Synopsis

MR Imaging at low field (LF) suffers from low SNR due to intrinsically low Boltzmann polarization. As a result, long acquisition times are needed to accommodate the additional signal averaging required to attain sufficient SNR. In the aim of improving SNR at LF, efficient pulse sequences, efficient radiofrequency (RF) coil designs and new image reconstruction methods are required. In this talk, we will discuss how refocused pulse sequences and have enabled high-quality LF imaging. On the reconstruction side, a noise-robust image reconstruction approach AUTOMAP was implemented with a deep neural network architecture so as to further increase sensitivity at LF.