Quantification of atrial fibrillation–related imaging markers using spectral photon-counting computed tomography

Background Structural remodelling of the cardiac conduction system, including fibrosis and impaired vascular supply, contributes to atrial fibrillation (AF). In vivo characterisation has been limited by conventional imaging. Spectral photon-counting CT (PCCT) provides high spatial resolution and quantitative tissue characterisation and may enable identification of AF-related imaging biomarkers. Purpose To assess PCCT markers of sinoatrial (SA) node vascular supply and interatrial septum (IAS) remodelling associated with AF. Methods In this retrospective cohort study, patients undergoing clinically indicated PCCT coronary CT angiography (July–December 2025) were screened. From 518 examinations screened, sixty-six patients were included (33 AF, 33 controls; mean age 66.9±12.8 years; 42.4% women). PCCT markers included SA node artery visibility, origin and length score; IAS attenuation and spectral iodine concentration; IAS fibrosis score (4-point scale); SA node attenuation and iodine; and left atrial wall thickness (LAWT>2 mm). Group comparisons, correlation analyses and a random forest model for feature importance were performed. Results AF patients had reduced SA artery visibility and shorter SA arteries (p<0.001), with no difference in origin. IAS fibrosis was more prevalent in AF (mixed/fibrotic 84.8% vs 12.1%; fully fibrotic 60.6% vs 3%; p<0.001). IAS attenuation and iodine concentration were higher in AF and strongly correlated (r=0.708; p<0.001). IAS fibrosis correlated with shorter SA artery length (r=0.731; p<0.001). LAWT>2 mm was associated with higher IAS fibrosis and higher attenuation and iodine (all p<0.001). SA node attenuation was higher in AF (p=0.014). In machine learning analysis, LAWT, IAS fibrosis/texture and SA artery length were the most important features, while clinical variables contributed less. Conclusions Spectral PCCT enables quantitative assessment of conduction system remodelling in AF. AF is characterised by shorter SA arteries and increased IAS density and iodine suggestive for fibrosis, supporting an interplay between vascular supply and structural remodelling. These markers may help identify patients with subclinical AF undergoing coronary PCCT and support targeted screening.