Klempfner, Robert

Abstract Background Mitral regurgitation (MR) and tricuspid regurgitation (TR) are prevalent valvular heart diseases associated with significant morbidity and mortality. Traditional echocardiography faces limitations in availability, cost, consistency, and reliability, leading to misdiagnosis and undertreatment. The application of artificial intelligence (AI) to echocardiographic scans has the potential to address these challenges. Methods This study evaluates the performance of an AI algorithm on an external population. The algorithm utilizes deep learning networks to analyze echocardiographic exams for diagnosing atrioventricular valve disorders. We tested the algorithm on transthoracic echocardiography data collected from a single center between 2013 and 2023. The model's performance was compared to ground truth values using two classification schemes: distinguishing between normal-mild and moderate-severe regurgitation, and categorizing results into four groups: normal, mild, moderate, and severe. Results The MR cohort included 280 patients, while the TR cohort comprised 298 patients. The model demonstrated a robust ability to identify clinically significant (moderate and above) atrioventricular valve regurgitation. The MR model achieved an area under the curve (AUC) of 0.98 (95% CI: 0.97–0.99), with 91% accuracy, 95% sensitivity, and 89% specificity. In comparison, the TR model exhibited an AUC of 0.96 (95% CI: 0.94–0.98), with 84% accuracy, 91% sensitivity, and 80% specificity. Conclusion The model demonstrated high diagnostic accuracy and reliability in assessing atrioventricular valve regurgitation severity, highlighting its potential as a valuable clinical tool. The findings underscore the role of AI in complementing expert evaluations and improving access to effective diagnostics, with future applications potentially including point-of-care diagnosis and monitoring of disease progression.

Abstract Aims Mitral and tricuspid regurgitation (MR and TR) are common in older adults and associated with substantial morbidity and mortality. While transthoracic echocardiography (TTE) is the diagnostic gold standard, access remains limited in many care settings. Artificial intelligence (AI)–based echocardiographic analysis may help address this diagnostic gap. Methods and results We externally validated a deep learning algorithm developed by Aisap.ai using TTE studies from the Mayo Clinic Health System (2013–23). The model analyses echocardiographic images to classify atrioventricular regurgitation severity and was evaluated against cardiologist interpretations. Performance was assessed using binary (normal–mild vs. moderate–severe) and ordinal (normal, mild, moderate, severe) classification schemes. Among 1541 eligible TTEs, the model returned predictions for 578 studies (38%). Performance analysis was limited to these cases. The MR cohort included 280 studies and the TR cohort 298. For MR, the model achieved an area under the receiver operating characteristic curve (AUC) of 0.98 [95% confidence interval (CI): 0.97–0.99], with 91% accuracy, 95% sensitivity, and 89% specificity. For TR, the AUC was 0.96 (95% CI: 0.94–0.98), with 84% accuracy, 91% sensitivity, and 80% specificity. Conclusion In cases where a prediction was generated, the model demonstrated high diagnostic performance in identifying clinically significant atrioventricular regurgitation. These findings support the feasibility of AI-assisted echocardiography in diverse populations, while underscoring the need for technical alignment between model requirements and local acquisition practices to ensure real-world applicability.

AbstractAimsIntraoperative defibrillation testing (DT) during implant or replacement of implantable cardioverter-defibrillators (ICDs) has been a matter of debate for many years. This debate was put to rest by the Simple and Nordic ICD trials, and the practice of testing during new implantations has essentially been almost abandoned. Old registries demonstrated an increased incidence of significant findings in DT during replacements. The aim of the present study was to evaluate frequency of significant findings and safety of DT in subjects undergoing device replacement.Methods and resultsA prospective observational multi-centre study included consecutive patients undergoing ICD generator replacement. The primary outcome was a failure to terminate induced ventricular fibrillation (VF) with a single shock 10 J below the maximal capacity of the device. Secondary outcomes included complications of DT. Patients were followed-up at 1- and 6-months post-procedure. A total of 92 patients were eligible, and consented to the study, of which 84 underwent DT during battery replacement. The median age was 68 years and 79.8% were males. Induction of VF was successful in 84 patients as was a successful conversion on the first attempt in all. There were no procedure-related complications. During follow up one patient had two appropriate ICD shock events. In four patients, ICD programming was changed. None suffered inappropriate shock. There was no evidence of lead malfunction. Two deaths occurred, none of which was related to arrhythmia.ConclusionThe present study found DT was not associated with complications in patients undergoing ICD generator replacement but produced no clinically important information.

Abstract Aims To evaluate the benefit of speckle tracking radial strain imaging (STRSI)-guided left ventricular (LV) lead (LVL) positioning in cardiac resynchronization therapy (CRT) in patients (pts) with ischaemic cardiomyopathy with CRT indication. Methods and results We conducted a prospective randomized controlled trial. Patients were enrolled in nine centres with 2:1 randomization into two groups (guided vs. control). Patients underwent STRSI to identify the optimal LV position from six LV segments at midventricular level. Implantation via STRSI was attempted for recommended segment in the guided group only. Follow-up included echocardiography (6 months) and clinical evaluation (6 and 12 months). The primary endpoint was comparison % reduction in LV end-systolic volume at 6 months with baseline. Secondary endpoints included hospitalizations for heart failure and death, and improvement in additional echocardiographic measurements and quality of life score. A total of 172 patients (115 guided vs. 57 control) were enrolled. In the guided group, 60% of the implanted LV leads were adjudicated to be successfully located at the recommended segment, whereas in the control group 44% reached the best STRSI determined segment. There was no difference between the groups in any of the primary or secondary endpoints at 6 and 12 months. Conclusion Our findings suggest that echo-guided implantation of an LV lead using STRSI does not improve the clinical or echocardiographic response compared with conventional implantation.

Background Detecting significant coronary artery disease (CAD) in the general population is complex and relies on combined assessment of traditional CAD risk factors and noninvasive testing. We hypothesized that a CAD‐specific heart rate variability (HRV) algorithm can be used to improve detection of subclinical or early ischemia in patients without known CAD. Methods and Results Between 2014 and 2018 we prospectively enrolled 1043 patients with low to intermediate pretest probability for CAD who were screened for myocardial ischemia in tertiary medical centers in the United States and Israel. Patients underwent 1‐hour Holter testing, with immediate HRV analysis using the HeartTrends DyDx algorithm, followed by exercise stress echocardiography (n=612) or exercise myocardial perfusion imaging (n=431). The threshold for low HRV was identified using receiver operating characteristic analysis based on sensitivity and specificity. The primary end point was the presence of myocardial ischemia detected by exercise stress echocardiography or exercise myocardial perfusion imaging. The mean age of patients was 61 years and 38% were women. Myocardial ischemia was detected in 66 (6.3%) patients. After adjustment for CAD risk factors and exercise stress testing results, low HRV was independently associated with a significant 2‐fold increased likelihood for myocardial ischemia (odds ratio, 2.00; 95% CI, 1.41–2.89 [ P =0.01]). Adding HRV to traditional CAD risk factors significantly improved the pretest probability for myocardial ischemia. Conclusions Our data from a large prospective international clinical study show that short‐term HRV testing can be used as a novel digital‐health modality for enhanced risk assessment in low‐ to intermediate‐risk individuals without known CAD. Clinical Trial Registration URL: http://www.ClinicalTrials.gov . Unique identifiers: NCT01657006, NCT02201017).