SLIM revisited.

SLIM (spectroscopic localization by imaging), a tool for in vivo magnetic resonance spectroscopy that provides compartmental spectra from relatively few phase-encoded data, is limited by the inhomogeneities within each compartment. It is shown that the SLIM-derived spectra approach the average compartmental spectra as the number of phase encoded data approaches infinity, and the convergent rate of the error for each compartment is proportional to the size of the compartment. Numerical simulation studies are conducted to confirm the convergence property. A detailed discussion of the result is provided.