Simultaneous analysis of moisture, active component and cake structure of lyophilized powder for injection with diffuse reflectance FT-NIR chemometrics

Abstract

This paper establishes a novel method for the simultaneous analysis of moisture, active component and cake structure of lyophilized powder for injection using diffuse reflectance Fourier transform near infrared (FT-NIR) chemometrics. The experiment indicated that the back-propagation artificial neural network (BP-ANN) was suitable for the content predictions of moisture and active component; the root mean square errors of prediction (RMSEPs) were 0.1471 and 0.0082, the correlation coefficients (Rs) of prediction 0.9553 and 0.9891. And the self-organizing map (SOM) was adapted to the discrimination of cake structures; the prediction accuracy was 100.0%.