RESUMEN
Aims: Machine-learning (ML)-based automated measurement of echocardiography images emerges as an option to reduce observer variability. The objective of the study is to improve the accuracy of a pre-existing automated reading tool ('original detector') by federated ML-based re-training. Methods and results: Automatisierte Vermessung der Echokardiographie was based on the echocardiography images of n = 4965 participants of the population-based Characteristics and Course of Heart Failure Stages A-B and Determinants of Progression Cohort Study. We implemented federated ML: echocardiography images were read by the Academic Core Lab Ultrasound-based Cardiovascular Imaging at the University Hospital Würzburg (UKW). A random algorithm selected 3226 participants for re-training of the original detector. According to data protection rules, the generation of ground truth and ML training cycles took place within the UKW network. Only non-personal training weights were exchanged with the external cooperation partner for the refinement of ML algorithms. Both the original detectors as the re-trained detector were then applied to the echocardiograms of n = 563 participants not used for training. With regard to the human referent, the re-trained detector revealed (i) superior accuracy when contrasted with the original detector's performance as it arrived at significantly smaller mean differences in all but one parameter, and a (ii) smaller absolute difference between measurements when compared with a group of different human observers. Conclusion: Population data-based ML in a federated ML set-up was feasible. The re-trained detector exhibited a much lower measurement variability than human readers. This gain in accuracy and precision strengthens the confidence in automated echocardiographic readings, which carries large potential for applications in various settings.
RESUMEN
AIMS: Agonistic antibodies against neurohumoral receptors can induce cardio-noxious effects by altering the baseline receptor activity. To estimate the prevalence of autoantibodies directed against the beta-1 receptor (b1-AAB) in patients admitted to the hospital for acute heart failure (HF) at (i) baseline and (ii) after 6 months of follow-up (F6) and (iii) after another 12 months of follow-up (i.e. 18 months after index hospitalization), to estimate their prognostic impact on clinical outcome (death or first hospitalization for HF). METHODS AND RESULTS: In 47 patients, b1-AAB were serially determined in serum samples collected at index hospitalization and at 6 months of follow-up (F6) with a flow cytometry-based assay: median age 71 years (quartiles 60, 80), 23 (49%) women, 24 (51%) HF with preserved ejection fraction. Beta1-AAB were detected in three subjects at index hospitalization (6%), and in eight subjects at F6 (17%). There were no differences apparent between patients with and without b1-AAB at F6 with regard to age, sex, type, duration, or main cause of HF. During the 12 month period following F6 (i.e. up to month 18), eight events occurred. Event-free survival was associated with prevalence of b1-AAB at F6. Compared with patients without b1-AAB at F6, age-adjusted Cox regression indicated a higher event risk in patients harbouring b1-AAB, with a hazard ratio of 8.96 (95% confidence interval 1.81-44.50, P = 0.007). CONCLUSIONS: Our results suggest a possible adverse prognostic relevance of b1-AAB in patients with acute HF, but this observation needs to be confirmed in larger patient collectives.
