Background: Type 1 Brugada syndrome (BrS) is a hereditary arrhythmogenic disease showing peculiar electrocardiographic (ECG) patterns, characterized by ST-segment elevation in the right precordial leads, and risk of Sudden Cardiac Death (SCD). Furthermore, although various ECG patterns are described in the literature, different individual ECG may show high-grade variability, making the diagnosis problematic. The study aims to develop an innovative system for an accurate diagnosis of Type 1 BrS based on ECG pattern recognition by Machine Learning (ML) models and blood markers analysis trough transcriptomic techniques.
Methods: The study is structured in 3 parts: (a) a retrospective study, with the first cohort of 300 anonymized ECG obtained in already diagnosed Type 1 BrS (75 spontaneous, 150 suspected) and 75 from control patients, which will be processed by ML analysis for pattern recognition; (b) a prospective study, with a cohort of 11 patients with spontaneous Type 1 BrS, 11 with drug-induced Type 1 BrS, 11 suspected BrS but negative to Na + channel blockers administration, and 11 controls, enrolled for ECG ML analysis and blood collection for transcriptomics and microvesicles analysis; (c) a validation study, with the third cohort of 100 patients (35 spontaneous and 35 drug-induced BrS, 30 controls) for ML algorithm and biomarkers testing.
Discussion: The BrAID system will help clinicians improve the diagnosis of Type 1 BrS by using multiple information, reducing the time between ECG recording and final diagnosis, integrating clinical, biochemical and ECG information thus favoring a more effective use of available resources. Trial registration Clinical Trial.gov, NCT04641585. Registered 17 November 2020, https://clinicaltrials.gov/ct2/show/NCT04641585.
Keywords: Brugada syndrome; Machine learning; RNA; Transcriptomic.
© 2021. The Author(s).