Robust calibrations on reduced sample sets for API content prediction in tablets: Definition of a cost-effective NIR model development strategy

Sigrid Pieters, Wouter Saeys, Tom Van den Kerkhof, Mohammad Goodarzi, Mario Hellings, Thomas De Beer, Yvan Vander Heyden

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

Owing to spectral variations from other sources than the component of interest, large investments in the NIR model development may be required to obtain satisfactory and robust prediction performance. To make the NIR model development for routine active pharmaceutical ingredient (API) prediction in tablets more cost-effective, alternative modelling strategies were proposed. They used a massive amount of prior spectral information on intra- and inter-batch variation and the pure component spectra to define a clutter, i.e., the detrimental spectral information. This was subsequently used for artificial data augmentation and/or orthogonal projections. The model performance improved statistically significantly, with a 34-40% reduction in RMSEP while needing fewer model latent variables, by applying the following procedure before PLS regression: (1) augmentation of the calibration spectra with the spectral shapes from the clutter, and (2) net analyte pre-processing (NAP). The improved prediction performance was not compromised when reducing the variability in the calibration set, making exhaustive calibration unnecessary. Strong water content variations in the tablets caused frequency shifts of the API absorption signals that could not be included in the clutter. Updating the model for this kind of variation demonstrated that the completeness of the clutter is critical for the performance of these models and that the model will only be more robust for spectral variation that is not co-linear with the one from the property of interest.

Original languageEnglish (US)
Pages (from-to)62-70
Number of pages9
JournalAnalytica Chimica Acta
Volume761
DOIs
StatePublished - Jan 25 2013

Fingerprint

development strategy
Calibration
Tablets
drug
calibration
Costs and Cost Analysis
prediction
cost
Pharmaceutical Preparations
Costs
Water
water content
Water content
modeling
Processing
development model

Keywords

  • Augmentation methods
  • Near infrared spectroscopy
  • Orthogonal projection methods
  • Prior spectral information
  • Tablets

ASJC Scopus subject areas

  • Biochemistry
  • Analytical Chemistry
  • Spectroscopy
  • Environmental Chemistry

Cite this

Robust calibrations on reduced sample sets for API content prediction in tablets : Definition of a cost-effective NIR model development strategy. / Pieters, Sigrid; Saeys, Wouter; Van den Kerkhof, Tom; Goodarzi, Mohammad; Hellings, Mario; De Beer, Thomas; Heyden, Yvan Vander.

In: Analytica Chimica Acta, Vol. 761, 25.01.2013, p. 62-70.

Research output: Contribution to journalArticle

Pieters, Sigrid ; Saeys, Wouter ; Van den Kerkhof, Tom ; Goodarzi, Mohammad ; Hellings, Mario ; De Beer, Thomas ; Heyden, Yvan Vander. / Robust calibrations on reduced sample sets for API content prediction in tablets : Definition of a cost-effective NIR model development strategy. In: Analytica Chimica Acta. 2013 ; Vol. 761. pp. 62-70.
@article{2005ff5ce75b4adbaaedae95390272ad,
title = "Robust calibrations on reduced sample sets for API content prediction in tablets: Definition of a cost-effective NIR model development strategy",
abstract = "Owing to spectral variations from other sources than the component of interest, large investments in the NIR model development may be required to obtain satisfactory and robust prediction performance. To make the NIR model development for routine active pharmaceutical ingredient (API) prediction in tablets more cost-effective, alternative modelling strategies were proposed. They used a massive amount of prior spectral information on intra- and inter-batch variation and the pure component spectra to define a clutter, i.e., the detrimental spectral information. This was subsequently used for artificial data augmentation and/or orthogonal projections. The model performance improved statistically significantly, with a 34-40{\%} reduction in RMSEP while needing fewer model latent variables, by applying the following procedure before PLS regression: (1) augmentation of the calibration spectra with the spectral shapes from the clutter, and (2) net analyte pre-processing (NAP). The improved prediction performance was not compromised when reducing the variability in the calibration set, making exhaustive calibration unnecessary. Strong water content variations in the tablets caused frequency shifts of the API absorption signals that could not be included in the clutter. Updating the model for this kind of variation demonstrated that the completeness of the clutter is critical for the performance of these models and that the model will only be more robust for spectral variation that is not co-linear with the one from the property of interest.",
keywords = "Augmentation methods, Near infrared spectroscopy, Orthogonal projection methods, Prior spectral information, Tablets",
author = "Sigrid Pieters and Wouter Saeys and {Van den Kerkhof}, Tom and Mohammad Goodarzi and Mario Hellings and {De Beer}, Thomas and Heyden, {Yvan Vander}",
year = "2013",
month = "1",
day = "25",
doi = "10.1016/j.aca.2012.11.034",
language = "English (US)",
volume = "761",
pages = "62--70",
journal = "Analytica Chimica Acta",
issn = "0003-2670",
publisher = "Elsevier",

}

TY - JOUR

T1 - Robust calibrations on reduced sample sets for API content prediction in tablets

T2 - Definition of a cost-effective NIR model development strategy

AU - Pieters, Sigrid

AU - Saeys, Wouter

AU - Van den Kerkhof, Tom

AU - Goodarzi, Mohammad

AU - Hellings, Mario

AU - De Beer, Thomas

AU - Heyden, Yvan Vander

PY - 2013/1/25

Y1 - 2013/1/25

N2 - Owing to spectral variations from other sources than the component of interest, large investments in the NIR model development may be required to obtain satisfactory and robust prediction performance. To make the NIR model development for routine active pharmaceutical ingredient (API) prediction in tablets more cost-effective, alternative modelling strategies were proposed. They used a massive amount of prior spectral information on intra- and inter-batch variation and the pure component spectra to define a clutter, i.e., the detrimental spectral information. This was subsequently used for artificial data augmentation and/or orthogonal projections. The model performance improved statistically significantly, with a 34-40% reduction in RMSEP while needing fewer model latent variables, by applying the following procedure before PLS regression: (1) augmentation of the calibration spectra with the spectral shapes from the clutter, and (2) net analyte pre-processing (NAP). The improved prediction performance was not compromised when reducing the variability in the calibration set, making exhaustive calibration unnecessary. Strong water content variations in the tablets caused frequency shifts of the API absorption signals that could not be included in the clutter. Updating the model for this kind of variation demonstrated that the completeness of the clutter is critical for the performance of these models and that the model will only be more robust for spectral variation that is not co-linear with the one from the property of interest.

AB - Owing to spectral variations from other sources than the component of interest, large investments in the NIR model development may be required to obtain satisfactory and robust prediction performance. To make the NIR model development for routine active pharmaceutical ingredient (API) prediction in tablets more cost-effective, alternative modelling strategies were proposed. They used a massive amount of prior spectral information on intra- and inter-batch variation and the pure component spectra to define a clutter, i.e., the detrimental spectral information. This was subsequently used for artificial data augmentation and/or orthogonal projections. The model performance improved statistically significantly, with a 34-40% reduction in RMSEP while needing fewer model latent variables, by applying the following procedure before PLS regression: (1) augmentation of the calibration spectra with the spectral shapes from the clutter, and (2) net analyte pre-processing (NAP). The improved prediction performance was not compromised when reducing the variability in the calibration set, making exhaustive calibration unnecessary. Strong water content variations in the tablets caused frequency shifts of the API absorption signals that could not be included in the clutter. Updating the model for this kind of variation demonstrated that the completeness of the clutter is critical for the performance of these models and that the model will only be more robust for spectral variation that is not co-linear with the one from the property of interest.

KW - Augmentation methods

KW - Near infrared spectroscopy

KW - Orthogonal projection methods

KW - Prior spectral information

KW - Tablets

UR - http://www.scopus.com/inward/record.url?scp=84872162159&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872162159&partnerID=8YFLogxK

U2 - 10.1016/j.aca.2012.11.034

DO - 10.1016/j.aca.2012.11.034

M3 - Article

C2 - 23312315

AN - SCOPUS:84872162159

VL - 761

SP - 62

EP - 70

JO - Analytica Chimica Acta

JF - Analytica Chimica Acta

SN - 0003-2670

ER -