欢迎来到《四川大学学报(医学版)》

健康志愿者心肌磁共振纹理特征初探

Myocardial Magnetic Resonance Texture Characteristics of Healthy Volunteers

  • 摘要:
      目的  探讨汉族健康志愿者心肌磁共振图像的纹理特征。
      方法  收集健康汉族志愿者59例,行3.0T心脏核磁共振电影序列及增强前后的T1 mapping序列扫描。计算每位健康志愿者的心功能参数。测量左室心肌的Native T1值,并计算细胞外容积(extracellular volume, ECV)值。在增强前的T1 mapping图像上提取40组心肌纹理特征,分析心肌纹理特征与年龄、Native T1、ECV值之间的相关性,并比较心肌纹理特征在性别、年龄分层组中的差异。
      结果  共纳入59例健康汉族志愿者(女性28例,18~76岁),其中低龄组(< 45岁)29例,高龄组(≥45岁)30例。男性左室心肌质量指数(left ventricular mass index, Lvmassi)高于女性、Native T1值低于女性(P<0.01),其余血压、心率、左室射血分数(left ventricular ejection fraction, LVEF)以及ECV值在不同性别组间差异均无统计学意义。在40个心肌纹理特征中,有10个心肌纹理特征存在性别差异(P<0.01),在低龄组中,男性和女性有13个心肌纹理特征存在差异;高龄组心肌纹理特征不存在性别差异(P>0.01)。有8个心肌纹理特征与年龄相关(最大r=0.483,P<0.01)。有6个一阶统计量与左心室心肌中层Native T1值相关(最大r=0.792,P<0.01),3个一阶统计量与ECV相关(最大r=0.409,P<0.01)。
      结论  基于T1 mapping的心肌纹理特征在中国汉族健康人群中存在年龄和性别差异。

     

    Abstract:
      Objective  To determine the myocardial texture features of cardiac magnetic resonance (CMR) in healthy adult Han populations.
      Methods  59 healthy Han volunteers were recruited for this study from May 2016 to November 2017. CMR examinations were performed on the participants with a 3.0T scanner (Tim Trio, Siemens Medical Solution) to estimate the functional parameters, Native T1 value and ECV. Texture analysis (TA) was performed on the region of interest (ROI) in the left ventricle myocardium on T1 mapping images, with 40 myocardial texture features being extracted. Differences in the myocardial texture features across gender and age groups were analyzed through Student's t-tests or Wilcoxon signed-rank tests. Spearman correlations were analyzed between the myocardial texture features and age, native T1 value and extracellular volume (ECV).
      Results  Of the 59 participants, 28 were women and 29 were in the younger age group (< 45 years old). The male participants had higher left ventricular mass index (Lvmassi) and lower native T1 than their female counterparts (P < 0.01). No gender differences in blood pressure, heart rate, left ventricular ejection fraction (LVEF) and ECV values were found. Ten of the forty myocardial texture features showed gender differences, including two first order features and eight Grey-level co-occurrence matrix (GLCM) features. Gender differences appeared in five first order features and eight GLCM features in the younger group (< 45 years old), but not in the older group (≥45 years old). Eight myocardial texture features were correlated with age, including five first order features and three GLCM features (all P < 0.01). Six first-order texture features were correlated with Native T1 values of the left ventricle middle myocardium. Three first-order texture features were correlated with ECV.
      Conclusion  Myocardial texture features in T1 mapping images vary by gender and age in healthy Han populations.

     

/

返回文章
返回