Wiener filter methods for baseline correction of magnetic resonance spectra

The problem of correcting the baseline distortion of signals measured in magnetic resonance spectroscopy is investigated in the context of Wiener filter theory. Based on the estimation of model parameters describing the spectrum, a method is proposed that estimates the correlation operators of the net signal and the baseline distortion. These operators are used to construct a Wiener filter for the measured spectrum. Results of numerical simulations that compare the proposed method with other methods are given.