Distinguishing normal from rejecting renal allografts: application of a three-stage classification strategy to MR and IR spectra of urine

We describe a three-stage classification strategy, specifically designed to classify infrared (IR) and/or nuclear magnetic resonance (MR) spectra of biofluids (F) and tissues (T). We apply the strategy to distinguish between normal renal transplants and rejected allografts, based on the MR and/or IR spectra of urine. The strategy produces accurate classifiers, thus promising the possibility of a rapid, non-invasive detection of graft pathology. (C) 2002 Elsevier Science B.V. All rights reserved.