Magnetic resonance spectroscopic imaging (MRSI) is an imaging modality used for studying tissues in-vivo in order to assess and quantify metabolites for diagnostic purposes. However, long scanning times, low spatial resolution, poor signal-to-noise ratio (SNR) and the subsequent noise-sensitive non-linear model fitting are major roadblocks in accurately quantifying the metabolite concentration. In this work, we propose a spectrally adaptive method for denoising of MRSI spectra based on the concept of Non-Local Means (NLM)1,2 which relies on spectral redundancy in the given data-set. Due to lack of phase information in the acquired data, we perform a ‘spectral dephasing’ step to consider a wide range of phase variations to improve the NLM-based pattern analysis. We evaluated our method on human in-vivo data and assessed the improvement in SNR for NAA.     «

Magnetic resonance spectroscopic imaging (MRSI) is an imaging modality used for studying tissues in-vivo in order to assess and quantify metabolites for diagnostic purposes. However, long scanning times, low spatial resolution, poor signal-to-noise ratio (SNR) and the subsequent noise-sensitive non-linear model fitting are major roadblocks in accurately quantifying the metabolite concentration. In this work, we propose a spectrally adaptive method for denoising of MRSI spectra based on the conce...     »