Contour-based Registration for Left Ventricular Myocardial Extracellular Volume Fraction Using T1 Mapping in Healthy Adults
- VernacularTitle:轮廓配准法在MRI评估健康成年人左心室心肌细胞外容积分数中的价值
- Author:
Xu-Hui ZHOU
1
;
Mei ZHENG
;
Qiu-Xia XIE
;
Ling LIN
Author Information
1. 中山大学附属第八医院影像科
- Keywords:
cardiovascular magnetic resonance imaging;
T1 mapping;
extracellular volume fraction;
contour-based registration
- From:
Journal of Sun Yat-sen University(Medical Sciences)
2018;39(1):130-137
- CountryChina
- Language:Chinese
-
Abstract:
[Objective]We aimed to evaluate the role of contour-based registration in quantification of myocardial extracellular volume fraction(ECV)based on T1 mapping technique.[Methods]T1 mapping images of the basal,mid-cavity and apical short axis slices of 26 healthy volunteers(16 males and 10 females)were obtained before and after administration of contrast agent using Modified Look-Locker Inversion Recovery(MOLLI)on a 3.0T Magnetic Resonance Imaging system.ECV was calculated by routine method and registered method.For each slice,subjects were divided into the deformation subgroup and the control subgroup.Routine ECVs and registered ECVs were compared for each subgroup.[Results]Left ventricular deformation among T1 mapping images occurred in 16 subjects(61.5%). In all three slices, registered ECV maps had better image quality and more highest grade images than routine ECV maps. Routine ECVs[(26.81 ± 2.78)%,(25.38 ± 3.05)%,(28.66 ± 4.10)%]were statistically different from registered ECVs in global [(25.75±2.42)%,P=0.001],mid-cavity[(24.30±2.45)%,P=0.016]and apical slices[(27.22±3.38)%,P=0.010]. In deformation subgroups of all the 3 slices,registered ECVs were lower than routine ECVs with smaller SDs(P=0.038, 0.012,0.016).The apical ECVs were higher with larger SDs than the other two slices and the global ECVs(routine:F=4.799,P=0.004;registered:F=4.822,P=0.003).[Conclusions]Contour-based registration can improve the image qual-ity and precision of ECV quantification in cases with ventricular shape deformation among source images.