A comparison of selective saturation and selective echo chemical shift imaging techniques.

We report on a comparative study of two methods of chemical shift imaging which can be used to selectively image fat and water in vivo. Both methods require a B0 field sufficiently homogenous to resolve the methylene and water spectral lines. One method, called CHESS, uses a chemically selective pulse to saturate the unwanted spectral line. The other method, called SECSI, achieves chemical selectivity by using a soft 180 degree RF pulse in forming a spin echo image. Both methods require that the strength of the B1 RF field be accurately calibrated and homogenous. We show by theoretical analysis that the suppression of unwanted spectral lines is sensitive to the first power of B1 errors in the CHESS method but to the second power of B1 errors in the SECSI method. Experiments with phantoms confirmed the expected non linearity of the SECSI method, and showed superior water suppression factors in phantoms with it. Experiments with a large phantom and a living rabbit showed superior results using the SECSI method, and the best results were obtained using a combination of the two techniques.