AI自动定位技术在主动脉CTA成像中的应用

Application of AI Automatic Positioning Technology in Aortic CTA Imaging

目的探讨人工智能(AI)自动定位技术在主动脉CT血管造影(CTA)成像中对图像质量和辐射剂量的影响。方法前瞻性搜集2021年1月至2023年11月行主动脉CTA患者480例,根据患者扫描体位不同分为A组(“仰卧头先进”组)和B组(“仰卧足先进”组),两组各240例,其中A组和B组根据定位方式进一步分为A1组(扫描体位为“仰卧头先进”,定位方式为自动定位)和A2组(扫描体位为“仰卧头先进”,定位方式为手动定位);B1组(扫描体位为“仰卧足先进”,定位方式为自动定位)和B2组(扫描体位为“仰卧足先进”,定位方式为手动定位),每组各120例。对各组定位的偏中心距离、客观图像质量、主观图像质量及辐射剂量进行评价。结果在两种体位患者中,自动组主动脉主干各层面及主动脉分支腹腔干、左肾动脉、右肾动脉、肠系膜上动脉的噪声较手动组均降低,A1组较A2组主动脉主干SNR提升了12%~21%,主动脉分支SNR提升了13%~17%,B1组较B2组主动脉主干信噪比(SNR)提升了12%~19%,主动脉分支SNR提升了10%~16%;A1组的CT容积剂量指数(CTDIvol)较A2组降低了7.4%,B1组的CTDIvol较B2组降低了8.2%,且自动组的平均剂量长度乘积(DLP)和有效辐射剂量(ED)均低于手动组。结论基于AI的自动定位技术在主动脉CTA成像中能够显著减小不同体位患者的偏中心距离,降低图像噪声,提高主动脉血管图像质量,减小患者辐射剂量。

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 positioning)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%compared 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 average DLP and ED of the automatic group were lower than those of the manual group.Conclusion AI based automatic positioning 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.