Articles
POTENTIAL OF NEAR-INFRARED REFLECTANCE SPECTROSCOPY FOR ESTIMATING TOTAL DIETARY FIBER CONTENT OF DEHYDRATED VEGETABLES
Article number
483_29
Pages
265 – 274
Language
Abstract
The conventional procedure for determining total dietary fiber (TDF) is very time consuming.
A near-infrared reflectance (NIR) spectrophotometric procedure was evaluated for potential use in assaying TDF of vegetables rapidly.
Dehydrated vegetables were milled, and TDF was measured by the AOAC (985.29) enzymatic-gravimetric method.
TDF values ranged from 8.41 to 37.19% of dry weight.
Milled vegetables were scanned from 1100 to 2498 nm with a NIR Systems 6500 monochrometer.
A dietary fiber calibration was obtained with multiple linear regression analysis as the regression method.
The selected wavelengths in the regression equation were 2244, 2214, 2392 and 1172 nm.
The standard error and multiple correlation coefficient of a calibration sample set (n=66) were 2.10% and 0.94%, respectively.
The root mean square of the differences of a prediction sample set (n=22) was 2.45%. But other samples (n=27) generated mostly bigger differences between the value of NIR-calculated TDF and that of AOAC-determined TDF. The samples used were viscous with a high TDF content as indicated by the brown color of the dehydrated vegetable, thus it was impossible to observe a peak top in the vicinity of 2248 nm, highly NIR-calculated TDF value (more than about 35%). As mentioned above, if applied samples are distinguished, NIR spectroscopy can estimate TDF content of dehydrated vegetables rapidly.
A near-infrared reflectance (NIR) spectrophotometric procedure was evaluated for potential use in assaying TDF of vegetables rapidly.
Dehydrated vegetables were milled, and TDF was measured by the AOAC (985.29) enzymatic-gravimetric method.
TDF values ranged from 8.41 to 37.19% of dry weight.
Milled vegetables were scanned from 1100 to 2498 nm with a NIR Systems 6500 monochrometer.
A dietary fiber calibration was obtained with multiple linear regression analysis as the regression method.
The selected wavelengths in the regression equation were 2244, 2214, 2392 and 1172 nm.
The standard error and multiple correlation coefficient of a calibration sample set (n=66) were 2.10% and 0.94%, respectively.
The root mean square of the differences of a prediction sample set (n=22) was 2.45%. But other samples (n=27) generated mostly bigger differences between the value of NIR-calculated TDF and that of AOAC-determined TDF. The samples used were viscous with a high TDF content as indicated by the brown color of the dehydrated vegetable, thus it was impossible to observe a peak top in the vicinity of 2248 nm, highly NIR-calculated TDF value (more than about 35%). As mentioned above, if applied samples are distinguished, NIR spectroscopy can estimate TDF content of dehydrated vegetables rapidly.
Authors
H. Ito, K. Ippoushi, K. Azuma, H. Higashio
Keywords
Spinacia, Lactuca, Brassica, Raphanus, Cucumis, Allium, Pisum, Capsicum, Daucus
Online Articles (49)