Non-Invasive Detection of Protein Content in Corn Distillers Dried Grains with Solubles: Method for Selecting Spectral Variables to Construct High-Performance Calibration Model Using near Infrared Reflectance Spectroscopy

Abstract

All rights reservedCorn distillers dried grains with solubles (DDGS) is the main byproduct of the bioethanol industry 1 . It has high nutritional content, especially protein. Using corn DDGS as protein feed material not only increases animal productivity but also saves cost. 2,3 However, the protein content in corn DDGS can vary due to different sources of corn grains and different fermentation conditions. 1,4,5 The protein content in corn DDGS is essential for livestock diet formulation and a major determinant of price. Conventional laboratory analysis methods for determining protein content are time-consuming and costly. Near infrared (NIR) reflectance spectroscopy is a rapid, noninvasive, reliable and green detection technology, which has Corn distillers dried grains with solubles (DDGS), a byproduct of the bioethanol industry, is commonly used as animal feed. This paper evaluates the use of backward variable selection partial least square (BVSPLS) and genetic algorithm (GA) methods to select the spectral variables of near infrared (NIR) reflectance spectroscopy and construct high-performance calibration models of protein content in corn DDGS. The BVSPLS analysis utilised 16% of the spectral variables. Compared to the full spectrum model, the model constructed from the variables selected by the BVSPLS analysis significantly improved the accuracy of the model fit and achieved a 19% decrease in the standard error of validation (SEP) and a 23% increase in the residual validation deviation (RPD). The GA analysis selected 8% of the total NIR spectral variables and the model constructed from these selected variables had a fitted accuracy comparable to that of the full spectrum model. The spectral variables selected by both the BVSPLS analysis and GA analysis significantly simplified the NIR calibration model and provided better correlation between the selected spectral variables and protein content of corn DDGS. These results also have important implications for the development of a rapid, non-invasive, online analysis system to detect protein content of corn DDGS in-situ.

Filter design for molecular factor computing using wavelet functions

“…7(a) for the wheat dataset, it can be seen that the filter only passes the spectral range of 568-598 nm, 980-1430 nm and 2144-2230 nm. The spectral bands around 1190 nm and 2170 nm can be assigned to CÀ ÀH stretching second overtone and NÀ ÀH bending first overtone and stretching combinations of amine [44], respectively. For the MFC filter in Fig.…”

“…7(b) of the blood samples, the passing spectral ranges are located at 1266-1362 nm, 1522-1696 nm and 2110-2326 nm. The spectral bands around 1550 nm and 2170 nm can be assigned to NÀ ÀH stretching first overtone and NÀ ÀH bending first overtone and stretching combinations of amine [44], respectively. The same spectral band around 2170 is selected as the informative variable for the three datasets for the model of protein in corn, wheat and blood samples.…”

“…It indicates that only partial spectral variables contribute to the model, and some of the variables are not used in the model. Although it is difficult to analyze the attribution of the variables in an NIR spectrum, the spectral bands around 2070 nm can be assigned to NÀ ÀH stretching or bending combinations of amine, and the spectral band around 2160 nm can be attributed to NÀ ÀH bending first overtone and stretching combinations of amine [44]. Furthermore, the spectral bands around 1800, 2200 and 2160 nm are selected as the informative variables in the results obtained by CARS [39] and MCS-RPLS [40].…”

Using FT-NIR Calibration to Determine Soybean Meal Amount in Concentrated Feed Ration Mixtures for Laying Hens

“…Determination of moisture, protein, fat and carbohydrate contents by wet methods is a laborious process that requires time and characteristically produces a lot of chemical waste. However, methods using the FT-NIR technique provide precise and rapid results in order to be an alternative to such determinations with minimal or blank sample preparation and no waste (Kandalave et al, 2012;SalgóveGergely, 2012;Zhou et al, 2012). However, multivariate calibration methods should be used to obtain quantitative information for the simultaneous evaluation of spectral properties and various parameters in the NIR region.…”

A Multi-Model Method in Near Infrared Spectra Analysis

“…It has the advantages of high speed, high efficiency, low cost, good test reproducibility and convenient measurement. It has been increasingly used in food industry, petrochemical industry, medicine and other fields [1]- [6]. In recent years, it has been applied more and more in wood science research [7]- [10].…”