Steady spectra of supreme resolution and lowest noise in high-order optimized derivative fast Fourier transform for ovarian NMR spectroscopy

Abstract

The optimized derivative fast Fourier transform (dFFT) simultaneously increases resolution and reduces noise in spectra reconstructed from encoded time signals. The pertinent applications have recently been published for time signals encoded with and without water suppression by in vitro and in vivo magnetic resonance spectroscopy (MRS). Even with the employed lower derivative orders, genuine resonances were narrowed, their intensities enhanced and the background baselines flattened. This unequivocally separated many overlapped peaks that are the thorniest problem in data analysis by signal processing. However, it has been common knowledge that higher-order derivative spectra quickly deteriorate with the increased derivative order. The optimized dFFT can challenge such findings. An unprecedented resilience of this processor to derivative-induced distortions is presently demonstrated for high derivative orders (up to 20). The salient illustrations are given for the water residual, lactate quartet and lactate doublet alongside their close surroundings. These applications of diagnostic relevance for patients with cancer are reported for time signals encoded with water suppression by in vitro proton MRS of human ovary.