Skip to main content
ARS Home » Research » Publications at this Location » Publication #125751
Title: THE PREDICTION OF RVA PARAMETERS BY NIR SPECTROSCOPY AND TWO-DIMENSIONAL CORRELATION

Author
item MEADOWS, FREDERICK
item BARTON II, FRANKLIN

Submitted to: United States Japan Natural Resources Protein Panel
Publication Type: Proceedings
Publication Acceptance Date: 10/15/2001
Publication Date: N/A
Citation: N/A

Interpretive Summary: RVAs (Rapid Visco Analyzers) are widely used in assessing the cooking and processing qualities in grains. However, the analysis time limits the method. In order to shorten the time needed determine cooking and processing qualities, a novel noninvasive rapid approach has been tried. We have used NIR spectroscopy to predict the traditional RVA parameters, but they are unable to be predicted precisely. Therefore, we have derived a new parameter using PLS2 calculations that may be used in lieu of, or in conjunction with the current RVA measures for cooking qualities

Technical Abstract: RVAs (Rapid Visco Analyzers) are widely used in assessing cooking and processing characteristics in rice. However, traditional parameters (peak viscosity, final viscosity, breakdown, consistency, and setback) rely on several factors. This makes them difficult to predict spectroscopically. For this reason we have investigated methods by which to obtain new RVA parameters that are more predictable. Short, medium, and long grain rice flour samples were used for NIR and RVA measurements. The content of amylose ranged from 0.41% to 24.90% (w/w) and that of protein ranged from 4.89% to 11.35%. Moisture in rice flour samples was approximately 12%. PLS1 (Partial Least Squares Regression) models were created for NIR prediction of RVA parameters. NIR absorptions and RVA viscosity intensities were used in a PLS2 regression model. The PLS2 model resulted in the construction of a 2D matrix map of regression coefficients. The new wparameters were found to correlate more strongly with the NIR (1,100-2,500 nm) region than traditional parameters. This finding suggests that NIR can be used to predict rheological properties.