Autonomous artificial intelligence for sorting results of preventive radiological examinations of chest organs: medical and economic efficiency

• DOI: 10.17816/dd641703
• Publication Date: 2025-03-04T10:14:12Z
• Authors: Yuriy A. Vasilev; Dmitry A. Sychev; Alexander V. Bazhin; Igor M. Shulkin; Anton V. Vladzymy...; Alexandra Yu. Gol...; Kirill M. Arzamasov; Andrei V. Mishchenko; Gevorg A. Bekdzha...; Arcadiy S. Goldberg; Larisa G. Rodionova

BACKGROUND: We propose a novel model for processing chest radiographs acquired through population screening using classification by autonomous artificial intelligence (AI) as medical device with maximum sensitivity of 1.0 (95% CI 1.0,1.0). The system categorizes examinations (X-ray and photofluorography the chest) into two groups: 'abnormal' 'normal'. abnormal category encompasses all deviations (i.e., pathological conditions, post-surgical changes, age-related variations, congenital features) to be reviewed radiologists. normal includes studies without findings that implicitly do not require radiologist interpretation. AIM: This study aims evaluate feasibility effectiveness AI in radiography screening. METHODS: conducted prospective, multicenter diagnostic assess safety performance studies. analyzed 575,549 images fluoroscopy machines processed three different AI-powered devices. scientific merit was achieved statistical analytical methods. RESULTS: classified 54.8% 'normal', proportionally reducing workload otherwise would have been spent on interpretation reporting. demonstrated 99.95% accuracy classification. Clinically significant discrepancies occurred 0.05% cases CI: 0.04, 0.06%). CONCLUSION: demonstrates clinical economic radiograph means Future steps should focus regulatory framework updates legitimize implementation devices specific preventive tasks.