Investigation On Quantitative Analysis Of Urinary Calculus Components Based On Infrared Spectroscopy

Objective Urinary lithiasis as a global disease with complex etiopathogenisis and mechanism is easy to recur.So it is crucial to qualitify and quantify urinary calculus constituents to cure and prevent recuring.In our study,we apply partial least-squares (PLS)to analyze infrared spectroscopy(IR)of urinary calculus through multiplet linear regression analysis.In order to seek to the feasibility of quantitative assessment of calculus composition through PLS.And providing strong evidence to clinical treatment.Materials and Methods:We use analytical pure of whewellite and Hydroxyl apatite and weddellite produced to allocate sixty-six samples per different proportion.The allocating principle is ensuring proportion of every sample uniform and no bias.Then we collect twenty-five urinary calculus,which qualified by IR and certificated to be binary or triangle mixtures containing these three constituents.And the content of every composition is measured by wet chemical procession in advance. Collecting all IR and then converting them to data spectrums,we apply PLS to analyze those data spectrums to do quantitative assessment of all calculus composition.Results:The mixtures of different content of every composition has different IR.The concentrations of oxalic ion,calcium,phosphate radical of urinary calculus and their IR are correlating.Using the predictive contents of constituents of every mixture and their ture contents to make correlation diagram.The correlation coefficient of the predictive concentrations of oxalic ion,calcium,phosphate radical and their true value are0.9472,0.9577,0.9483 respectively,closing to 1.Standard deviations are 9.3261,5.2362,10.2186,which close to their root mean square error of prediction (RMSEP).Conclusions:The method of use IR to quantify the compositions of urinary calculus is convenient,quick and less samples.And the same time applying PLS regression to model.Compared with others LS,PLS model gives more stable and accurate prediction results.Through an investigation of parameters involved in the modeling, it is demonstrated that applying IR to quantify composition of urinary calculus is a reliable and feasible method.