A diagnostic method for stage determination of chronic kidney diseases using Fourier transform infrared spectroscopy

Abstract

Background: Fourier transform infrared spectroscopy (FTIR ) combined with chemometric techniques have been employed for qualitative analysis of proteins in blood serum. Objective: The aim of this study was to suggest a diagnostic method for stage determination of chronic kidney disease (CKD) . Methods: This laboratory study was conducted in Faculty of Science of Imam Khomeini International University during 2010-2011. 50 samples were evaluated using Fourier transform infrared spectroscopy. 0.5 ml of blood serum was poured in FTIR cell for each sample. The samples were from different stages of CKD and their stages were verified by Jaffe method as gold standard (Distribution of stages was as follows: 9 stage I/ healthy, 7 stage II, 11 stage III, 8 stage IV, 5 stage V and 10 stage VI). The samples were evaluated by Quadratic discriminant analysis (QDA). Data were modified with wavelengths selection algorithms using the Successive Projection Algorithm ( SPA ) and the Genetic Algorithm ( GA ) and were applied for prediction of serum samples status with different stages of CKD. Findings: Quadratic discriminant analysis together with successive projection algorithm and genetic algorithm as wavelength selection strategies are suitable for prediction of CKDs. The correct classification rates of the samples were 85% and 97% for SPA and GA models, respectively. Conclusion: With regards to the results, this new diagnostic method can be used for multi-stage classification of the chronic kidney disease as an improved method.