A NOVEL DIAGNOSTIC-TEST FOR ARTHRITIS - MULTIVARIATE-ANALYSIS OF INFRARED-SPECTRA OF SYNOVIAL-FLUID

Fourier transform infrared spectroscopy has been applied to the invest igation of synovial fluids (SFs) aspirated from arthritic joints. Sign ificant differences, related to differences in the composition of the fluid as a result of the disease processes, were found between spectra of SFs from joints affected by rheumatoid arthritis, osteoarthritis, spondyloarthropathies, and meniscal injuries. Linear discriminant anal ysis with leave-one-out cross validation was used to classify 239 SF f ilm spectra obtained from 86 patients. Using a patient-based approach, in which the consensus of results obtained from three spectra of each fluid was taken as the diagnosis, multivariate analysis successfully classified spectra into four classes, in excellent agreement with clin ical diagnosis (96.5% correct classification). These results demonstra te that when combined with a properly trained classifier, infrared spe ctra of SF films can be used as an aid in the diagnosis of arthritic d isorders. (C) 1997 John Wiley & Sons, Inc.