Editorial for "Automatic Time-Resolved Cardiovascular Segmentation of 4D Flow MRI Using Deep Learning"
Editorial for "Automatic Time-Resolved Cardiovascular Segmentation of 4D Flow MRI Using Deep Learning"
复制标题
“使用深度学习对 4D 流 MRI 进行自动时间分辨心血管分割”的社论
DOI:
10.1002/jmri.28220
复制
发表时间:
2022
影响因子:
4.4
通讯作者:
Montalt-Tordera J
中科院分区:
文献类型:
--
作者:
Montalt-Tordera J
Flow quantification using 2D phase-contrast cine MRI (PC-MRI) has long been a part of clinical cardiovascular protocols and is extremely useful in the evaluation and quantification of cardiac and valve function (1). Over the last decade, 3D cine PC-MRI (more often called 4D flow MRI) has received increasing research interest and undergone important developments that have brought it closer to widespread clinical adoption (2). Unlike standard PC-MRI, 4D flow MRI offers comprehensive 3D anatomic coverage and flow quantification in all three spatial directions.This is beneficial for three main reasons. Firstly, it allows visualization of 3D anatomy and flow patterns, using techniques such as streamlines, vector fields, isosurfaces and volume renderings. Secondly, it enables retrospective multiplanar analysis at any location. This simplifies imaging protocols when it is necessary to measure flow in multiple locations, such as in cases of complex congenital heart disease (CHD). Finally, it enables the computation of advanced hemodynamic parameters, such as vorticity, helicity, wall shear stress, kinetic energy, viscous energy loss and blood stasis. Many useful applications of this type of analysis have been demonstrated (2, 3), and further research is warranted to establish its clinical value in different settings.