Preliminary Study On Quantitative Method Of Oral Solid Dosage Forms By Near Infrared Spectra

Objective The study explores simultaneous quantification of near infrared spectrum for accessories in the complicated analysis system of solid pharmaceutical preparation and tries to break down initial bottlenecks that multiple accessories are hard to quantify simultaneously in implementation of consistency evaluation and have no suitable analysis methods for generic drugs caused by shortage of existing test standards in domestic accessories,difficult preparation of solid pharmaceutical adjuncts for analyzing and detecting suitable solutions.In addition,it further discusses spectrum pre-treatment,wave-length choice,model optimization and confirmation of near-infrared quantitative analysis of several drugs in consistency evaluation of generic drugs,so as to provide rapid detection of pharmaceutical adjuncts and provide practical and feasible analysis methods for consistency evaluation of solid generic drugs.Methods(1)Combine the near infrared spectrum with chemometrics to detect the content of Amlodipine Besylate Tablets excipient rapidly.Select characteristic wavelength variation points through random frog method,variable importance in the projection and stability competitive adaptive reweighted sampling,adopt 9spectrum pre-processing methods for processing the original spectrum,and then establish the partial least squares model and support vector regression analytical model respectively,and then compare these two models.Use the optimally selected model to test the samples.(2)In order to improve the performance of the quantitative model of near infrared spectrum,the study combine the Monte Carlo uninformative variable elimination with the genetic algorithm(MCUVE-GA),to select the glipizide pharmaceutical excipient spectral variables,and establish the principal component analysis(PCA)and Artificial Neural Network Algorithm(ANN)model,and then make comparative analysis of the model performance and the partial least squares(PLS)model,to select out the modeling method.(3)It probes into the modeling approach of pharmaceutical excipient content in the fast and nondestructive testing valsartan capsules,and use near infrared spectrum method combined the interval partial least squares(iPLS),backward interval partial least squares(BiPLS),synergic interval partial least squares(SiPLS)respectively for modeling,and then compare the impact of each algorithm with different interval division number and interval selection.(4)It uses near infrared diffuse reflectance analysis combined with partial least squares to establish the quantitative analysis model of domperidone pharmaceutical excipient.It constantly optimizes the model through researching the spectral region selection,best primary factor number selection and other aspects,to confirm the best modeling parameters finally.Results(1)The results of using near infrared diffuse reflection spectrometry to test the Amlodipine Besylate Tablets excipient show that: for the samples involved,random frog method effect is the best in the optimal characteristic wavelength variable selection;in the model forecasted results,compared with the support vector regression analytical model,there will be good effect when using the 5 indexes of partial least squares quantitative model to forecast the root-mean-square error evaluation parameter,and the relative analysis error values are all bigger than 3.0.The standard error of the test value of the sample and the measured value is less than 1.30.The paired t test shows that there is no significant difference between the two groups at the significance level of α = 0.05.(2)Near-infrared spectroscopy for the determination of glipizide tablets showed that the evaluation parameters of the partial least squares(PLS)model of starch,dextrin and magnesium stearate were superior to those of principal component analysis(PCA)and artificial Neural network algorithm(ANN)model.The sucrose was trained by the conjugate gradient learning algorithm to obtain better performance parameters than the PLS model.Pairwise t test showed that there was no significant difference between the two groups at the significant level of α = 0.05.(3)The results of using near infrared spectrum method to test the valsartan capsules excipient show that: the BiPLS model prediction precision of excipient microcrystalline cellulose and sodium carboxymethyl starch is better than the iPLS and SiPLS model precision,the iPLS model prediction precision of excipient povidone and sodium dodecyl sulfate is better than the BPLS and SiPLS model precision.Pairwise t test showed that there was no significant difference between the two groups at the significant level of α = 0.05.(4)The performance of PLS quantitative analysis model of the dopantrione tablets was good,and the validation set correlation coefficient and the predicted root mean square error were respectively 0.9657,1.29;0.9870,0.877;0.9734,0.688;0.9474,0.734;0.9303,0.880;0.9777,0.0495.Conclusions The Near infrared diffuse reflectance spectroscopy combined with chemometrics can be used to detect the content of Amlodipine Besylate Tablets,Glipizide Tablets,valsartan capsules and Domperidone Tablets rapidly.Through selecting 6 groups of sample not involved in modeling to confirm the validity the the model,and paired t test analysis was made to determine that there was no significant difference between the results of nearinfrared measurement and the measured value.The near infrared diffuse reflectance spectroscopy can be used for the rapid detection of the excipients′contents in amlodipine besylate tablets,with the advantages of being operated simply and rapidly,green,accurate and reliable tested results,and good repeatability,stability,linearity,accuracy,which may provide a reference for the rapid detection of other pharmaceutical excipients′contents.In practical application,by increasing the sample volume of calibration set and prediction set to further optimize,validate and perfect the model,its applicability and reliability can be continuously improved,it can meet the needs of actual production better,and has important guiding significance for on-line detection;furthermore,it can further expand the application of near infrared spectroscopy,and it’s expected to give power to the consistency evaluation of generic drugs.