Acceleration of high-quality Raman imaging via a locality enhanced transformer network.

Raman imaging (RI) is an outstanding technique that enables molecular-level medical diagnostics and therapy assessment by providing characteristic fingerprint and morphological information about molecules. However, obtaining high-quality Raman images generally requires a long acquisition time, up to hours, which is prohibitive for RI applications of timely cytopathology and histopathology analyses. To address this issue, image super-resolution (SR) based on deep learning, including convolutional neural networks and transformers, has been widely recognized as an effective solution to reduce the time required for achieving high-quality RI. In this study, a locality enhanced transformer network (LETNet) is proposed to perform Raman image SR. Specifically, the general architecture of the transformer is adopted with the replacement of self-attention by convolution to generate high-fidelity and detailed SR images. Additionally, the convolution in the LETNet is further optimized by utilizing depth-wise convolution to improve the computational efficiency of the model. Experiments on hyperspectral Raman images of breast cancer cells and Raman images of a few channels of brain tumor tissues demonstrate that the LETNet achieves superior 2×, 4×, and 8× SR with fewer parameters compared with other SR methods. Consequently, high-quality Raman images can be obtained with a significant reduction in time, ranging from 4 to 64 times. Overall, the proposed method provides a novel, efficient, and reliable solution to expedite high-quality RI and promote its application in real-time diagnosis and therapy.

Deep learning networks for the recognition and quantitation of surface-enhanced Raman spectroscopy.

Surface-enhanced Raman spectroscopy (SERS) based on machine learning methods has been applied in material analysis, biological detection, food safety, and intelligent analysis. However, machine learning methods generally require extra preprocessing or feature engineering, and handling large-scale data using these methods is challenging. In this study, deep learning networks were used as fully connected networks, convolutional neural networks (CNN), fully convolutional networks (FCN), and principal component analysis networks (PCANet) to determine their abilities to recognise drugs in human urine and measure pirimiphos-methyl in wheat extract in the two input forms of a one-dimensional vector or a two-dimensional matrix. The best recognition result for drugs in urine with an accuracy of 98.05% in the prediction set was obtained using CNN with spectra as input in the matrix form. The optimal quantitation for pirimiphos-methyl was obtained using FCN with spectra in the matrix form, and the analysis was accomplished with a determination coefficient of 0.9997 and a root mean square error of 0.1574 in the prediction set. These networks performed better than the common machine learning methods. Overall, the deep learning networks provide feasible alternatives for the recognition and quantitation of SERS.

Non-Destructive Detection of Strawberry Quality Using Multi-Features of Hyperspectral Imaging and Multivariate Methods.

Soluble solid content (SSC), pH, and vitamin C (VC) are considered as key parameters for strawberry quality. Spectral, color, and textural features from hyperspectral reflectance imaging of 400-1000 nm was to develop the non-destructive detection approaches for SSC, pH, and VC of strawberries by integrating various multivariate methods as partial least-squares regression (PLSR), support vector regression, and locally weighted regression (LWR). SSC, pH, and VC of 120 strawberries were statistically analyzed to facilitate the partitioning of data sets, which helped optimize the model. PLSR, with spectral and color features, obtained the optimal prediction of SSC with determination coefficient of prediction (Rp2) of 0.9370 and the root mean square error of prediction (RMSEP) of 0.1145. Through spectral features, the best prediction for pH was obtained by LWR with Rp2 = 0.8493 and RMSEP = 0.0501. Combination of spectral and textural features with PLSR provided the best results of VC with Rp2 = 0.8769 and RMSEP = 0.0279. Competitive adaptive reweighted sampling and uninformative variable elimination (UVE) were used to select important variables from the above features. Based on the important variables, the accuracy of SSC, pH, and VC prediction both gain the promotion. Finally, the distribution maps of SSC, pH, and VC over time were generated, and the change trend of three quality parameters was observed. Thus, the proposed method can nondestructively and accurately determine SSC, pH, and VC of strawberries and is expected to design and construct the simple sensors for the above quality parameters of strawberries.

Rapid Detection of Pesticide Residues in Paddy Water Using Surface-Enhanced Raman Spectroscopy.

Pesticide residue in paddy water is one of the main factors affecting the quality and safety of rice, however, the negative effect of this residue can be effectively prevented and reduced through early detection. This study developed a rapid detection method for fonofos, phosmet, and sulfoxaflor in paddy water through chemometric methods and surface-enhanced Raman spectroscopy (SERS). Residue from paddy water samples was directly used for SERS measurement. The obtained spectra from the SERS can detect 0.5 mg/L fonofos, 0.25 mg/L phosmet, and 1 mg/L sulfoxaflor through the appearance of major characteristic peaks. Then, we used chemometric methods to develop models for the intelligent analysis of pesticides, alongside the SERS spectra. The classification models developed by K-nearest neighbor identified all of the samples, with an accuracy of 100%. For the quantitative analysis, the partial least squares regression models obtained the best predicted performance for fonofos and sulfoxaflor, and the support vector machine model provided optimal results, with a root-mean-square error of validation of 0.207 and a coefficient of determination of validation of 0.99952, for phosmet. Experiments for actual contaminated samples also showed that the above models predicted the pesticide residue values with high accuracy. Overall, using SERS with chemometric methods provided a simple and convenient approach for the detection of pesticide residues in paddy water.

Detection of Pirimiphos-Methyl in Wheat Using Surface-Enhanced Raman Spectroscopy and Chemometric Methods.

Pesticide residue detection is a hot issue in the quality and safety of agricultural grains. A novel method for accurate detection of pirimiphos-methyl residues in wheat was developed using surface-enhanced Raman spectroscopy (SERS) and chemometric methods. A simple pretreatment method was conducted to extract pirimiphos-methyl residue from wheat samples, and highly effective gold nanorods were prepared for SERS measurement. Raman peaks assignment was calculated using density functional theory. The Raman signal of pirimiphos-methyl can be detected when the concentrations of residue in wheat extraction solution and contaminated wheat is as low as 0.2 mg/L and 0.25 mg/L, respectively. Quantification of pirimiphos-methyl was performed by applying regression models developed by partial least squares regression, support vector machine regression and random forest with principal component analysis using different preprocessed methods. As for the contaminated wheat samples, the relative deviation between gas chromatography-mass spectrometry value and predicted value is in the range of 0.10%-6.63%, and predicted recovery is 94.12%-106.63%, ranging from 23.93 mg/L to 0.25 mg/L. Results demonstrated that the proposed SERS method is an effective and efficient analytical tool for detecting pirimiphos-methyl in wheat with high accuracy and excellent sensitivity.

Optimal Hotspots of Dynamic Surfaced-Enhanced Raman Spectroscopy for Drugs Quantitative Detection.

Surface-enhanced Raman spectroscopy (SERS) as a powerful qualitative analysis method has been widely applied in many fields. However, SERS for quantitative analysis still suffers from several challenges partially because of the absence of stable and credible analytical strategy. Here, we demonstrate that the optimal hotspots created from dynamic surfaced-enhanced Raman spectroscopy (D-SERS) can be used for quantitative SERS measurements. In situ small-angle X-ray scattering was carried out to in situ real-time monitor the formation of the optimal hotspots, where the optimal hotspots with the most efficient hotspots were generated during the monodisperse Au-sol evaporating process. Importantly, the natural evaporation of Au-sol avoids the nanoparticles instability of salt-induced, and formation of ordered three-dimensional hotspots allows SERS detection with excellent reproducibility. Considering SERS signal variability in the D-SERS process, 4-mercaptopyridine (4-mpy) acted as internal standard to validly correct and improve stability as well as reduce fluctuation of signals. The strongest SERS spectra at the optimal hotspots of D-SERS have been extracted to statistics analysis. By using the SERS signal of 4-mpy as a stable internal calibration standard, the relative SERS intensity of target molecules demonstrated a linear response versus the negative logarithm of concentrations at the point of strongest SERS signals, which illustrates the great potential for quantitative analysis. The public drugs 3,4-methylenedioxymethamphetamine and α-methyltryptamine hydrochloride obtained precise analysis with internal standard D-SERS strategy. As a consequence, one has reason to believe our approach is promising to challenge quantitative problems in conventional SERS analysis.

Detection and direct readout of drugs in human urine using dynamic surface-enhanced Raman spectroscopy and support vector machines.

A new, novel, rapid method to detect and direct readout of drugs in human urine has been developed using dynamic surface-enhanced Raman spectroscopy (D-SERS) with portable Raman spectrometer on gold nanorods (GNRs) and a classification algorithm called support vector machines (SVM). The high-performance GNRs can generate gigantic enhancement and the SERS signals obtained using D-SERS on it have high reproducibility. On the basis of this feature of D-SERS, we have obtained SERS spectra of urine and urine containing methamphetamine (MAMP). SVM model was built using these data for fast identified and visual results. This general method was successfully applied to the detection of 3, 4-methylenedioxy methamphetamine (MDMA) in human urine. To verify the accuracy of the model, drug addicts' urine containing MAMP were detected and identified correctly and rapidly with accuracy more than 90%. The detection results were displayed directly without analysis of their SERS spectra manually. Compared with the conventional method in lab, the method only needs a 2 µL sample volume and takes no more than 2 min on the portable Raman spectrometer. It is anticipated that this method will enable rapid, convenient detection of drugs on site for the police.

Three-dimensional surface-enhanced Raman scattering hotspots in spherical colloidal superstructure for identification and detection of drugs in human urine.

Rapid component separation and robust surface-enhanced Raman scattering (SERS) identification of drugs in real human urine remain an attractive challenge because of the sample complexity, low molecular affinity for metal surface, and inefficient use of hotspots in one- or two-dimensional (2D) geometries. Here, we developed a 5 min strategy of cyclohexane (CYH) extraction for separating amphetamines from human urine. Simultaneously, an oil-in-water emulsion method is used to assemble monodisperse Ag nanoparticles in the CYH phase into spherical colloidal superstructures in the aqueous phase. These superstructures create three-dimensional (3D) SERS hotspots which exist between every two adjacent particles in 3D space, break the traditional 2D limitation, and extend the hotspots into the third dimension along the z-axis. In this platform, a conservative estimate of Raman enhancement factor is larger than 10(7), and the same CYH extraction processing results in a high acceptability and enrichment of drug molecules in 3D hotspots which demonstrates excellent stability and reproducibility and is suitable for the quantitative examination of amphetamines in both aqueous and organic phases. Parallel ultraperformance liquid chromatography (UPLC) examinations corroborate an excellent performance of our SERS platform for the quantitative analysis of methamphetamine (MA) in both aqueous solution and real human urine, of which the detection limits reach 1 and 10 ppb, respectively, with tolerable signal-to-noise ratios. Moreover, SERS examinations on different proportions of MA and 3,4-methylenedioxymethamphetamine (MDMA) in human urine demonstrate an excellent capability of multiplex quantification of ultratrace analytes. By virtue of a spectral classification algorithm, we realize the rapid and accurate recognition of weak Raman signals of amphetamines at trace levels and also clearly distinguish various proportions of multiplex components. Our platform for detecting drugs promises to be a great prospect for a rapid, reliable, and on-spot analyzer.

A Quantum Dot Fluorescence Sensor System Design for Hg²âº Trace Detection.

The detection of Hg²âº ions usually requires large laboratory equipment, which encounters difficulties for rapid field test in most applications. In this paper, we design a reflective sensor for trace Hg²âº analysis based on the fluorescent quenching of Quantum dots, which contains two major modules, i. e. the fluorescent sensing module and the signal processing module. The fluorescence sensing module is composed of a laser source, a light collimated system and a photo-detector, which enables the realization of the fluorescence excitation as well as its detection. The signal processing module realized the further amplification of the detected signal and hereafter the filtering of noises. Furthermore, the Hg²âº concentration will displayed on the QT interface using a Linux embedded system. The sensor system is low cost and small, which makes it available for rapid field test or portable applications. Experimental results show that the sensor has a good linear relationship for the Hg²âº concentration range from 15.0 x 10⁻9 to 1.8 x 10⁻6 mol · L⁻¹. The regression equation is V0/V = 1.309 13 + 3.37c, where c is Hg²âº concentration, and V0 is the voltage value for the blank case. In our work, the linearity is determined as 0. 989 26. The experiments exhibit that Ca²âº, Mn²âº and Pb²âº ions have small influence on the Hg²âº detection, and the interfere of other common ions can be neglected, which indicates a good selectivity of the sensor. Finally, it shows that our sensor has a rapid response time of 35 s and a good repeatability, thus it is potential for field test of trace Hg²âº.

Rapid and accurate identification of bakanae pathogens carried by rice seeds based on hyperspectral imaging and deep transfer learning.

Bakanae disease is a common seed-borne disease of rice. Rapid and accurate detection of bakanae pathogens carried by rice seeds is essential for the health of rice germplasm resources and the safety of rice production. This study aims to propose a general framework for species identification of major bakanae pathogens carried by rice seeds based on hyperspectral imaging and deep transfer learning. Seven varieties of rice seeds and four kinds of bakanae pathogens were analyzed. One-dimensional deep convolution neural networks (DCNNs) were first constructed using complete datasets. They achieved accuracies larger than 96.5% on the testing sets of most datasets, exceeding the conventional SVM and PLS-DA models. Then the developed DCNNs were transferred to detect other complete training sets. Most of the deep transferred models achieved comparable or even better performance than the original DCNNs. Two smaller target training sets were further constructed by randomly selecting spectra from the complete training sets. As the size of the target training sets reduced, the accuracies of all models on the corresponding testing sets also decreased gradually. Visualization analysis were conducted using the t-distribution stochastic neighbor embedding (t-SNE) algorithm and a proposed gradient-weighted activation wavelength (Grad-AW) method. They all showed that deep transfer learning could utilize the representation patterns in the source datasets to improve the target tasks. The overall results indicated that the bakanae pathogens were all identified accurately under our proposed framework. Hyperspectral imaging combined with deep transfer learning provided a new idea for the quality detection of large-scale seeds in modern seed industry.

Identification of surface-enhanced Raman spectroscopy using hybrid transformer network.

Surface-enhanced Raman Spectroscopy (SERS) is extensively implemented in drug detection due to its sensitivity and non-destructive nature. Deep learning methods, which are represented by convolutional neural network (CNN), have been widely applied in identifying the spectra from SERS for powerful learning ability. However, the local receptive field of CNN limits the feature extraction of sequential spectra for suppressing the analysis results. In this study, a hybrid Transformer network, TMNet, was developed to identify SERS spectra by integrating the Transformer encoder and the multi-layer perceptron. The Transformer encoder can obtain precise feature representations of sequential spectra with the aid of self-attention, and the multi-layer perceptron efficiently transforms the representations to the final identification results. TMNet performed excellently, with identification accuracies of 99.07% for the spectra of hair containing drugs and 97.12% for those of urine containing drugs. For the spectra with additive white Gaussian, baseline background, and mixed noises, TMNet still exhibited the best performance among all the methods. Overall, the proposed method can accurately identify SERS spectra with outstanding noise resistance and excellent generalization and holds great potential for the analysis of other spectroscopy data.

[Quantitative detection of ethyl paraoxon based on SERS and PCA-SLR].

In the present paper, the surface-enhanced Raman spectroscopy (SERS) was used to build the model for the quantitative detection of ethyl paraoxon by the principal component analysis and segmented linear regression (PCA-SLR). Firstly, SERS in 820-1630 cm(-1) of ethyl paraoxon solution were measured and the spectra in 820-1630 cm(-1)(complete range) and 845-875 cm(-1) (characteristic range) of ethyl paraoxon solution were preprocessed by standard normal transformation (SNV), multiplicative scatter correction (MSC), the absolute values of first derivative and the second derivative respectively. Additionally, the number of dimensions of the spectra was reduced by PCA. Finally, the models were established by SLR It was found that the model developed with MSC preprocessed spectroscopy of characteristic range performed best (RMSEP: 0.33) by comparing the predictive accuracy of the different models. The result could meet with the needs in the quantitative detection of ethyl paraoxon.

Statistic and Network Features of RGB and Hyperspectral Imaging for Determination of Black Root Mold Infection in Apples.

Apples damaged by black root mold (BRM) lose moisture, vitamins, and minerals as well as carry dangerous toxins. Determination of the infection degree can allow for customized use of apples, reduce financial losses, and ensure food safety. In this study, red-green-blue (RGB) imaging and hyperspectral imaging (HSI) are combined to detect the infection degree of BRM in apple fruits. First, RGB and HSI images of healthy, mildly, moderately, and severely infected fruits are measured, and those with effective wavelengths (EWs) are screened from HSI by random frog. Second, the statistic and network features of images are extracted by using color moment and convolutional neural network. Meanwhile, random forest (RF), K-nearest neighbor, and support vector machine are used to construct classification models with the above two features of RGB and HSI images of EWs. Optimal results with the 100% accuracy of training set and 96% accuracy of prediction set are obtained by RF with the statistic and network features of the two images, outperforming the other cases. The proposed method furnishes an accurate and effective solution for determining the BRM infection degree in apples.

Detection of 1-OHPyr in human urine using SERS with injection under wet liquid-liquid self-assembled films of ß-CD-coated gold nanoparticles and deep learning.

1-Hydroxypyrene (1-OHPyr), a typical hydroxylated polycyclic aromatic hydrocarbon (OH-PAH), has been commonly regarded as a urinary biomarker for assessing human exposure and health risks of PAHs. Herein, a fast and sensitive method was developed for the determination of 1-OHPyr in urine using surface-enhanced Raman spectroscopy (SERS) combined with deep learning (DL). After emulsification, urinary 1-OHPyr was separated using simple liquid-liquid extraction. Gold nanoparticles with ß-cyclodextrin (ß-CD@AuNPs) were synthesized, and homogeneous and ordered ß-CD@AuNP films were prepared through a liquid-liquid interface self-assembly process. The separated 1-OHPyr was injected under wet assembled films for SERS detection. Concentration as low as 0.05 µg mL-1 of 1-OHPyr in urine could still be detected, and the relative standard deviation was 5.5 %, and this was ascribed to the adsorption of ß-CD and the high-probability contact between 1-OHPyr molecules and the nanogap of assembled films under the action of capillary force. Meanwhile, a convolutional neural network (CNN), a classical DL network architecture, was adopted to build the prediction model, and the model was further simplified by genetic algorithm (GA). CNN combined with a GA obtained optimized results with determination coefficient and a root mean square error of prediction sets of 0.9639 and 0.6327, respectively, outperforming other models. Overall, the proposed method achieves fast and accurate detection of 1-OHPyr in urine, improves the assessment human exposure to PAHs and is expected to have applications in the analysis of other OH-PAHs in complex environments.

SERS with Flexible ß-CD@AuNP/PTFE Substrates for In Situ Detection and Identification of PAH Residues on Fruit and Vegetable Surfaces Combined with Lightweight Network.

The detection of polycyclic aromatic hydrocarbons (PAHs) on fruit and vegetable surfaces is important for protecting human health and ensuring food safety. In this study, a method for the in situ detection and identification of PAH residues on fruit and vegetable surfaces was developed using surface-enhanced Raman spectroscopy (SERS) based on a flexible substrate and lightweight deep learning network. The flexible SERS substrate was fabricated by assembling ß-cyclodextrin-modified gold nanoparticles (ß-CD@AuNPs) on polytetrafluoroethylene (PTFE) film coated with perfluorinated liquid (ß-CD@AuNP/PTFE). The concentrations of benzo(a)pyrene (BaP), naphthalene (Nap), and pyrene (Pyr) residues on fruit and vegetable surfaces could be detected at 0.25, 0.5, and 0.25 µg/cm2, respectively, and all the relative standard deviations (RSD) were less than 10%, indicating that the ß-CD@AuNP/PTFE exhibited high sensitivity and stability. The lightweight network was then used to construct a classification model for identifying various PAH residues. ShuffleNet obtained the best results with accuracies of 100%, 96.61%, and 97.63% for the training, validation, and prediction datasets, respectively. The proposed method realised the in situ detection and identification of various PAH residues on fruit and vegetables with simplicity, celerity, and sensitivity, demonstrating great potential for the rapid, nondestructive analysis of surface contaminant residues in the food-safety field.

Detection of amylase activity and moisture content in rice by reflectance spectroscopy combined with spectral data transformation.

In this study, reflectance spectroscopy was used to achieve rapid and non-destructive detection of amylase activity and moisture content in rice. Since rice husk can interfere with spectral measurements, spectral data transformation was used to remove the husk interference. Reflectance spectra of rice were transformed by direct standardization, convolutional autoencoder network, and kernel regression (KR). Then, random frog and elliptical envelope were adopted to select effective wavelengths, and partial least squares regression (PLSR) and support vector regression were used to establish analysis models. The optimal transformation was from KR, and PLSR and effective wavelengths of the transformed spectra obtained excellent performance with coefficient of determination of test of 0.6987 and 0.8317 and root-mean-square error of test of 0.3359 and 2.2239, respectively. The result was better than that of the rice spectra and was close to that of the husked rice spectra. When the moisture content was integrated into the regression model of amylase activity, a better result was obtained. Thus, the proposed method can detect amylase activity and moisture content in rice accurately.

Surface-enhanced Raman spectroscopy charged probes under inverted superhydrophobic platform for detection of agricultural chemicals residues in rice combined with lightweight deep learning network.

In this study, surface-enhanced Raman spectroscopy (SERS) charged probes and an inverted superhydrophobic platform were used to develop a detection method for agricultural chemicals residues (ACRs) in rice combined with lightweight deep learning network. First, positively and negatively charged probes were prepared to adsorb ACRs molecules to SERS substrate. An inverted superhydrophobic platform was prepared to alleviate the coffee ring effect and induce tight self-assembly of nanoparticles for high sensitivity. Chlormequat chloride of 15.5-0.05 mg/L and acephate of 100.2-0.2 mg/L in rice were measured with the relative standard deviation of 4.15% and 6.25%. SqueezeNet were used to develop regression models for the analysis of chlormequat chloride and acephate. And the excellent performances were obtained with the coefficients of determination of prediction of 0.9836 and 0.9826 and root-mean-square errors of prediction of 0.49 and 4.08. Therefore, the proposed method can realize sensitive and accurate detection of ACRs in rice.

Detection of apple fruit damages through Raman spectroscopy with cascade forest.

Apple fruit damages seriously cause product and economic losses, infringe consumer rights and interests, and have harmful effects on human and livestock health. In this study, Raman spectroscopy (RS) and cascade forest (CForest) were adopted to determine apple fruit damages. First, the RS spectra of healthy, bruised, Rhizopus-infected, and Botrytis-infected apples were measured. Spectral changes and band attribution were analyzed. Different modeling methods were combined with various pre-processing and dimension reduction methods to construct recognition models. Among all models, CForest constructed with full spectra processed by Savitsky-Golay smoothing obtained the best performance with accuracies of 100%, 91.96%, and 92.80% in the training, validation, and test sets (ACCTE). And the modeling time is reduced to 1/3 of the full-spectra model with a similar ACCTE of 91.56% after principal component analysis. Overall, RS and CForest provided a non-destructive, rapid, and accurate identification of apple fruit damages and could be used in disease recognition and safety assurance of other fruits.

Drought stress identification of tomato plant using multi-features of hyperspectral imaging and subsample fusion.

Drought stress (DS) is one of the most frequently occurring stresses in tomato plants. Detecting tomato plant DS is vital for optimizing irrigation and improving fruit quality. In this study, a DS identification method using the multi-features of hyperspectral imaging (HSI) and subsample fusion was proposed. First, the HSI images were measured under imaging condition with supplemental blue lights, and the reflectance spectra were extracted from the HSI images of young and mature leaves at different DS levels (well-watered, reduced-watered, and deficient-watered treatment). The effective wavelengths (EWs) were screened by the genetic algorithm. Second, the reference image was determined by ReliefF, and the first four reflectance images of EWs that are weakly correlated with the reference image and mutually irrelevant were obtained using Pearson's correlation analysis. The reflectance image set (RIS) was determined by evaluating the superposition effect of reflectance images on identification. The spectra of EWs and the image features extracted from the RIS by LeNet-5 were adopted to construct DS identification models based on support vector machine (SVM), random forest, and dense convolutional network. Third, the subsample fusion integrating the spectra and image features of young and mature leaves was used to improve the identification further. The results showed that supplemental blue lights can effectively remove the high-frequency noise and obtain high-quality HSI images. The positive effect of the combination of spectra of EWs and image features for DS identification proved that RIS contains feature information pointing to DS. Global optimal classification performance was achieved by SVM and subsample fusion, with a classification accuracy of 95.90% and 95.78% for calibration and prediction sets, respectively. Overall, the proposed method can provide an accurate and reliable analysis for tomato plant DS and is hoped to be applied to other crop stresses.

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

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.