loading page

Calibration transfer for bioprocess Raman monitoring using Kennard Stone Piecewise Direct Standardization and multivariate algorithms
  • +7
  • Laure Pétillot,
  • Fiona Pewny,
  • Martin Wolf,
  • Célia Sanchez,
  • Fabrice Thomas,
  • Johan Sarrazin,
  • Katharina Fauland,
  • Herman Katinger,
  • Charlotte Javalet,
  • Christophe Bonneville
Laure Pétillot
RESOLUTION Spectra Systems

Corresponding Author:petillot.laure@gmail.com

Author Profile
Fiona Pewny
Polymun Scientific
Author Profile
Martin Wolf
Polymun Scientific
Author Profile
Célia Sanchez
RESOLUTION Spectra Systems
Author Profile
Fabrice Thomas
RESOLUTION Spectra Systems
Author Profile
Johan Sarrazin
RESOLUTION Spectra Systems
Author Profile
Katharina Fauland
Polymun Scientific
Author Profile
Herman Katinger
Polymun Scientific
Author Profile
Charlotte Javalet
RESOLUTION Spectra Systems
Author Profile
Christophe Bonneville
RESOLUTION Spectra Systems
Author Profile

Abstract

In the biopharmaceutical industry, Raman spectroscopy is now a proven PAT tool that enables in-line simultaneous monitoring of several CPPs and CQAs in real-time. However, as Raman monitoring requires multivariate modeling, variabilities unknown by models can impact the monitoring prediction accuracy. With the widespread use of Raman PAT tools, it is necessary to fix instrumental variability impacts, encountered for instance during a device replacement. In this work, we investigated the impact of instrumental variability between probes inside a multi-channel analyzer and between two analyzers, and explored solutions to correct them on model prediction errors in cell cultures. We found that the Kennard Stone Piecewise Direct Standardization (KS PDS) method enables to lower model prediction errors and that only one batch with the unknown device in the calibration dataset was sufficient to correct the prediction gap induced by instrumental variability. As a matter of fact, during device replacement a first cell culture monitoring can be performed with the KS PDS method. Then, the new data obtained can be inserted in the calibration dataset to integrate instrumental variability in the chemometric model. This methodology provides good multivariate calibration model prediction errors throughout the instrumental changes.
09 Jan 2020Submitted to Engineering Reports
24 Jan 2020Submission Checks Completed
24 Jan 2020Assigned to Editor
06 Feb 2020Reviewer(s) Assigned
16 Mar 2020Editorial Decision: Revise Major
14 May 20201st Revision Received
15 May 2020Submission Checks Completed
15 May 2020Assigned to Editor
15 May 2020Reviewer(s) Assigned
08 Jun 2020Editorial Decision: Revise Minor
11 Jun 20202nd Revision Received
11 Jun 2020Submission Checks Completed
11 Jun 2020Assigned to Editor
11 Jun 2020Editorial Decision: Accept
15 Sep 2020Published in Engineering Reports. 10.1002/eng2.12230