The influence of the analysis technique on myocardial T1 estimation using MRI

Cardiovascular disease is the main cause of death worldwide. Magnetic resonance imaging (MRI) provides unprecedented capabilities for myocardial tissue characterization. Specifically, T1 mapping showed to be a valuable technique for identifying myocardial fibrosis and scar tissues without the need for contrast agent administration. Nevertheless, various factors can influence the analysis technique and affect the resulting T1 measurements. These factors include the signal calculation method (Average, Median, or pixel-wise Mapping), size and location of the analyzed region-of-interest (ROI), and signal-to-noise ratio (SNR) level of the acquired images. In this study, we evaluate the influence of these factors on T1 measurement using numerical and calibrated phantoms. The numerical phantom results showed that the percentage error for T1 estimation reached 2.3%, 3.1%, and 12.28% for the Median, Average, and Mapping methods, respectively, at SNR level of 7 dB, compared to 0% error at high SNR (44 dB) .The calibrated phantom experiments revealed high correlation (R>0.9) between the estimated and reference T1 values at high SNR levels. Based on the results in this study, the Median calculation method outperforms the Average and Mapping methods due to its lower estimation error at low SNR and higher correlation values with reference measurements. Further, large ROI's that are placed at the center of the analyzed region result in more accurate T1 estimates than smaller ROI's or those placed closer to the object boundary.