Studies On Nondestructive Quantitative Pharmaceutical Analysis Using NIR Spectroscopy With Genetic Algorithm

In recent years, near infrared (NIR) spectroscopy has made rapid progress as a powerful analytical tool. NIR spectroscopy analysis techniques is an efficient, simple, nondestructive and no contamination method that has been used in chemical analysis in diverse fields, such as agriculture, food, petrochemical, textile and pharmaceutical industries etc. NIR has been widely used in pharmaceutical analysis, and it's widely prospect in this field is attractive. The NIR spectral region is generally defined as the wavelength range from 780nm～2500nm. It is customarily divided into two ranges, short-wavelength NIR spectroscopy (780nm～1100nm) and long-wavelength NIR spectroscopy (1100nm～2500nm). The spectra in this region are dominated by the absorption for overtones and combinations of fundamental vibrational modes of O-H,C-H and N-H groups in compound. However, NIR spectroscopy has some disadvantages resulting in weak, partly overlapped, non-specific bands, so it is hard to interpret of the data only depending on the conventional analytical methods. In order to obtain the exact result in the qualitative and quantitative analysis, some data processing methods must be applied. The modern NIR technique is the outcome of NIR spectroscopy combined with the chemometrics method. Combined with the strong information process ability of chemometrics, NIR spectroscopy can fully demonstrate its rapidness, precision and nondestructive analysis which has great potential in these application.The main research content of this paper is nondestructive quantitative pharmaceutical analysis using genetic algorithm on NIR spectroscopy. It includes three stages. The first stage is pretreating of spectra and deleting noisy signals from spectra. The methods include first-derivative, second-derivative, multiplicative scatter correction (MSC) and standard normal variate (SVN). The second stage is variables selection and removing noise and random errors from the original spectra. The methods include the principal component analysis (PC), uninformative variable elimination (UVE), mutual information (MI) and genetic algorithm (GA). The last stage is multivariate calibration. The methods include partial least-squares (PLS) and artificial neural nets (ANNs). A modified genetic algorithm with fixed number of select variables was developed in this paper. Modified GA reduces the number of input variables to ANNs and reduces training time, which is a favorable objective as it allows more networks design to be evaluated in a given time. Modified GA is used to select fixed number of variables that are used as the input variables of ANNs, which can decrease the effect of structure of ANNs and increase the convergence capacity of GA-ANNs model.First, different variables selection methods were used for quantitative analysis of cefalexin powdered drug. PC, UVE, MI and modified GA methods is used to variables selection, and these selected variables are used to set up PLS and ANNs models. In the paper, the methods for selecting factor number of PLS and number of hidden nodes of ANNs are given. It illuminates that PLS calibration results is poor due to NIR spectra overlapped bands, complicated absorptance, and nonlinearity. On the contrary, ANNs method can result in the robust models due to its anti-jamming, anti-noise and nonlinear transfer ability. ANN method has the superiority in quantitative analysis of complicated drugs. The modified GA coupled to ANNs gets the best result, so modified GA is super to other variables selection methods.Then different pretreating methods were used for quantitative analysis of trimethoprim powdered drug. First-derivative, second-derivative, MSC and SNV methods are used to pretreat original spectra, modified GA-ANNs method is used to set up models and giving results of test set. Original spectra contain not only the information from the components of the samples, but many noises from any aspects. These noisy signals can interfere in spectral information. In this paper original spectra give the best result, so original spectra are super to other pretreating methodsBased on the method discussed former, modified GA-ANNs is used for quantitative analysis of compound cefalexin and trimethoprim powdered drug. Modified GA-ANNs is used to select variables from original spectra and ANNs is used to set up models. The model gives good results. This method is a robust mothed due to its anti-jamming, anti-noise and nonlinear transfer ability.The method of modified GA-ANNs on NIR spectroscopy is explored for nondestructive quantitative analysis of solid pharmaceutical samples. It is feasible to obtain quantitative information by processing NIR spectroscopy with modified GA-ANNs models. Modified GA-ANNs must have been expansive foreground and applied values.