Font Size: a A A

Study On The Non-invasive Quantitative Pharmaceutical Analysis With Near-Infrared Spectroscopy Technique

Posted on:2008-06-24Degree:MasterType:Thesis
Country:ChinaCandidate:L LuoFull Text:PDF
GTID:2144360215490407Subject:Biomedical engineering
Abstract/Summary:PDF Full Text Request
The near-infrared(NIR) light is the electromagnetic wave between the visible and mid-infrared light. The NIR region covers the range from 780nm to 2500 nm. The qualitative and quantitative analyses of NIR must have the aid of chemometrics and computer because the NIR spctra are much weaker than the fundamental vibrations absorption encountered in the mid-IR region and the absorption of overtones and combinations are overlapping seriously. NIR is advantageous because quick and accurate measurements can often be made with out destroying the sample.Usually the drug concentration in blood is mesured with method for the blood drawed out from vein at present. It brings pain, and is dufficult for the continous and homely measurement. Significant efforts have been devoted to develop the method and technique for noninvasive blood drug measurement. But all the work is in its basic studying stage. Because NIR can penetrate through skin and body tissue, and there is correlation between the concentration of drug and its NIR absorption, it is much widely used to measure the concentration of blood drug than other methods.In this paper, the near-infrared spectrum of theophylline and carbamazepine admixture is measured first. Then calibrated model for vary concentration of sample is made using multiple linear regression method, principal component analysis method, partial least squares regression method and artificial neural network method. It is concluded that artificial neural network combined with principal component analysis method has received the best result.
Keywords/Search Tags:near-infrared spectroscopy, noninvasive measurement, multiple linear regression, principal component analysis, partial least squares regression, artificial neural network
PDF Full Text Request
Related items