Protocol optimization for MRI studies in radiation oncology: III. MRI relaxometry

One factor that impedes the uptake of (Formula presented) (Formula presented) and (Formula presented) (Formula presented) maps as quantitative imaging biomarkers in clinical radiotherapy is the large bias and disparity in acquired values across different centers and scanners. In this study, it is shown that the (Formula presented) (Formula presented) fitting method results in a large dependence of the acquired (Formula presented) (Formula presented) on the applied flip angles. A modification of the signal equation is suggested that reduces this variability. The use of mono-exponential fitting in (Formula presented) (Formula presented) mapping is responsible for a large bias that is dependent on the echo time spacing in a multi spin echo pulse sequence. A bi-exponential fit provides 3 parametric maps with distinct physiological features and results in more reliable and reproducible (Formula presented) (Formula presented) maps. QA phantoms that exhibit mono-exponential decay are unable to pick up or assess the impact of non-mono-exponential (Formula presented) (Formula presented) decay in tissue. Numerical simulations have been applied to determine the optimum range of imaging parameters to minimize the stochastic and systematic uncertainty in (Formula presented) (Formula presented) and (Formula presented) (Formula presented) maps.