肋骨骨折CT智能辅助诊断系统开发、集成与应用OA

This study explores the integration scheme of an Artificial Intelligence(AI)-assisted diagnostic system for rib fractures on chest CT with the Picture Archiving and Communication System(PACS),aiming to significantly improve the efficiency and accuracy of rib fracture diagnosis.An efficient AI-assisted diagnostic system for rib fractures on chest CT is constructed,and an integration interface between the system and PACS is designed.A total of 187 patients with chest trauma are enrolled,and all patients undergo two CT scans on the day of chest trauma and within 1 to 3 months after injury.The presence of definite callus on the second follow-up chest CT is used as the gold standard for the diagnosis of rib fractures.The consistency between manual diagnosis and AI-assisted diagnosis based on the first chest CT,as well as the difference in diagnostic time,are compared and analyzed.The integrated AI-assisted diagnostic system improves the diagnostic efficiency of rib fractures,with the average diagnostic time shortened by 15.12%.Meanwhile,the system enhances the diagnostic efficacy of rib fractures that the average sensitivity for diagnosing complete rib fractures is increased by 5.06%,and the average positive predictive value is increased by 1.51%.The average sensitivity for diagnosing incomplete rib fractures is increased by 16.45%,and the average positive predictive value is increased by 4.65%,both of which are significantly superior to traditional manual diagnosis.This innovative integration scheme has important clinical application value and broad application prospects,and is expected to play an important role in clinical practice.