Abstract: In parallel magnetic resonance imaging (MRI)， the signal to noise ratio (SNR) of reconstruction image would be obviously reduced under the high acceleration factors because of the illposed problem in the process of sensitivity encoding (SENSE) reconstruction. Through indepth analysis of total variation (TV) regularized SENSE reconstruction model， an efficient and fast split Bregman iteration algorithm was introduced to obtain the optimal solution and effectively improve the image reconstruction results. The simulation experiments were carried on the phantom data and brain data of MRI， respectively. The experimental results demonstrated that compared with the traditional TV regularized SENSE reconstruction algorithm， the proposed algorithm not only has fewer iterations and faster convergence speed， but also can alleviate the aliasing artifacts， significantly improves the SNR and decreases the normalized mean squared error of reconstruction image.

Key words: sensitivity encoding (SENSE), magnetic resonance image reconstruction, total variation regularization, artificial time marching method, split Bregman iteration