Objective Exploring the impact of artificial intelligence automatic positioning technology on image quality and radiation dose in aortic CTA imaging.Methods Prospective collection of 480 patients undergoing aortic CTA in our hospital from January 2021 to November 2023,divided into Group A("Head First-Supine"group)and Group B("Feet First-Supine"group)based on different scanning positions,with 240 patients in each group,Among them,Group A and Group B are further divided into Group A1(scanning position is"Head First-Supine",positioning method is automatic po-sitioning)and Group A2(scanning position is"Head First-Supine",positioning method is manual positioning)according to the positioning method;B1 group(scanning position is"Feet First-Supine",positioning method is automatic positioning)and B2 group(scanning position is"Feet First-Supine",positioning method is manual positioning),with 120 cases in each group.Evaluate the off-center distance,objective image quality,subjective image quality,and radiation dose of each group.Results In two different positions of patients,the noise levels of the aortic trunk and branches of the abdominal trunk,left renal artery,right renal artery,and superior mesenteric artery in the automatic group were lower than those in the manual group.The SNR of the aortic trunk in group A1 increased by 12％to 21％compared to group A2,and the SNR of the aortic branch increased by 13％to 17％.The SNR of the aortic trunk in group B1 increased by 12％to 19％com-pared to group B2,and the SNR of the aortic branch increased by 10％to 16％;The CTDIvol of group A1 decreased by 7.4％compared to group A2,and the CTDIvol of group B1 decreased by 8.2％compared to group B2.Moreover,the aver-age DLP and ED of the automatic group were lower than those of the manual group.Conclusion AI based automatic po-sitioning technology can significantly reduce the off center distance of patients in different positions,reduce image noise,improve the quality of aortic vessel images,and reduce patient radiation dose in aortic CTA imaging.
NETL
NSTL
万方数据
