next up previous
Next: Cost Function Formulation Up: Theory Previous: Theory

Problem Definition

Let the measured (wrapped) phase be $\phi_j$ where $j$ is an n-dimensional index specifying the spatial location (or voxel). The objective is to find the unwrapped phase denoted $\theta_j$ such that:

\begin{displaymath}
\theta_j = \phi_j + 2\pi m_j
\end{displaymath} (1)

where the integer value $m_j$ at each voxel specifies the corrective offset required. The unwrapped phase should be smooth and not contain the $2\pi$ jumps that are typically seen in the wrapped phase.



Mark Jenkinson 2001-10-12