One novel parameters estimation method for magnetic resonance spectroscopy based on decimation and prior knowledge of signal poles

Parameter estimation of Magnetic Resonance Spectroscopy (MRS) signal is a key step for the application of MRS. To the best of our knowledge, there are no parameter estimation methods that are based on decimation in combination with prior knowledge to estimate the parameters of MRS signals. This paper puts forward a novel method for parameter estimation of MRS signal, which takes advantage of prior knowledge of signal poles to improve the accuracy of quantification. At the same time decimation is exploited to further increase the performance of subspace-based parameter estimation in the case of oversampling. The theory analysis and experiment results show that in terms of computational complexity and accuracy the novel method performs much better than the commonly used methods such as HTLS and HTLSPK.