Browse views: by Year, by Function, by GLF, by Subfunction, by Conference, by Journal

Global regression model for moisture content determination using near-infrared spectroscopy

Roggo, Yves (2017) Global regression model for moisture content determination using near-infrared spectroscopy. European journal of pharmaceutics and biopharmaceutics. ISSN 1873-3441; 0939-6411

Abstract

Near-infrared (NIR) global quantitative models were evaluated for the moisture content (MC) determination of three different freeze-dried drug products. The quantitative models were based on 3822 spectra measured on two identical spectrometers to include variability. The MC, measured with the reference Karl Fischer (KF) method, were ranged from 0.05% to 4.96%. Linear and non-linear regression models using Partial Least Square (PLS), Decision Tree (DT), Bayesian Ridge Regression (Bayes-RR), K-Nearest Neighbors (KNN), and Support Vector Regression (SVR) algorithms were created and evaluated. Among them, the SVR model was retained for a global application. The Standard Error of Calibration (SEC) and the Standard Error of Prediction (SEP) were respectively 0.12% and 0.15%. This model was then evaluated in terms of total error and risk-based assessment, linearity, and accuracy. It was observed that MC can be fastly and simultaneously determined in freeze-dried pharmaceutical products thanks to a global NIR model created with different medicines. This innovative approach allows to speed up the validation time and the in-lab release analyses.

Item Type: Article
Date Deposited: 10 Jul 2018 00:45
Last Modified: 10 Jul 2018 00:45
URI: https://oak.novartis.com/id/eprint/36927

Search

Email Alerts

Register with OAK to receive email alerts for saved searches.