Raman Spectroscopy and Improved Inception Network for Determination of FHB-Infected Wheat Kernels.

Mengqing Qiu,Shouguo Zheng,Shizhuang Weng,Ling Zheng,Qingshan Xu,Xujin Hu,Le Tang

doi:10.3390/foods11040578

Abstract

Detection of infected kernels is important for Fusarium head blight (FHB) prevention and product quality assurance in wheat. In this study, Raman spectroscopy (RS) and deep learning networks were used for the determination of FHB-infected wheat kernels. First, the RS spectra of healthy, mild, and severe infection kernels were measured and spectral changes and band attribution were analyzed. Then, the Inception network was improved by residual and channel attention modules to develop the recognition models of FHB infection. The Inception–attention network produced the best determination with accuracies in training set, validation set, and prediction set of 97.13%, 91.49%, and 93.62%, among all models. The average feature map of the channel clarified the important information in feature extraction, itself required to clarify the decision-making strategy. Overall, RS and the Inception–attention network provide a noninvasive, rapid, and accurate determination of FHB-infected wheat kernels and are expected to be applied to other pathogens or diseases in various crops.

Highlights

Wheat, the third largest cereal crop in terms of total production, is grown around the world and has become a staple food in Europe and Asia
The process of Fusarium head blight (FHB) infection is accompanied by the accumulation of toxic secondary metabolites, such as deoxynivalenol and zearalenone, which endanger human and livestock health via oxidative stress damage [2]
This study aims to develop a determination method for FHB-infected wheat kernels using Raman spectroscopy (RS) combined with an improved Inception network (Figure 1)

Summary

Introduction

The third largest cereal crop in terms of total production, is grown around the world and has become a staple food in Europe and Asia. Caused by Fusarium graminearum and Fusarium culmorm, Fusarium head blight (FHB) is prone to pandemics in the middle and lower Yangtze River and Jianghuai regions in China, in southern Huanghuai [1]. Because FHB primarily infects the wheat ear, causing shriveled kernels with a chalky or pink color, the yield and quality of wheat are seriously threatened. The process of FHB infection is accompanied by the accumulation of toxic secondary metabolites, such as deoxynivalenol and zearalenone, which endanger human and livestock health via oxidative stress damage [2]. The detection of FHB infection of wheat kernels can ensure rational chemical control [3], guide agricultural practices, and screen FHB-resistant wheat varieties [4] and prevent and evaluate diseases as well as guarantee agricultural production safety. Visual, and biochemical methods are readily available for FHB detection [5]

Objectives

Methods

Results

Conclusion