南方医科大学学报

南方医科大学学报 ›› 2022, Vol. 42 ›› Issue (11): 1646-1654.doi: 10.12122/j.issn.1673-4254.2022.11.08

• • 上一篇    下一篇

经导管主动脉瓣置换术前测量主动脉根部参数的一致性及与短期预后的关联性：基于Anythink和3mensio CT软件比较

刘长福，孙泽瑜，王 晶，王民汉，辛 然，丁熠璞，王 玺，穆 洋，陈 韬，蒋 博，王 琳，章 明，单冬凯，陈韵岱   

  1. 中国人民解放军总医院第六医学中心心血管病医学部，北京 100853；南开大学医学院，天津 300071；北京思创贯宇科技开发有限公司，北京 100853
  • 出版日期:2022-11-20 发布日期:2022-11-30

Anythink for CT-based aorta root measurements before transcatheter aortic valve replacement: measurement consistency with 3mensio and impact on short-term prognosis

LIU Changfu, SUN Zeyu, WANG Jing, WANG Minhan, XIN Ran, DING Yipu, WANG Xi, MU Yang, CHEN Tao, JIANG Bo, WANG Lin, ZHANG Ming, SHAN Dongkai, CHEN Yundai   

  1. Senior Department of Cardiology, Sixth Medical Center of PLA General Hospital, Beijing 100853, China; School of Medicine, Nankai University, Tianjin 300071, China; Beijing Crealife Technology Co., Ltd, Beijing 100853, China
  • Online:2022-11-20 Published:2022-11-30

RichHTML

35

PDF

295

摘要： 目的 通过对比国产半自动术前CT分析软件Anythink与3mensio应用于经导管主动脉瓣置换术前测量的数据，评价Anythink软件测量主动脉根部参数的一致性与短期预后间的关联性。方法 回顾性纳入2016年12月~2022年2月于中国人民解放军总医院行TAVR手术的67例患者，由一名完成专业培训的心内科医生分别使用新型半自动分析软件Anythink与“金标准”3mensio软件同步重建主动脉根部模型，测量主动脉瓣环及周围结构，分析两种软件结果的相关性及一致性；由两名独立医生应用Anythink软件重复测量，评价Anythink对于同一研究对象测量的可重复性；依据Anythink和3mensio软件测量结果选取瓣膜型号，分析两软件在实际用于指导临床瓣膜选择时的异同。结果 Anythink与3mensio测量的左冠脉开口高度（13.37±3.35 mm vs 13.19±3.19 mm）、右冠脉开口高度（15.45±2.89 mm vs 15.75±2.93mm）、瓣环直径（24.0±2.2 mm vs 23.9±2.3 mm）、瓣环面积（450.6±88.3 mm2 vs 447.5±90.0 mm2）、瓣环周长（76.5±7.2 mm vs 76.1±7.6 mm）、主动脉瓣环与水平面夹角（53.5°±10.0°vs 51.1°±9.7°）差异均无统计学意义（P>0.05）。Pearson相关分析结果显示，两软件测量数据呈正相关（r=0.884~0.981，P<0.01）；Anythink测量瓣环数据组内相关系数ICC=0.894~0.992，组间相关系数ICC=0.651~0.954；Anythink与3mensio依据平均直径、面积直径、周长直径的选取瓣膜型号Kappa检验值分别为0.886、0.796、0.775；依据两次Anythink测量的平均直径、面积直径、周长直径选取瓣膜型号组内Kappa值为0.819、0.841、0.795，组间Kappa值为0.812、0.812、0.768。与患者短期预后关联性部分，在出现术后瓣周漏的患者中，Anythink测量面积直径相较于3mensio软件推荐的瓣膜型号略大，而在出现术后新发传导阻滞的患者中，Anythink推荐的瓣膜型号略小。结论 国产半自动TAVR术前CT分析软件Anythink与3mensio软件在主动脉根部参数测量方面具有较好的一致性与较高的可重复性。经过专业训练的心内科医生可运用Anythink软件来获得相对准确的主动脉根部数据，为TAVR术前瓣膜型号的选择提供指导和参考。

关键词: 主动脉瓣膜病；经导管主动脉瓣置换术；计算机断层扫描；瓣环大小

Abstract: Objective To evaluate the consistency and reproducibility of aortic root measurements by Anythink, a semi-automated preoperative CT analysis software, with those of 3mensio. Methods Sixty-seven patients undergoing transcatheter aortic valve replacement (TAVR) in the First Medical Center of Chinese PLA General Hospital from December, 2016 to February, 2022 were retrospectively enrolled in this study. A cardiology resident who completed his professional training used both the software Anythink and 3mensio (as the gold standard) to reconstruct the aortic root model and analyze the parameters of the aortic annulus and the surrounding structures. The correlation and consistency of the measurement results of two software were analyzed. Two independent residents also used Anythink software to repeat the measurements for the same patient for assessment of the reproducibility of Anythink measurements. The valve models were selected based on the measurements by Anythink and 3mensio, and similarities and differences of the two software in clinical valve selection were assessed. Results The measurements of the distances from the anulus plane to the left and right coronary ostium, average diameter of the anulus, anulus area, anulus perimeter, and the angle between the annulus and horizontal plane did not differ significantly between the two software (P>0.05), and their measurements showed positive correlations (r=0.884- 0.981, P<0.01). The intra-group and inter-group correlation coefficients of the anulus parameters measured by Anythink ranged from 0.894 to 0.992 and from 0.651 to 0.954, respectively. The Kappa-test values of valve models selected by Anythink and 3mensio based on the average diameter, area diameter and perimeter diameter were 0.886, 0.796 and 0.775, respectively. The intra-group Kappa values for the valve models selected based on Anythink measurements were 0.819, 0.841, and 0.795, and the inter-group Kappa values were 0.812, 0.812, and 0.768, respectively. Compared with the measurements by 3mensio, the recommended area diameter measured by Anythink was slightly greater in patients with postoperative paravalvular leakage, but slightly smaller in patients with postoperative new-onset conduction block. Conclusion Anythink has excellent measurement consistency and high reproducibility for aortic root measurements, and trained cardiologists can use Anythink to obtain accurate aortic root parameters before TAVR.

Key words: aortic valve disease; transcatheter aortic valve replacement; computed tomography; annulus sizing