Signal-to-Noise Improvement in in Vivo Spin-Echo Spectroscopy in the Presence of Motion

Abstract In the presence of macroscopic motion, spin-echo techniques suffer from signal losses due to motion-induced phase shifts. Several simple and efficient techniques for reducing losses are presented. They are based on coaddition of phased, modulus, or power spectra, instead of coaddition of FIDs. The advantages and drawbacks of each method are analyzed: For it to be performed, phasing of individual spectra requires a sufficiently high signal-to-noise ratio at each scan, at least for one peak. Coaddition of modulus or power spectra does not have this limitation. However coaddition of modulus spectra introduces a baseline artifact. This prevents peaks smaller than the rms value of the noise on individual scans to be observed. Furthermore this mode is nonlinear for low-intensity signals. In contrast, coaddition of power spectra is a linear mode with respect to the peak intensities. The drawback is that, for peaks smaller than the rms value of the noise on individual scans, the noise level depends on the signal-to-noise ratio. These methods have been shown to increase the signal-to-noise ratio in in vivo1H spectroscopy of the rat brain. Furthermore utilization of these techniques greatly improve the accuracy of T2 measurements and reduce t1, noise in in vivo 2D experiments.