The applicability of Fourier transform near‐infrared (FTNIR), or mid‐infrared (FTMIR) spectroscopy, and their combination for rapid analysis of DON in wheat bran, was investigated for the classification of samples into compliant and non‐compliant groups regarding the EU legal limit of 750 µg kg−1. Partial least squares‐discriminant analysis (PLS‐DA) and principal component‐linear discriminant analysis (PC‐LDA) were employed as classification techniques using a cutoff value of 400 µg kg−1 DON to distinguish the two classes. Depending on the classification model, overall discrimination rates were from 87% to 91% for FTNIR and from 86% to 87% for the FTMIR spectral range. The FTNIR spectroscopy gave the highest overall classification rate of wheat bran samples, with no false compliant samples and 18% false noncompliant samples when the PC‐LDA classification model was applied. The combination of the two spectral ranges did not provide a substantial improvement in classification results in comparison with FTNIR.
DE GIROLAMO A.;
CERVELLIERI Salvatore;
CORTESE Marina;
PORRICELLI Anna Chiara Raffaella;
PASCALE Michelangelo;
LONGOBARDI Francesco;
VON HOLST Christoph;
CIACCHERI Leonardo;
LIPPOLIS V.;
2019-06-06
WILEY-BLACKWELL
JRC109635
0022-5142 (online),
https://publications.jrc.ec.europa.eu/repository/handle/JRC109635,
10.1002/jsfa.9392 (online),
| Name | Country | City | Type |
|---|
This document is only visible at the Commission level.
You are not authorized to publish or distribute it outside the European Commission.
This is a public document. You can share this publication.
Datasets
| ID | Title | Public URL |
|---|
Dataset collections
| ID | Acronym | Title | Public URL |
|---|
Scripts / source codes
| Description | Public URL |
|---|
Additional supporting files
| File name | Description | File type |
|---|