Automatic triage of twelve-lead electrocardiograms using deep convolutional neural networks: a first implementation study.

Aims: Expert knowledge to correctly interpret electrocardiograms (ECGs) is not always readily available. An artificial intelligence (AI)-based triage algorithm (DELTAnet), able to support physicians in ECG prioritization, could help reduce current logistic burden of overreading ECGs and improve time to treatment for acute and life-threatening disorders. However, the effect of clinical implementation of such AI algorithms is rarely investigated. Methods and results: Adult patients at non-cardiology departments who underwent ECG testing as a part of routine clinical care were included in this prospective cohort study. DELTAnet was used to classify 12-lead ECGs into one of the following triage classes: normal, abnormal not acute, subacute, and acute. Performance was compared with triage classes based on the final clinical diagnosis. Moreover, the associations between predicted classes and clinical outcomes were investigated. A total of 1061 patients and ECGs were included. Performance was good with a mean concordance statistic of 0.96 (95% confidence interval 0.95-0.97) when comparing DELTAnet with the clinical triage classes. Moreover, zero ECGs that required a change in policy or referral to the cardiologist were missed and there was a limited number of cases predicted as acute that did not require follow-up (2.6%). Conclusion: This study is the first to prospectively investigate the impact of clinical implementation of an ECG-based AI triage algorithm. It shows that DELTAnet is efficacious and safe to be used in clinical practice for triage of 12-lead ECGs in non-cardiology hospital departments.

The circadian clock remains intact, but with dampened hormonal output in heart failure.

BACKGROUND: Circadian (24-h) rhythms are important regulators in physiology and disease, but systemic disease may disrupt circadian rhythmicity. Heart failure (HF) is a systemic disease affecting hormonal regulation. We investigate whether HF affects the rhythmic expression of melatonin and cortisol, main endocrine products of the central clock, and cardiac-specific troponin in patients. We corroborate the functionality of the peripheral clock directly in the organs of translational models, inaccessible in human participants. METHODS: We included 46 HF patients (71.7% male, median age of 60 years, NYHA class II (32.6%) or III (67.4%), ischemic cardiomyopathy (43.5%), comorbidities: diabetes 21.7%, atrial fibrillation 30.4%), and 24 matched controls. Blood was collected at seven time-points during a 24-h period (totalling 320 HF and 167 control samples) for melatonin, cortisol, and cardiac troponin T (cTnT) measurements after which circadian rhythms were assessed through cosinor analyses, both on the individual and the group level. Next, we analysed peripheral circadian clock functionality using cosinor analysis in male animal HF models: nocturnal mice and diurnal zebrafish, based on expression of core clock genes in heart, kidneys, and liver, every 4 h during a 24-h period in a light/darkness synchronised environment. FINDINGS: Melatonin and cortisol concentrations followed a physiological 24-h pattern in both patients and controls. For melatonin, acrophase occurred during the night for both groups, with significantly decreased amplitude (median 5.2 vs 8.8, P = 0.0001) and circadian variation ([maximum]/[minimum]) in heart failure patients. For cortisol, mesor showed a significant increase for HF patients (mean 331.9 vs 275.1, P = 0.017) with a difference of 56.8 (95% CI 10.3-103.3) again resulting in a relatively lower variation: median 3.9 vs 6.3 (P = 0.0058). A nocturnal blood pressure dip was absent in 77.8% of HF patients. Clock gene expression profiles (Bmal, Clock, Per, Cry) were similar and with expected phase relations in animal HF models and controls, demonstrating preserved peripheral clock functionality in HF. Furthermore, oscillations in diurnal zebrafish were expectedly in opposite phases to those of nocturnal mice. Concordantly, cTnT concentrations in HF patients revealed significant circadian oscillations. INTERPRETATION: Central clock output is dampened in HF patients while the molecular peripheral clock, as confirmed in animal models, remains intact. This emphasises the importance of taking timing into account in research and therapy for HF, setting the stage for another dimension of diagnostic, prognostic and therapeutic approaches. FUNDING: Hartstichting.

Xenotransplantation of Human Cardiomyocyte Progenitor Cells Does Not Improve Cardiac Function in a Porcine Model of Chronic Ischemic Heart Failure. Results from a Randomized, Blinded, Placebo Controlled Trial.

BACKGROUND: Recently cardiomyocyte progenitor cells (CMPCs) were successfully isolated from fetal and adult human hearts. Direct intramyocardial injection of human CMPCs (hCMPCs) in experimental mouse models of acute myocardial infarction significantly improved cardiac function compared to controls. AIM: Here, our aim was to investigate whether xenotransplantation via intracoronary infusion of fetal hCMPCs in a pig model of chronic myocardial infarction is safe and efficacious, in view of translation purposes. METHODS & RESULTS: We performed a randomized, blinded, placebo controlled trial. Four weeks after ischemia/reperfusion injury by 90 minutes of percutaneous left anterior descending artery occlusion, pigs (n = 16, 68.5 ± 5.4 kg) received intracoronary infusion of 10 million fetal hCMPCs or placebo. All animals were immunosuppressed by cyclosporin (CsA). Four weeks after infusion, endpoint analysis by MRI displayed no difference in left ventricular ejection fraction, left ventricular end diastolic and left ventricular end systolic volumes between both groups. Serial pressure volume (PV-)loop and echocardiography showed no differences in functional parameters between groups at any timepoint. Infarct size at follow-up, measured by late gadolinium enhancement MRI showed no difference between groups. Intracoronary pressure and flow measurements showed no signs of coronary obstruction 30 minutes after cell infusion. No premature death occurred in cell treated animals. CONCLUSION: Xenotransplantation via intracoronary infusion of hCMPCs is feasible and safe, but not associated with improved left ventricular performance and infarct size compared to placebo in a porcine model of chronic myocardial infarction.

Expanding mouse ventricular cardiomyocytes through GSK-3 inhibition.

Controlled proliferation of cardiomyocytes remains a major limitation in cell biology and one of the main underlying hurdles for true modern regenerative medicine. Here, a technique is described for robust expansion of early fetal-derived mouse ventricular cardiomyocytes on a platform usable for high-throughput molecular screening, tissue engineering and, potentially, in vivo translational experiments. This method provides a small-molecule approach to control proliferation or differentiation of early beating cardiomyocytes through modulation of the Wnt/ß-catenin signaling pathway. Moreover, isolation and expansion of fetal cardiomyocytes takes less than 3 weeks, yields a relatively pure (∼70%) functional myogenic population, and is highly reproducible.