Automated Analysis of Thoracic Aortic Aneurysms Based on Chest Computed Tomography Data: a Population Study in Moscow

Abstract

Background. Cardiovascular diseases remain a pressing issue associated with great economic burden and loss of earning capacity. Among them, thoracic aortic aneurysms pose a serious threat as they commonly develop asymptomatically. Patients with aneurysms are often diagnosed late, which contributes to higher mortality rates due to complications. For this reason, conducting new epidemiological studies on this problem in our country is relevant.Objective: to study the prevalence of pathological dilatation of the thoracic aorta in Moscow by means of artificial intelligence technologies (AITs) using chest computed tomography (CT) data.Material and methods. A retrospective analysis of chest CT data from 227,149 patients obtained in the period from October 2022 to October 2023 was performed using AITs.Results. The analysis revealed that 13.3% of patients exhibited signs of thoracic aortic dilatation, while 0.8% had aneurysm signs. The prevalence of thoracic aortic aneurysms in Moscow was 12.4 cases per 100,000 individuals. Males were more susceptible than females; aneurysms of the descending aorta were more typical for them, while aneurysms of the ascending aorta were more common in females. The incidence of thoracic aortic dilatation increased with age. Potential causes and strategies to minimize AIT errors were discussed.Conclusion. The results highlight the importance of opportunistic screening for thoracic aortic aneurysms to ensure timely detection and prevention of complications. It would be especially beneficial to men and elderly population. The effectiveness of AITs to support decision-making by radiologists analyzing chest CT results was confirmed. The study provides an important update to the data on the prevalence of thoracic aortic aneurysms in asymptomatic patients in Russia. The identified features make it possible to direct the diagnostic and prevention efforts more effectively.

Keywords

• computed tomography
• ct
• thoracic aortic dilatation
• thoracic aortic aneurysm
• artificial intelligence
• artificial intelligence technologies
• aits