Sr. Amendolia et al., CLASSIFICATION AND QUANTITATION OF H-1-NMR SPECTRA OF ALDITOLS BINARY-MIXTURES USING ARTIFICIAL NEURAL NETWORKS, Analytical chemistry, 70(7), 1998, pp. 1249-1254
A pattern recognition method based on artificial neural networks (ANNs
) to analyze and quantify the components of six alditol binary mixture
s is presented, This method is suitable to classify the spectra of the
15 mixtures obtained from the six alditols and to produce quantitativ
e estimates of the component concentrations. The system is user-friend
ly and is helpful in solving the problem of greatly overlapping signal
s, often encountered in NMR spectroscopy of carbohydrates. A ''classif
ication'' ANN uses 200 intensity values of the H-1 NMR spectrum in the
range 3.5-4 ppm, When the correct mixture is identified, the quantifi
cation is solved by assigning a specific ANN to each mixture, These AN
Ns use the same 200 values of the spectrum and output the values of th
e two concentrations. The error in the ANN responses is studied, and a
method is developed to estimate the accuracy in determining the conce
ntrations. The networks' abilities to recognize previously unseen mixt
ures are tested. When the classification ANN (trained on the 15 binary
mixtures) is exposed to complex (i.e., more than binary) mixtures of
the six known alditols, it successfully identifies the components if t
heir minimum concentration is 10%. Given the precision of the results
and the small number of errors reported, we believe that the method ca
n be used in all fields in which the recognition and quantification of
components are necessary.