GENETIC PROGRAMMING FOR CLASSIFICATION AND FEATURE-SELECTION - ANALYSIS OF H-1 NUCLEAR-MAGNETIC-RESONANCE SPECTRA FROM HUMAN BRAIN-TUMOR BIOPSIES

Genetic programming (GP) is used to classify tumours based on H-1 nucl ear magnetic resonance (NMR) spectra of biopsy extracts. Analysis of s uch data would ideally give not only a classification result but also indicate which parts of the spectra are driving the classification (i. e. feature selection). Experiments on a database of variables derived from 1H NMR spectra from human brain tumour extracts (n = 75) are repo rted, showing GP's classification abilities and comparing them with th at of a neural network. GP successfully classified the data into menin gioma and non-meningioma classes. The advantage over the neural networ k method was that it made use of simple combinations of a small group of metabolites, in particular glutamine, glutamate and alanine. This m ay help in the choice of the most informative NMR spectroscopy methods for future non-invasive studies in patients. (C) 1998 John Wiley & So ns, Ltd.