Figure 1.

Schematic of the trace‐based correction method. A, Magnitude and phase images of a single sagittal slice are acquired with a FLASH sequence (data shown for subject 4). The phase images are here shown at a peak of expiration ($\mathrm{\Delta }{B}_{\mathit{ex}}$) and inspiration ($\mathrm{\Delta }{B}_{\mathit{in}}$). B, The coupling parameter ($\mathrm{\Delta }{B}_{\mathit{ref}}(z)$) between the respiratory trace and the field offset ($\mathrm{\Delta }{B}_0(z,t)$) inside the spinal cord is determined for each axial slice based on the FLASH calibration data. The residual field offsets ($\mathrm{\Delta }{B}_{\mathit{res}}(z,t)$) show that the linear model explains a large part of the measured temporal field variations. C, During later scans, the field offset ($\mathrm{\Delta }{\widehat{B}}_0(z,t)$) can be estimated based on $\mathrm{\Delta }{B}_{\mathit{ref}}(z)$ and the respiratory trace. The color scale to indicate vertebral level is the same as in B. D, The estimated field offset yields corresponding phase evolutions ($\mathrm{\Delta }{\widehat{\varphi }}_0(z,t)$), here shown for a multi‐echo gradient‐echo readout at 5 different time points indicated by vertical lines in the plot in C. The acquired image data are demodulated by the estimated $\mathrm{\Delta }{\widehat{\varphi }}_0(z,t)$