摘要
目的:建立三磷酸腺苷二钠片通用型定性模型,用于药品现场快速筛查。方法:采集全国12个生产企业共60批样品的近红外光谱,结合化学计量学最近邻法建立两层定性分析模型,并对模型专属性、耐用性进行验证。结果:优化后的定性判别模型在三种建模方案中专属性好,能准确识别16批验证集样品,区分出4批结构类似物;该模型标准差(SDev)、平均光谱距离(MD)与阈值均更小。结论:本文建立的模型快速、准确、操作简便,可在药品快速检测车上应用。
Objective： To establish a general qualitative model for the rapid on-site identification of adenosine triphosphate disodium tablets. Methods： Near infrared-spectra （NIRS） of 60 batches of samples from 12 manufacturers were collected to establish the two-layer qualitative model in combination with the nearest-neighbor method. The model was validated for specificity and durability. Results： The optimized qualitative model showed excellent specificity which accurately identified the 16 batches of testing samples and distinguished 4 batches of structural analogues. The standard deviation （SD_ev）, the average spectral distance and the threshold of the model were smaller compared to other modeling schemes. Conclusion： This mode! is rapid, accurate and easy to operate, which is applicable in drug testing vehicles for drug rapid screening.
出处
《中国药事》
CAS
2015年第3期309-312,共4页
Chinese Pharmaceutical Affairs
关键词
药品快速检测车
近红外定性模型
最近邻法
因子分析
三磷酸腺苷二钠片
现场快速筛查
drug rapid testing vehicles
near-infrared qualitative model
nearest-neighbor method
factorization
adenosine triphosphate disodium tablets, on-site rapid screening
