Abstract

OBJECTIVES
The determination of oxalate in urine is required for the diagnosis and treatment of primary hyperoxaluria, idiopathic stone disease and various intestinal diseases. We examined the possibility of using Fourier transform near-infrared (FT-NIR) spectroscopy analysis to quantitate urinary oxalate. The practical advantages of this method include ease of the sample preparation and operation technique, the absence of sample pre-treatments, rapid determination and noninvasiveness. METHODS: The range of oxalate concentration in standard urine solutions was 0-221 mg/l. These 80 different samples were scanned in the region of 780-1,300 nm with a 0.5 nm data interval by a Spectrum One NTS FT-NIR spectrometer. PCR, PLSR and MLR regression models were used to calculate and evaluate the calibration equation. RESULTS: The PCR and PLSR calibration models were obtained from the spectral data and they are exactly same. The standard error of estimation (SEE) and the % variance were 10.34 mg/l and 97.86%, respectively. After full cross validation of this model, the standard error of estimation was 5.278 mg/l, which was much smaller than that of the pre-validation. Furthermore, the MCC (multiple correlation coefficient) was 0.998, which was compatible with the 0.923 or 0.999 obtained from the previous enzymatic methods. CONCLUSIONS: These results showed that FT-NIR spectroscopy can be used for rapid determination of the concentration of oxalate in human urine samples.

• Keywords: UrineOxalateNIR spectroscopyRegression analysis