Fig. 6.

Image quality as a function of reconstruction method in a volunteer study. (a) Axial slice for (left) NVS-PEVS, (center) ESP-PEVS, and (right) RSGR-PEVS. Images normalized by maximum value of reconstructed intensity and displayed using equal window/level settings. (b) Mean rESF across breathing phases at lung-liver interface as a function of reconstruction technique. In regions with large temporal changes RSGR-PEVS provides a significant improvement in spatial resolution while ESP-PEVS does not improve resolution. Bar-plots depict average ± standard error of mean. (c) Displacement of lung-liver interface as a function of breathing phase. (d) Effect of reconstruction technique on noise. The noise floor (σ) is estimated by the standard deviation of signal intensity (across phases) in a large background region with no NMR signal.