Fast quantitative T1 mapping based on simultaneous-multi-slice acquisitions using spiral gradients and model-based reconstruction algorithms applied to whole head MR imaging at 3 Tesla

Quantitative Magnetic Resonance Imaging is a technique that focuses on obtaining maps of specific biomarkers in MRI. In comparison to qualitative MRI, it enables: 1- comparison of different tissues within the same individual from different locations or points in time; 2- comparison between images from different individuals; 3- achievement of disease-specificity, with studies correlating abnormalities in the absolute values of certain biomarkers and health conditions. Longitudinal relaxation times, also referred to as T1 times, are a biomarker that can be quantified using MRI and this variable can be used for tissue characterization. It stands for the time needed for the spins to realign with the main magnetic field after a perturbation is applied to the system. One method to determine T1 is by tracing the transition from an initial state to the steady state of a spoiled gradient echo sequence (GRE). If several inversion preparation pulses are used, waiting periods should be introduced before each inversion, to allow the reestablishment of the equilibrium magnetization. That leads to long measurement times, making this type of acquisition detrimental for patient and clinical routine applications. Here a new method that excludes the waiting intervals and measurement repetitions by combining an innovative acquisition scheme with an iterative model-based reconstruction algorithm is proposed. It combines a single-shot inversion recovery (IR) with a non-IR-prepared acquisition without waiting intervals. Shorter acquisition times can be achieved using a single IR preparation combined with an undersampled spoiled GRE, spiral trajectories, and simultaneous multi-slice excitation. The experiments were performed on 3T Siemens Magnetom PRISMA scanner in healthy volunteers and the acquisition parameters were: TE: 1.1 ms, TR: 7.5 ms, slice thickness: 5 mm, spatial resolution: 0.8 x 0.8 mm². Combination of the information from non-IR prepared and IR prepared magnetization allows for the calculation of quantitative T1 maps by using an iterative model-based reconstruction scheme. In this algorithm, multiple images are reconstructed for different time points along the T1-recovery process, leading to T1 determination keeping reasonably short total acquisition times (< 1 minute).