文献速递心血管MR

全自动人工智能心脏运动参数提取——应用于长轴电影磁共振图像上的二尖瓣与三尖瓣

Fully automated AI-based cardiac motion parameter extraction - application to mitral and tricuspid valves on long-axis cine MR images.
PubMed 文献 · Eur J Radiol · 2023-07-13

研究开发了全自动人工智能瓣膜追踪系统,在2腔或4腔电影序列上自动提取房室平面位移、速度、二尖瓣/三尖瓣环收缩期位移及纵向缩短等运动参数,并基于5289例患者数据库评估。系统对AVPD、AVPV、MAPSE、TAPSE和LS的偏差分别为1.44±1.32 mm、1.51±1.46 cm/s、2.21±1.81 mm、2.40±1.97 mm、2.50±2.06 mm。

要点

  • 全自动AI瓣膜追踪系统可自动提取二尖瓣和三尖瓣相关运动参数,平均MAPSE为9.5±3.0 mm,平均LS为15.9±3.9%。
  • 平均TAPSE为13.4±4.7 mm,平均LS为21.4±6.9%。
  • 系统精度以偏差表示:AVPD 1.44±1.32 mm、AVPV 1.51±1.46 cm/s、MAPSE 2.21±1.81 mm、TAPSE 2.40±1.97 mm、LS 2.50±2.06 mm。

核心结论

  • 结果表明该系统可自动提取多种瓣膜相关运动参数,具有多功能性。

研究设计与样本

  • 研究开发并评估了基于两个序贯神经网络(定位网络和配准网络)的全自动AI瓣膜追踪系统,在大型心脏磁共振数据集(N=5289)的2腔或4腔电影序列上进行评估。

关键影像表现/征象

  • 在2腔和4腔视图上,平均MAPSE和LS分别为9.5±3.0 mm和15.9±3.9%。
  • 平均TAPSE和LS分别为13.4±4.7 mm和21.4±6.9%。

诊断标准与关键数值

  • 系统精度以偏差表示:AVPD为1.44±1.32 mm,AVPV为1.51±1.46 cm/s,MAPSE为2.21±1.81 mm,TAPSE为2.40±1.97 mm,LS为2.50±2.06 mm。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

PURPOSE: In cardiac MRI, valve motion parameters can be useful for the diagnosis of cardiac dysfunction. In this study, a fully automated AI-based valve tracking system was developed and evaluated on 2- or 4-chamber view cine series on a large cardiac MR dataset. Automatically derived motion parameters include atrioventricular plane displacement (AVPD), velocities (AVPV), mitral or tricuspid annular plane systolic excursion (MAPSE, TAPSE), or longitudinal shortening (LS).

METHOD: Two sequential neural networks with an intermediate processing step are applied to localize the target and track the landmarks throughout the cardiac cycle. Initially, a localisation network is used to perform heatmap regression of the target landmarks, such as mitral, tricuspid valve annulus as well as apex points. Then, a registration network is applied to track these landmarks using deformation fields. Based on these outputs, motion parameters were derived.

RESULTS: The accuracy of the system resulted in deviations of 1.44 ± 1.32 mm, 1.51 ± 1.46 cm/s, 2.21 ± 1.81 mm, 2.40 ± 1.97 mm, 2.50 ± 2.06 mm for AVPD, AVPV, MAPSE, TAPSE and LS, respectively. Application on a large patient database (N = 5289) revealed a mean MAPSE and LS of 9.5 ± 3.0 mm and 15.9 ± 3.9 % on 2-chamber and 4-chamber views, respectively. A mean TAPSE and LS of 13.4 ± 4.7 mm and 21.4 ± 6.9 % was measured.

CONCLUSION: The results demonstrate the versatility of the proposed system for automatic extraction of various valve-related motion parameters.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Yoon SS, Fischer C, Amsel D, et al. Fully automated AI-based cardiac motion parameter extraction - application to mitral and tricuspid valves on long-axis cine MR images. Eur J Radiol. 2023. PMID: 37517314. doi:10.1016/j.ejrad.2023.110978

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。