An Automatic Abdomen MR-Quantification based on a Simultaneous Multi-Parameter Mapping within 3 Breath-holds and deep-learning Segmentation

Abstract

For automatic abdomen quantification, a rapid and simultaneous multi-parameter mapping method was evaluated, which could acquire quantitative R1, R2* and PDFF indices within 3-4 breath-holds. By adopting automatic organ segmentation, it could directly generate MR quantitative information for liver, kidney and spleen, which could be potentially used for liver pathology evaluation. The result showed that the multi-parameter sequence could generate high-quality R1, R2* and PDFF maps, with well B1 field correction and fat signal separation.