Atrial fibrillation (AF) has been extensively studied for decades, typically categorized by episode duration: paroxysmal, persistent, long-standing persistent, and permanent AF. Alternative classifications focus on mechanisms like multiple rotors, singular mother rotors, focal sources, multiple wavelet hypothesis, epi-endo dissociation, and micro reentry, all supported by data and computer models.
Despite this knowledge, pulmonary vein isolation (PVI) remains the sole proven ablation treatment for paroxysmal AF. This presentation explores AF from a new angle and poses critical questions:
Complexity: There's a trend towards complex models to understand AF, including digital twins. Before advancing model complexity, we must ask whether we fully understand the simpler models. I will discuss how a basic theorem in atrial tachycardia was overlooked for 35 years despite available evidence.
The scarcity of publicly available datasets for scientific research is a significant barrier. This limitation is not only due to privacy concerns but also reflects a broader cultural reluctance to share data freely. Moreover, many different types of datasets exists, which can also have an influence on the interpretation of the data.
Open Source Code: There is a lack of open-source codes available for analyzing AF datasets. Researchers often develop code that is seldom shared post-publication. Additionally, many companies operate as 'black boxes,' further restricting access to valuable tools and information.
Heterogeneous Patient Populations: Atrial fibrillation (AF) presents a unique challenge due to its heterogeneity; a wide array of pathologies can manifest as AF, each potentially requiring different treatment approaches. This diversity within the AF patient population suggests that a one-size-fits-all model may not be appropriate.
Through this talk, I hope to challenge existing paradigms and encourage a more collaborative and transparent approach to AF research.