Optimization of MP2RAGE sequence parameters for head and neck T1-mapping and OE-MRI

Abstract

Cancer is the collective name for more than 200 types of diseases that annually causes the death of 8.8 million people, cancer in the head and neck (H&N) accounts for 330.000 of these. It has been shown that hypoxic tumors are more resistant to treatment than well oxygenated tumors. The possibility to examine hypoxia in a tumor could be used for predicting external radiation treatment response, and hence enable adaptive treatment. Today, there is no standardized method to ex amine hypoxia. Oxygen Enhanced Magnetic Resonance Imaging (OE-MRI) is an in-vivo technique that evaluates the change in T1-value when breathing oxygen com pared to when breating air. While breathing oxygen, if the excess oxygen does not bind to hemoglobin molecules, inhomogeneities will occur since oxygen is paramag netic which shortens the T1-value of the tissue. T1-mapping is a technique where T1-values of tissues are calculated, and by combining OE-MRI and T1-mapping the change of T1-value caused by the excess oxygen can be detected. By using the pulse sequence Magnetization Prepared 2 Rapid Acquistition Gradient Echoes (MP2RAGE), the measured signal will be T1-weighted and independent of M0, T∗2 and B1−, which is desirable for T1-mapping. The aim of the study was to find the optimal MP2RAGE parameters in order to create a Signal-to-T1 calibration curve to be used for T1-mapping and OE-MRI. To establish the optimal MP2RAGE parameters, the inversion times, repetition time and flip angles were optimized in three subsequent steps in order to find a combination parameters that gave rise to a high signal difference for T1-values in the H&N area. The optimized parameters were validated via phantom measurements by comparing the calculated T1-values with the specified T1-values in the phantom. In addition, the MP2RAGE sequence was used for OE-MRI on a patient with H&N cancer. In this study the optimization of MP2RAGE sequence was successful. The phantom validation showed that several combinations of MP2RAGE parameters resulted in accurate T1-maps. It was also shown that MP2RAGE parameters resulting in the largest difference in signal did not necessarily produce the most accurate T1-map.