Quantitative magnetic resonance imaging (MRI) is a robust and reproducible analysis method that has become the gold standard for cardiac analysis in clinical care. This thesis aims to optimize quantitative MRI characterization of myocardial tissue. A new method was introduced to obtain a synthetic image for drawing the left ventricle myocardial contours used for quantifying T2* for iron deposition detection, called the contrast-optimized composite image method, followed by a semi-automatic myocardial segmentation to ease segmentation difficulties in the bright blood multi-gradient echo image series. Two monoexponential T2* truncation methods were compared in quantifying iron loading at different iron content classifications. In T1 mapping, mean and median pixel value quantification were compared for their ability to detect heart function abnormality. Meanwhile, different perfusion models for quantifying myocardial blood flow (MBF) were equally sensitive to heart function. Higher reproducibility of T2* quantification was obtained by improving the TE image quality for myocardial segmentation and performing a semi-automatic myocardial segmentation. The two truncation methods showed similar T2* values at severe myocardial iron loading and diverged in lower myocardial iron content regions. Median quantification of pixel-wise measurement can replace the mean regardless of statistical data distribution, as shown in the early abnormality of heart function detected by native T1. Different MBF estimations are found between different perfusion models with the extended Toft and Fermi models allowing for better differentiation of the presence of significant stenosis. To conclude, this thesis has provided improved methods for characterizing myocardial tissue quantitatively.