Citations

Background Bicuspid aortic valve (BAV) is the most common congenital heart defect, associated with a lifetime morbidity burden exceeding 80% and an estimated 7% probability of transmission to first-degree relatives. Hence, echocardiographic screening of relatives is recommended. The aim of our study is to develop an artificial intelligence-enabled electrocardiogram (AI-ECG) model for the detection of BAV. Methods Between January 1, 1990, and June 30, 2023, the Mayo Clinic healthcare system database was searched for patients aged ≥18 years with echocardiograms performed for family history of BAV, abnormal auscultation, a BAV-related diagnosis, or to rule out the condition, and with an electrocardiogram (ECG) obtained within 6 months. Cases were patients with a confirmed BAV diagnosis, and controls were those with normal aortic valves. ECGs were randomly assigned into training (80%), internal validation (10%), and test (10%) groups. The AI-ECG model was built using Keras under TensorFlow. Results A total of 13,065 valid ECG–echocardiogram pairs were included (mean [SD] age, 58 [17] years; 44% women), with a median time of 1 day (interquartile range, 0–3 days) between studies. With a 45% BAV prevalence in the test group ( n = 1301), the area under the curve was 0.704, positive predictive value (PPV) 60.1%, and negative predictive value (NPV) 69.4%. At a prevalence of 7%, the NPV increased to 96.1%, but the PPV decreased to 12.2%. Conclusions In this retrospective cohort and using the current AI-ECG model, BAV detection showed modest diagnostic performance and limited standalone clinical utility, suggesting that ECG-only screening for BAV may require additional modalities or more refined phenotyping.

Abstract Introduction The relationship between sleep and brain metabolism remains unclear. This study aimed to examine whether overnight sleep features, using a home sleep apnea system based on peripheral arterial tonometry, are associated with regional FDG-PET changes in older adults. Methods We analyzed cognitively unimpaired adults (≥60 years) from the Mayo Clinic Study of Aging (MCSA) cohort who were not receiving sleep apnea treatment. Participants underwent overnight sleep assessment with the WatchPAT system, followed by FDG-PET imaging the next morning. FDG-PET data were co-registered to 3T MRI and automatically segmented into 122 anatomical regions. PET signal was referenced to the pons (SUVR). We computed voxel-weighted median FDG signal across 47 bilateral cortical and subcortical regions of interest (ROIs), excluding brainstem and cerebellum. Associations between sleep architecture and sleep apnea/hypoxemia burden and regional metabolism were assessed using partial Spearman’s rank correlation, adjusting for age. Results Thirty-seven participants were included (70.3% female; median [IQR] age: 72.0 [68.0–78.0] years). Wake after sleep onset (WASO) showed widespread negative associations with FDG uptake in cortical regions (p< 0.05), particularly the ventromedial/ventrolateral (rs=–0.47 to –0.33) prefrontal cortex, lateral temporal (rs=–0.38 to –0.33), and occipital (rs=–0.47 to –0.37) cortices. Total sleep time (TST) was positively associated with medial structures (p< 0.05), including the posterior cingulate (rs=0.39), retrosplenial cortex (rs=0.48), hippocampus/parahippocampal (rs=0.34 to 0.36), calcarine cortex (ρ=0.38), thalamus (rs=0.37), and caudate (rs=0.43). Hypoxic burden was positively associated with medial occipital regions (rs=0.35 to 0.39) and AHI-4% with cuneus (rs=0.33). Regional gray matter volume did not mediate these associations. WASO correlated with Insomnia Severity Index (rs=0.43, p=0.008) and self-reported refreshing sleep (rs=0.046). Conclusion Overnight sleep macrostructure features (WASO and TST) and hypoxic burden are significantly associated with brain metabolism, potentially underlying symptoms of sleep disturbance and cognitive aging. Support (if any) Mayo Clinic Alzheimer’s Disease Research Center Development Award (P30 AG062677), R01 AG056366, U01 AG006786, R01 AG056366, R01 AG034676, R37 AG011378, R01 AG041851, R01 NS097495, R01 AG068206, P30 AG062677, R01 HL065176. CLARA project work with funding from the European Union’s HORIZON EUROPE research and innovation programme (No 101136607).

Abstract Aims Rheumatic heart disease (RHD) remains the most prevalent acquired cardiovascular disease among young adults in developing countries, and there is a critical need for an efficient, affordable, and available screening strategy to detect RHD in these resource-limited areas. The 12-lead electrocardiogram (ECG) is a ubiquitous, inexpensive diagnostic tool, and application of artificial intelligence to the ECG (AI-ECG) has allowed for screening of various cardiovascular pathologies. Methods and results Patients aged >18 years with severe RHD with at least one transthoracic echocardiogram and one standard 12-lead ECG performed at Groote Schuur Hospital, Cape Town, within 180 days of each other were eligible. A total of 920 RHD patients were identified, with 2199 corresponding controls from the same medical centre. Patients were allocated to derivation (80%), validation (10%), and testing (10%) groups, respectively. The AI-ECG to detect RHD from a standard 12-lead ECG was built using the Keras framework under Tensorflow backend. In the test group, the AI-ECG for RHD exhibited strong algorithm performance with an AUC of 0.890 (95% Confidence Interval 0.848–0.940) and an accuracy of 83.8%. The algorithm remained robust between RHD presentation: mitral valve alone (AUC 0.913) and mixed mitral and aortic valve disease (AUC 0.858). There were minor performance differences in the subgroup analysis of patient sex, age, and ECG characteristics. Conclusion In this study, an AI-ECG algorithm was successfully developed to screen for advanced RHD from a standard 12-lead ECG, and the AI-ECG performance remained acceptable across most subgroups, including RHD manifestation, patient sex, age <70 years, and ECG characteristics.

INTRODUCTION: Building on prior results, we hypothesized that an electrocardiogram (ECG)-enabled machine learning (ML) model could be used to detect advanced chronic liver disease (CLD). METHODS: A cohort with CLD and 12-lead ECGs was matched with controls from electronic health records. A ML model was trained as a binary classifier. RESULTS: There are 12,930 patients with CLD and 64,577 controls in the cohort. The model's discriminative ability to classify CLD showed an area under the receiver-operating characteristic curve 0.858 (95% confidence interval: 0.850–0.866), and at the chosen threshold, CLD ECGs had 12 times higher odds of being classified as CLD (diagnostic odds ratio 12.33, 95% confidence interval: 11.16–13.63). DISCUSSION: An ECG-enabled ML model affords great promise in identifying advanced CLD in low resource areas.

Background: Risk stratification for suspected pulmonary emboli (PE) remains an area of active research. Current risk stratification algorithms have moderate discriminatory capabilities and are guideline-recommended, but they are poorly utilized in practice, with most clinicians relying predominantly on their clinical training and personal “gestalt.” Our group recently derived an Artificial Intelligence (AI) algorithm using electrocardiograms (ECG) to predict the presence of PE (AUC 0.69) on computed tomography scans of the pulmonary arteries (CTPA) in a cohort of 79K patients. Aim: Validate the AI-ECG algorithm in a contemporary dataset of patients with suspected PE. Methods: Emergency Department (ED) visits between 9/1/2022 and 12/31/2023 across the Mayo Clinic Enterprise were searched using a unified data platform. ED visits in patients over 18 years of age with a discretely documented Wells' PE or PERC Rule, a D-dimer result, or a CTPA were included. Only CTPA imaging and D-dimers performed within 48 hours of ED admission were eligible to be included. Patients were included in the final cohort if an ECG was performed within 6 hours of ED admission, D-dimer testing, or CTPA. Patients were considered negative for PE if CTPA was negative, and when this wasn't performed, a negative D-dimer (high sensitivity) or PERC rule of 0 was considered negative. A previously derived Artificial Neural Network algorithm to analyze the ECG to predict PE was applied to the dataset for validation. Results: A total of 27,028 ED visits in 24,288 unique patients meet inclusion criteria. Among this group, the first ED visit with a CTPA and an ECG within 6 hours was included, leaving 18,936 patients for the final validation cohort. The mean age of the cohort was 54.7 (SD 18.7), and 57% were male. Among the entire cohort, 13,573 underwent D-dimer testing for which 4,895 (36%) were positive at standard thresholds. A total of 10,486 patients underwent CTPA (55% of patients evaluated), of which 455 were positive for PE (4.3%). Using the AI-ECG model from the derivation cohort applied to this validation cohort resulted in an AUC of 0.69, identical to the derivation cohort. Using ECG model estimates alone, low (0.74%, n=2447, NPV 99.3%), elevated (2.27%, n=15404, NPV, 97.72%), and high-risk (8.02%, n=1085, NPV 91.98%) categories for PE can be created. A negative D-dimer in the elevated risk group increased the negative predictive value (NPV) to >99.9% and could be used to reduce imaging in this group further. Using the proposed ECG to D-dimer cascade risk stratification could further reduce the need for PE imaging for an overall NPV of 99.9% and sensitivity of 95.2%. We then used additional clinical parameters to train a new algorithm with additional clinical data and the AI-ECG algorithm estimates. Multiple machine learning algorithms were then evaluated in a dataset divided into training (2/3) and testing (1/3) datasets, after downsampling (to address class imbalance). The best results were obtained with AI-ECG results and D-dimer: Ada Boost (AUC 0.93, F1 score 0.895) compared to standard logistic regression (AUC 0.79, F1 score 0.746). The addition of age, sex, and oxygen saturation did not improve model performance. Conclusion: We have now validated the previously derived AI-ANN ECG algorithm for PE in a contemporary and independent cohort and demonstrate potential for integration into clinical PE assessment algorithms. Unlike prior risk stratification processes, ECGs are invariably performed in patients with cardiopulmonary symptoms and have the potential to prevent missed or delayed PE diagnosis by proactively identifying high-risk patients. Additionally, we demonstrate that the results of this algorithm can be combined with D-dimer values, resulting in a simple, powerful, predictive algorithm (AUC 0.93) that pairs cardiovascular electrophysiology with laboratory biomarkers and could lead to a new paradigm in clinical risk stratification for PE.

Background: Many medications, electrolyte perturbations, and several diseases can prolong the QTc beyond its 99 th percentile value resulting in acquired QT prolongation (AQTP). In contrast, approximately 1 in 2,000 people have congenital long QT syndrome (LQTS) hallmarked by pathological QT prolongation. Recently, we developed an AI-ECG model to distinguish these two groups, however this model does not elucidate which ECG segments distinguish LQTS from AQTP. Methods: Using the cohort from our previous study, all patients with LQTS evaluated in the Mayo Clinic Genetic Heart Rhythm Clinic and controls were included. Next, every patient/control with ≥1 ECG above age- and sex-specific 99 th percentile for QTc was selected and matched at 1:100 ratio. Subsequently, 5 ML algorithms (logistic regression, Naïve Bayes, random forest, Gradient boosting and XGBoost) were trained using 52 features representing segments of the 12-lead ECG (e.g. wave peak times, amplitudes, areas under wave, duration; 52x12 features). Training and testing sets were spilt at 80:20 ratio and 5-fold cross validation was performed to prevent overfitting. Following analysis, the top 100 features were re-analyzed to refine results. Results: Among >1,600 patients with LQTS, 690 had ≥ 1 ECG with a QTc above the established threshold compared to 28,186 controls. Following age- and sex-matching and splitting, 22,410 (training), and 5,776 (testing) ECGs were used. Of the 5 ML algorithms, XGBoost demonstrated best performance (AUC 0.913, accuracy 0.836, sensitivity 0.857, specificity 0.835) to distinguish LQTS from AQTP. Following filtering for the top 100 features, performance of the algorithm remained high with AUC of 0.912 (accuracy 0.84, sensitivity 0.841, specificity 0.84). The features with the greatest impact on classification output included time from T wave onset to T wave peak (leads V1, V4, V4), QRS interval duration (leads V1, V3, V4, aVR), and amplitude at end of ST segment (aVF). Conclusions: Distinguishing monogenetically driven LQTS from multi-factorial AQTP is critical in clinical practice for proper management, mitigation of offending QT-prolonging factors, and treatment. Building on our previously developed AI-ECG-LQTS ‘mutation detector’, this model identifies some of the elements within the ‘black box’ that distinguish an ECG and QTc stemming from a patient with LQTS compared to a similar prolonged QTc value that is arising due to non-genetic, acquired QT-prolonging factors.

Abstract Background Existing clinical preoperative risk assessment tools such as the Revised Cardiac Risk Index (RCRI) have been used extensively to predict perioperative major adverse cardiovascular events (MACE). Electrocardiogram (ECG) analysis using Deep Learning can identify hidden features that human eyes cannot identify and might help further risk stratify patients undergoing noncardiac surgery. Objective We aimed to develop a Deep Learning model that predicts perioperative MACE in patients undergoing elective noncardiac surgeries. Methods We included patients ≥ 18-year-old who underwent elective non-cardiac surgery between 2000-2021. MACE was validated in duplicate and were defined as a composite of in-hospital myocardial infarction, cardiac arrest or mortality. A convolutional neural network was developed using ECGs within 30 days before surgery. We randomly split subjects into training, internal validation, and testing datasets in a 7:1:2 ratio. Performance was evaluated using the area under the receiver operator characteristics curve (AUC) values on the testing dataset. Models trained included AI-ECG models to predict in-hospital MACE and to predict in-hospital mortality, and AI-ECG+RCRI models to predict in-hospital MACE and to predict in-hospital mortality. All models were validated using 30-day validated outcomes (MACE or all-cause mortality) in a community-based cohort and were compared with the Revised Cardiac Risk Index (RCRI) score using AUC values. Findings: We included 195,214 patients who underwent a total of 241,999 surgeries. 169,214 ECGs were used in the training dataset, 24,234 in the internal validation dataset, and 48,551 in the testing dataset. In the test cohort, the AI-ECG algorithm discriminated risk for in-hospital MACE with an AUC of 0.82 (95% CI 0.79-0.85). The algorithm similarly discriminated 30-day MACE with an AUC value of 0.79 (95% CI 0.77-0.81) surpassing the discrimination of the RCRI score AUC of 0.69 (95% CI 0.66-0.72). All models’ performances are shown in Figure1A. The AI-ECG model to predict in-hospital mortality had an AUC of 0.83 (95% CI 0.80-0.86) and predicted 30-day mortality with an AUC of 0.80 (95% CI 0.77-0.82). Other models’ performances are shown Figure1B. Training the AI-ECG models including RCRI data did not improve the AI-ECG models’ performance. Conclusion AI-ECG models with no additional clinical data can improve perioperative risk prediction in non-cardiac surgery, outperforming a conventional and widely used risk stratification tool such as the RCRI.

AbstractBackgroundThe clinical relevance of cardiac troponin (cTn) elevation in takotsubo syndrome (TTS) remains uncertain. The present study sought to investigate the role of cardiac troponin (cTn) elevations in mortality prediction of patients with Takotsubo syndrome (TTS).MethodsPatients enrolled in the International Takotsubo (InterTAK) Registry from January 2011 to February 2020 with available data on peak cTn levels were included in the analysis. Peak cTn levels during the index hospitalization were used to define clinically relevant myocardial injury. The threshold at which clinically relevant myocardial injury drives mortality at 1 year was identified using restricted cubic spline analysis.ResultsOut of 2′938 patients, 222 (7.6%) patients died during 1‐year follow‐up. A more than 28.8‐fold increase of cTn above the upper reference limit was identified as threshold for clinically relevant myocardial injury. The presence of clinically relevant myocardial injury was significantly associated with an increased risk of mortality at 5 years (adjusted HR 1.58, 95% CI 1.18–2.12, p =.002). Clinically relevant myocardial injury was related to an increased 5‐year mortality risk in patients with apical TTS (adjusted HR 1.57, 95% CI 1.21–2.03, p =.001), in presence of physical stressors (adjusted HR 1.60, 95% CI 1.22–2.11, p =.001), and in absence of emotional stressors (adjusted HR 1.49, 95% CI, 1.17–1.89, p =.001).ConclusionThis study for the first time determined a troponin threshold for the identification of TTS patients at excess risk of mortality. These findings advance risk stratification in TTS and assist in identifying patients in need for close monitoring and follow‐up.

Key Points Measuring blood potassium has always required access to blood. The surface electrocardiogram, analyzed using an artificial intelligence algorithm, can detect hyperkalemia bloodlessly. The artificial intelligence-analyzed electrocardiogram exhibited a high negative predictive value but substantially lower positive predictive value. Background Artificial intelligence (AI) electrocardiogram (ECG) analysis can enable detection of hyperkalemia. In this validation, we assessed the algorithm's performance in two high acuity settings. Methods An emergency department (ED) cohort (February to August 2021) and a mixed intensive care unit (ICU) cohort (August 2017 to February 2018) were identified and analyzed separately. For each group, pairs of laboratory-collected potassium and 12 lead ECGs obtained within 4 hours of each other were identified. The previously developed AI ECG algorithm was subsequently applied to leads 1 and 2 of the 12 lead ECGs to screen for hyperkalemia (potassium >6.0 mEq/L). Results The ED cohort ( N =40,128) had a mean age of 60 years, 48% were male, and 1% ( N =351) had hyperkalemia. The area under the curve (AUC) of the AI-enhanced ECG (AI-ECG) to detect hyperkalemia was 0.88, with sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and positive likelihood ratio (LR+) of 80%, 80%, 3%, 99.8%, and 4.0, respectively, in the ED cohort. Low-eGFR (<30 ml/min) subanalysis yielded AUC, sensitivity, specificity, PPV, NPV, and LR+ of 0.83, 86%, 60%, 15%, 98%, and 2.2, respectively, in the ED cohort. The ICU cohort ( N =2636) had a mean age of 65 years, 60% were male, and 3% ( N =87) had hyperkalemia. The AUC for the AI-ECG was 0.88 and yielded sensitivity, specificity, PPV, NPV, and LR+ of 82%, 82%, 14%, 99%, and 4.6, respectively in the ICU cohort. Low-eGFR subanalysis yielded AUC, sensitivity, specificity, PPV, NPV, and LR+ of 0.85, 88%, 67%, 29%, 97%, and 2.7, respectively in the ICU cohort. Conclusions The AI-ECG algorithm demonstrated a high NPV, suggesting that it is useful for ruling out hyperkalemia, but a low PPV, suggesting that it is insufficient for treating hyperkalemia.