Improving aortic aneurysm detection with artificial intelligence based on chest computed tomography data

• DOI: 10.17816/dd569388
• Publication Date: 2024-03-11T10:33:55Z
• Authors: Alexander V. Solovev; Yuriy A. Vasilev; Valentin E. Sinitsyn; Alexey V. Petraikin; Anton V. Vladzymy...; Igor M. Shulkin; Daria E. Sharova; Dmitry S. Semenov
• Keywords: 03 medical and health sciences; 0302 clinical medicine

BACKGROUND: Aortic aneurysms are known as “silent killers” because this potentially fatal condition can be asymptomatic. The annual incidence of thoracic aortic and ruptures is approximately 10 1.6 per 100,000 individuals, respectively. mortality rate for ruptured ranges from 94% to 100%. Early diagnosis treatment life-saving. Artificial intelligence technologies significantly improve diagnostic accuracy save the lives patients with aneurysms. AIM: This study aimed assess efficacy artificial detecting on chest computed tomography scans, well possibility using a clinical decision support system radiologists during primary interpretation radiological images. MATERIALS AND METHODS: results non-contrast scans were evaluated. A sample 84,405 18 years old was generated, 86 cases suspected based data selected retrospectively assessed by vascular surgeons. To age distribution diameter, an additional 968 randomly total number. RESULTS: In 44 cases, initially identified radiologists, whereas in 31 not detected initially; however, aided their detection. Six studies excluded, five had false-positive results. aids highlighting pathological changes medical images, increasing detection 41% when interpreting scans. use interpretations retrospective assessments advisable prevent underdiagnosis clinically significant pathologies enlargement. sample, dilation adults 14.5% 1.2%, findings also revealed age-dependent diameter aorta both men women. CONCLUSION: such Expanding screening concomitant negative consequences.