Journal of Associated Medical Sciences (Sep 2010)
A Comparison Between Breath hold and Free Breathing MRI Techniques of Myocardial T2* Measurement
Abstract
Objectives : To compare the effectiveness and reproducibility of myocardial T2* measurement between breath-hold and free-breathing techniques using GRE black blood multi-echoes pulse sequence. Methods : The techniques were validated by a known iron concentration phantom prior to study in 15 healthy volunteers. The images of single short-axis view of the mid left ventricle were acquired at 10 echo times using both Breath-hold and Free-breathing techniques. The images from both techniques were drawn the Regions of interest (ROI) at the left ventricular septum for T2* analysis. The parameters to be analyzed include Signal to Noise Ratio (SNR), reproducibility and percent coefficient of Variation (%CV). The fitting models used to evaluate for T2* values were the mono-exponential and offset models done by MATLAB 7.01. Results : The correlation between known iron concentration phantom and R2*(1000/T2*) using Pearson‘s test showed strongly linear correlation with a correlation coefficient over 0.96 (p<0.001). In human study, the images obtaining from the Free-breathing technique provided higher SNR than that of the Breath-hold approximately 40% in all TE images as well as the mean T2* values acquiring from different fitting models were no significant different (36.59±3.18 ms., and 36.31±2.72 ms., ( p = 0.799)), while the mean T2* values of Breath-hold technique obtaining from different fitting models were significant different (35.69±2.90 ms., and 30.67±7.21, (p = 0.036)). For the coefficients of variation (%CV), the T2* from the Free-breathing technique was lower than that of the Breath-hold technique, 6.34% and 8.41%, respectively. Conclusion : The Free-breathing T2* technique provided higher resolution and SNR compared to that of the Breath-hold T2* technique. In addition, the Free-breathing technique showed higher reproducibility when evaluated at different time frames of study, different fitting models, and different observers.
