A new approach to improving automated analysis of proton NMR spectra through global spectral deconvolution (GSD)

GSD is automatically reduces a frequency domain spectrum to a set of Lorentzian or near Lorentzian lines leaving out any baseline drift and noise. A reliable calculation of the derivatives is especially important, particularly in spectra with poor signal-to-noise ratio. Another tricky problem in the case of water is that it does not only show up as a singlet but also as a more complex multiplet presenting a convoluted fine structure. When the solvent presents a clear multiplet structure, like the quintuplet in DM50, and is matched during the evaluation process, it results in a good premium in the multiplet pattern recognition test of the scoring system. The combination of a Bayesian algorithm with GSD and the automatic detection of solvent and reference peaks proved to be very effective for the automatic identification of the correct integration values.