate the performance of neural networks trained to classify and directly estimate BMI. Methods: We identified all adult patients in the Mayo Clinic system who had a 10-second, 12-lead ECG within 15 days of a BMI measurement. Of the&nbsp;174,863 patients in the cohort, 2,895 were underweight (BMI < 18.5), 42,387 were of normal weight (BMI 18.5-24.9), and 57,939 were overweight only (BMI 25.0-29.9). Among patients categorized as obese, 39,788 had class I obesity (BMI 30.0-34.9), 18,978 had class II obesity (BMI 35.0-39.9), and 12,876 had class III obesity (BMI ≥ 40.0). The cohort was split into training, validation, and testing datasets in an 8:1:1 ratio to develop networks for BMI classification and a direct&nbsp;estimation of BMI. Results: In detecting overweight in the testing set, the model achieved an AUC of 0.86 (95% CI 0.85, 0.87). In detecting class I, class II, and class III obesity, the model yielded AUCs of 0.84 (95% CI 0.83, 0.85), 0.84 (95% CI 0.83, 0.85), and 0.85 (95% CI 0.85, 0.86), respectively. The model to estimate BMI achieved a mean absolute error of 3.70. Conclusions: AI-enabled ECGs demonstrate efficient detection of overweight patients and all classes of obesity. Further research is needed to determine if discrepancies between the traditionally measured BMI and the AI-ECG BMI reflect the presence or lack of metabolic abnormalities.

Introduction: Artificial Intelligence-enabled electrocardiogram (AI-ECG) can estimate the probability of having atrial fibrillation (AFib) while in sinus rhythm. Recurrent stroke in subjects without known AFib may be due to undiagnosed AFib. We evaluated the association between AI-ECG AFib and recurrent ischemic stroke and with incident AFib Methods: Consecutive adults admitted to a Mayo Clinic or Mayo Clinic Health System hospital or to Olmsted Community Hospital for ischemic stroke between January,1,2000 and June,30,2022 were identified using the Rochester Epidemiology Project. Recurrent ischemic strokes were identified after reviewing medical records and imaging studies. We excluded those with known AFib or without ECGs within 3 years prior to the first ischemic stroke. The AI-ECG network was previously developed using 649,931 ECGs from > 180,000 subjects. The primary endpoint was recurrent ischemic stroke; secondary endpoint was incident AFib. Cox model adjusted for age, sex, and CHA2DS2-VASc2 score variables. Results: We included 4,816 subjects, mean age 70±15 yr, 54% male, 92% white, median CHA2DS2-VASc2 score 4 [3-5], median follow-up 3 [1-7] yr. 280 subjects (6%) had a second stroke and 366 (8%) were diagnosed with incident AFib over follow-up. Those with a positive AI-ECG AFib had an increased risk for recurrent ischemic stroke(HR 1.54, 95% CI 1.21-1.96, P<0.001 Figure A) This association remained significant after adjustment for CHA2DS2-VASc2 score(HR 1.37 95% CI 1.07-1.76, P=0.01). These associations were not affected by anticoagulation use over follow-up. AI-ECG AFib was also associated with incident AFib(HR 2.42 95% CI 1.96-2.98, P<0.001 Figure B) Conclusions: AI-ECG AFib predicts recurrent ischemic stroke in the community, possibly reflecting undiagnosed AFib that can be detected by AI-ECG, as it also predicts incident AFib. These results underscore the potential utility of AI-ECG for risk stratification of patients with ischemic stroke.