Signal correction in magnetic resonance spectroscopy

Two methods of adaptive, non-linear filtering in Fourier- and wavelet-space are presented and applied to the correction of baseline and phase distortion of magnetic resonance spectroscopic signals. Both methods consist of estimating the model parameters describing the signal and constructing a multiplicative filter based on these estimations. While the Fourier filter method can be used for baseline correction only, the wavelet-based method is able to perform both baseline and phase correction in a single step.