Pasting cell: An alternative sample cell for detection of Aspergillus flavus infected milled rice by NIR spectroscopy

• Main Authors: Klaithin R., Phetkaeo T., Theanjumpol P., Kunasakdakul K., Thanapornpoonpong S., Vearasilp S.
• 格式: Article
• 出版: Chiang Mai University 2015
• 主題: Multidisciplinary
• 在線閱讀: http://www.scopus.com/inward/record.url?partnerID=HzOxMe3b&scp=84872255488&origin=inward; http://cmuir.cmu.ac.th/handle/6653943832/38056

The objective of this study was to reduce the quantity of milled rice sample used in spectra acquisition system for near infrared spectroscopy (NIRS). The milled rice (non-infected) and the milled rice mixing with Aspergillus flavus infected milled rice at the ratio of 5, 10, 15 and 20% w/w were investigated. The samples were contained in the coarse sample cell and were measured by the reflectance spectra using NIRSystem6500 in wavelength region from 700 to 2500 nm. Then, the same samples were packed in the pasting cell before measuring the spectra. The smoothing and second derivative techniques were used to transform the spectral data. The calibration equation was developed by partial least squares regression (PLSR). It was found that the value of the correlation coefficients (R) of the PLSR calibration result of the pasting cell were higher than the coarse sample cell, which were 0.97 and 0.82 respectively. Moreover, the standard errors of calibration (SEC), the standard errors of prediction (SEP), and the averages of difference between actual and NIR values (Bias) of the pasting cell were lower than the coarse sample cell. They were 1.60, 1.82, -0.28% and 4.04, 4.18, 0.02% respectively. Therefore, the pasting cell could be replaced by the coarse sample cell effectively, and the quantity of milled rice can be reduced.