NEAR-OPTIMAL REGION SELECTION FOR FEATURE SPACE REDUCTION - NOVEL PREPROCESSING METHODS FOR CLASSIFYING MR SPECTRA

We introduce a global feature extraction method specifically designed to preprocess magnetic resonance spectra of biomedical origin. Such pr eprocessing is essential for the accurate and reliable classification of diseases or disease stages manifest in the spectra. The new method is genetic algorithm-guided. it is compared with our enhanced version of the standard forward selection algorithm. Both seek and select opti mal spectral subregions. These subregions necessarily retain spectral information, thus aiding the eventual identification of the biochemist ry of disease presence and progression. The power of the methods is de monstrated on two biomedical examples: the discrimination between meni ngioma and astrocytoma in brain tissue biopsies, and the classificatio n of colorectal biopsies into normal and tumour classes. Both preproce ssing methods lead to classification accuracies over 97% for the two e xamples. (C) 1998 John Wiley & Sons, Ltd.