Application of near infrared (NIR) spectroscopy coupled to chemometrics for dried egg-pasta characterization and egg content quantification

Research output: Contribution to journalJournal articleResearchpeer-review

Dried egg pasta is an important and traditional food in the Italian cuisine, and the eggs in pasta improve its nutritional value and organoleptic properties. For this reason the percentage of eggs present in the products sold as “egg pasta” has to always be clearly reported in the label. In this respect, the present research addresses the possibility of developing a method which would allow fast, simple and economic determination of egg content in dried egg-pasta, using near-infrared spectroscopy and chemometric analysis. However, as it is very likely that the spectroscopic fingerprint can also be affected by the manufacturing process of this product, in particular by drying temperature and time, the effect of the manufacturing process on the spectral profile of egg-pasta samples was thoroughly investigated, using experimental design coupled to a multivariate exploratory data analytical technique called ANOVA–Simultaneous Component Analysis (ASCA). Moreover, once confirmed the significance of the drying effect on spectral shape, with the aim of building a calibration model to quantify the egg content in pasta samples irrespective of the manufacturing protocol adopted, a non-linear approach based on local regression, namely LWR-PLS, was investigated. This method allowed the determination of the egg content in external validation samples with low error (RMSEP=1.25), resulting in predictions more accurate and precise than those obtained by a global PLS model.
Original languageEnglish
JournalFood Chemistry
Issue number4
Pages (from-to)726 - 734
Publication statusPublished - 2013
Externally publishedYes

Bibliographical note

Special Issue: Food Quality Evaluation

    Research areas

  • Dried egg-pasta, Near infrared (NIR) spectroscopy, Chemometrics, ANOVA–Simultaneous Component Analysis (ASCA), Locally weighted partial least squares regression (LWR-PLS)

ID: 228373079