Hyperspectral Signatures Research Articles

RGB to hyperspectral: Spectral reconstruction for enhanced surgical imaging

AbstractThis study investigates the reconstruction of hyperspectral signatures from RGB data to enhance surgical imaging, utilizing the publicly available HeiPorSPECTRAL dataset from porcine surgery and an in‐house neurosurgery dataset. Various architectures based on convolutional neural networks (CNNs) and transformer models are evaluated using comprehensive metrics. Transformer models exhibit superior performance in terms of RMSE, SAM, PSNR and SSIM by effectively integrating spatial information to predict accurate spectral profiles, encompassing both visible and extended spectral ranges. Qualitative assessments demonstrate the capability to predict spectral profiles critical for informed surgical decision‐making during procedures. Challenges associated with capturing both the visible and extended hyperspectral ranges are highlighted using the MAE, emphasizing the complexities involved. The findings open up the new research direction of hyperspectral reconstruction for surgical applications and clinical use cases in real‐time surgical environments.

Non‐destructively characterizing sandstones, orthoquartzites, agates, and petrified wood for provenance research: Perspectives from the Southeastern Coastal Plain, United States

AbstractSiliceous sandstone (including quartzites), petrified wood, and agates located in Alabama and Mississippi were utilized as a toolstone resource during every recognized cultural period in the Lower Mississippi Valley region of the Southeastern United States. Regrettably, these materials have not been the focus of many provenance‐related investigations. Recent analyses of quartzite and sandstone from other regions in North America and from the Pyrenees were successful in discriminating sources using petrographic techniques. The current study examines the application of visible/near‐infrared reflectance and Fourier transform infrared reflectance (FTIR) spectroscopy on sourcing siliceous materials besides chert, particularly sandstones, orthoquartzites (quartz sandstone), petrified woods, and agates. This source characterization investigation focuses on a case study involving materials gathered from eight distinct collection sites, encompassing nine different siliceous resources collected in Alabama and Mississippi. These materials were sourced from two distinct geological formations: the Hattiesburg and Tallahatta. Results demonstrate the ability of non‐destructive reflectance spectroscopy and introduces a new outlier modeling method that detects, clusters, and separately models outliers with their own set of basis vectors. Principal component analyses, least absolute shrinkage and selection operator regression, linear discriminant function analysis (LDA), and random forest classification are used in this paper to better identify outlier elements as well as discriminate for stone materials accurately (between 67% and 100%). Although this is the first reflectance spectroscopy investigation used to characterize these materials for provenance applications, the preliminary results compare favorably with other provenance techniques whose aim is to quantify between‐formation (inter) and within‐formation (intra) outcrop variation. The quantified and differentiated sources, based on the hyperspectral signatures of the material, will provide a better understanding of prehistoric reliance on these lithic resources and produces a proxy to determine mobility, social interaction, and other past behavior.

Retinal hyperspectral imaging in mouse models of Parkinson’s disease and healthy aging

Retinal hyperspectral imaging (HSI) is a non-invasive in vivo approach that has shown promise in Alzheimer’s disease. Parkinson’s disease is another neurodegenerative disease where brain pathobiology such as alpha-synuclein and iron overaccumulation have been implicated in the retina. However, it remains unknown whether HSI is altered in in vivo models of Parkinson’s disease, whether it differs from healthy aging, and the mechanisms which drive these changes. To address this, we conducted HSI in two mouse models of Parkinson’s disease across different ages; an alpha-synuclein overaccumulation model (hA53T transgenic line M83, A53T) and an iron deposition model (Tau knock out, TauKO). In comparison to wild-type littermates the A53T and TauKO mice both demonstrated increased reflectivity at short wavelengths ~ 450 to 600 nm. In contrast, healthy aging in three background strains exhibited the opposite effect, a decreased reflectance in the short wavelength spectrum. We also demonstrate that the Parkinson’s hyperspectral signature is similar to that from an Alzheimer’s disease model, 5xFAD mice. Multivariate analyses of HSI were significant when plotted against age. Moreover, when alpha-synuclein, iron or retinal nerve fibre layer thickness were added as a cofactor this improved the R2 values of the correlations in certain groups. This study demonstrates an in vivo hyperspectral signature in Parkinson’s disease that is consistent in two mouse models and is distinct from healthy aging. There is also a suggestion that factors including retinal deposition of alpha-synuclein and iron may play a role in driving the Parkinson’s disease hyperspectral profile and retinal nerve fibre layer thickness in advanced aging. These findings suggest that HSI may be a promising translation tool in Parkinson’s disease.

Using computational learning for non-melanoma skin cancer and actinic keratosis near-infrared hyperspectral signature classification

BackgroundThe early detection of Non-Melanoma Skin Cancer (NMSC) is essential to ensure patients receive the most effective treatment. Diagnostic screening tools for NMSC are crucial due to high confusion rates with other types of skin lesions, such as Actinic Keratosis. Nevertheless, current means of diagnosing and screening patients rely on either visual criteria, that are often conditioned by subjectivity and experience, or highly invasive, slow, and costly methods, such as histological diagnoses. From this, the objectives of the present study are to test if classification accuracies improve in the Near-Infrared region of the electromagnetic spectrum, as opposed to previous research in shorter wavelengths. MethodsThis study utilizes near-infrared hyperspectral imaging, within the range of 900.6 and 1454.8 nm. Images were captured for a total of 125 patients, including 66 patients with Basal Cell Carcinoma, 42 with cutaneous Squamous Cell Carcinoma, and 17 with Actinic Keratosis, to differentiate between healthy and unhealthy skin lesions. A combination of hybrid convolutional neural networks (for feature extraction) and support vector machine algorithms (as a final activation layer) was employed for analysis. In addition, we test whether transfer learning is feasible from networks trained on shorter wavelengths of the electromagnetic spectrum. ResultsThe implemented method achieved a general accuracy of over 80 %, with some tasks reaching over 90 %. F1 scores were also found to generally be over the optimal threshold of 0.8. The best results were obtained when detecting Actinic Keratosis, however differentiation between the two types of malignant lesions was often noted to be more difficult. These results demonstrate the potential of near-infrared hyperspectral imaging combined with advanced machine learning techniques in distinguishing NMSC from other skin lesions. Transfer learning was unsuccessful in improving the training of these algorithms. ConclusionsWe have shown that the Near-Infrared region of the electromagnetic spectrum is highly useful for the identification and study of non-melanoma type skin lesions. While the results are promising, further research is required to develop more robust algorithms that can minimize the impact of noise in these datasets before clinical application is feasible.

Predicting Mechanical Properties of Carbonate Rocks Using Spectroscopy Across 0.4–12 μm

Determining the mechanical characteristics of rocks is crucial in various civil engineering sectors. Traditionally, the mechanical properties of rocks are determined through on-site and laboratory tests carried out during geotechnical surveys. However, these extensive surveys require considerable time and resources. In contrast, hyperspectral remote sensing techniques offer a rapid and simple means to determine the mineral composition and crystallographic structure of rocks. These features, in turn, influence the rocks' mechanical properties. This study focuses on characterizing the mechanical properties of carbonate rocks in a laboratory setting, using hyperspectral sensors. Approximately 150 cylindrical carbonate rock samples, spanning a wide strength range, were collected from diverse Israeli rock outcrops. Employing a point spectrometer (0.4 to 2.5 µm) and a spectral image sensor (8.0 to 12.0 µm), we captured samples' light reflections and spectral emissivity. Mechanical attributes, including density, porosity, water absorption, and uniaxial compressive strength (UCS), were measured. Advanced data mining techniques identified statistical correlations between hyperspectral signatures and mechanical properties, pinpointing key wavelengths for prediction. The developed models exhibited excellent predictability for the specified properties, attributing accuracy to discernible mineralogy and internal crystalline structure through spectroscopy. However, predicting UCS showed slightly weaker results due to influences from internal flaws not entirely reflected in spectroscopic data. Nonetheless, outcomes regarding rock UCS were deemed satisfactory. These findings open avenues for non-destructive tools in assessing the mechanical properties of rocks in quarrying operations.

Classification of Grapevine Varieties Using UAV Hyperspectral Imaging

Classifying grapevine varieties is crucial in precision viticulture, as it allows for accurate estimation of vineyard row growth for different varieties and ensures authenticity in the wine industry. This task can be performed with time-consuming destructive methods, including data collection and analysis in the laboratory. In contrast, unmanned aerial vehicles (UAVs) offer a markedly more efficient and less restrictive method for gathering hyperspectral data, even though they may yield data with higher levels of noise. Therefore, the first task is the processing of these data to correct and downsample large amounts of data. In addition, the hyperspectral signatures of grape varieties are very similar. In this study, we propose the use of a convolutional neural network (CNN) to classify seventeen different varieties of red and white grape cultivars. Instead of classifying individual samples, our approach involves processing samples alongside their surrounding neighborhood for enhanced accuracy. The extraction of spatial and spectral features is addressed with (1) a spatial attention layer and (2) inception blocks. The pipeline goes from data preparation to dataset elaboration, finishing with the training phase. The fitted model is evaluated in terms of response time, accuracy and data separability and is compared with other state-of-the-art CNNs for classifying hyperspectral data. Our network was proven to be much more lightweight by using a limited number of input bands (40) and a reduced number of trainable weights (560 k parameters). Hence, it reduced training time (1 h on average) over the collected hyperspectral dataset. In contrast, other state-of-the-art research requires large networks with several million parameters that require hours to be trained. Despite this, the evaluated metrics showed much better results for our network (approximately 99% overall accuracy), in comparison with previous works barely achieving 81% OA over UAV imagery. This notable OA was similarly observed over satellite data. These results demonstrate the efficiency and robustness of our proposed method across different hyperspectral data sources.

Multiscale reweighted smoothing regularization in curvelet domain for hyperspectral image denoising

ABSTRACT Hyperspectral images are easily contaminated by multi-modal noise in the process of data acquisition. Generally, hyperspectral signatures and noise appear unexpectedly scattered in the spatial-spectral domain. Conversely, in the transform domain, the principle components of an image are gathered in the low-frequency bands, while the details and noise often occur in high-frequency parts. The existing denoising methods usually divide the whole frequency domain into two parts, that is, high frequency and low frequency, which can not make the most of the knowledge contained at different frequencies or scales, especially in the multiscale transform domain. Furthermore, aiming to restore a clean image from the noise-corrupted data, hyperspectral image denoising often amounts to an ill-posed problem, in which reshaping cubic data into matrix form results in the loss of structure information. To address the aforementioned issues, this paper incorporates the reweighted smoothing regularization in the transform domain into the noise-perturbed degradation model to reformulate a novel optimization problem. First, the multiscale curvelet is leveraged to transform the spatial matrix into frequency domain, where the coefficients are carefully designed to weight the components at different scales and orientations. Second, l 1 -norm is imposed on the abundance term to promote sparsity in the transform domain for performance enhancement. Finally, using proximal alternating optimization, an efficient algorithm is well developed to obtain the closed-form solutions to the proposed denoising problem. The experimental results on several synthetic and real datasets, illustrate that the proposed method improves PSNR by at least 1.5 dB and reduces ERGAS error by more than 35% than the state-of-the-art approaches under the multi-modal noise mixed environments, which verifies the validity of regularization terms in hyperspectral images denoising.

Using hyperspectral signatures for predicting foliar nitrogen and calcium content of tissue cultured little-leaf mockorange (Philadelphus microphyllus A. Gray) shoots

Determining foliar mineral status of tissue cultured shoots can be costly and time consuming, yet hyperspectral signatures might be useful for determining mineral contents of these shoots. In this study, hyperspectral signatures were acquired from tissue cultured little-leaf mockorange (Philadelphus microphillus) shoots to determine the feasibility of using this technology to predict foliar nitrogen and calcium contents. After using a spectroradiometer to take hyperspectral images for determining foliar N and Ca contents, the correlation between the hyperspectral bands, vegetation indices, and hyperspectral features were calculated from the spectra. Features with high correlations were selected to develop the models via different regression methods including linear, random forest (RF), and support vector machines. The results showed that non-linear regression models developed through machine learning techniques, including RF methods and support vector machines provided satisfactory prediction models with high R2 values (%N by RF with R2 = 0.72, and %Ca by RF with R2 = 0.99), that can estimate nitrogen and calcium content of little-leaf mockorange shoots grown in vitro. Overall, the RF regression method provided the most accurate and satisfactory models for both foliar N and Ca estimation of little-leaf mockorange shoots grown in tissue culture.

Raman hyperspectroscopy of saliva and machine learning for Sjögren’s disease diagnostics

Sjögren's disease is an autoimmune disorder affecting exocrine glands, causing dry eyes and mouth and other morbidities. Polypharmacy or a history of radiation to the head and neck can also lead to dry mouth. Sjogren's disease is often underdiagnosed due to its non-specific symptoms, limited awareness among healthcare professionals, and the complexity of diagnostic criteria, limiting the ability to provide therapy early. Current diagnostic methods suffer from limitations including the variation in individuals, the absence of a single diagnostic marker, and the low sensitivity and specificity, high cost, complexity, and invasiveness of current procedures. Here we utilized Raman hyperspectroscopy combined with machine learning to develop a novel screening test for Sjögren's disease. The method effectively distinguished Sjögren's disease patients from healthy controls and radiation patients. This technique shows potential for development of a single non-invasive, efficient, rapid, and inexpensive medical screening test for Sjögren's disease using a Raman hyper-spectral signature.

Oil Spill Classification Using an Autoencoder and Hyperspectral Technology

Hyperspectral technology has been playing a leading role in monitoring oil spills in marine environments, which is an issue of international concern. In the case of monitoring oil spills in local areas, hyperspectral technology of small dimensions is the ideal solution. This research explores the use of encoded hyperspectral signatures to develop automated classifiers capable of discriminating between polluted and clean water and distinguishing between various types of oil. The overall objective is to leverage these classifiers to be able to improve the performance of conventional systems that rely solely on hyperspectral imagery. The acquisition of the hyperspectral signatures of water and hydrocarbons was carried out with a spectroradiometer. The range of the spectroradiometer used in this study covers the ranges between [350–1000] (visible near-infrared) and [1000–2500] (short-wavelength infrared). This gives detailed information regarding the targets of interest. Different neural autoencoders (AEs) have been developed to reduce inputs into different dimensions, from 1 to 15. Each of these encoded sets was used to train decision tree (DT) classifiers. The results are very promising, as they show that the AE models encoded data with correlation coefficients above 0.95. The classifiers trained with the different sets provide accuracies close to 1.

Mapping Leaf Area Index at Various Rice Growth Stages in Southern India Using Airborne Hyperspectral Remote Sensing

Globally, rice is one of the most important staple food crops. The most significant metric for evaluating the rice growth and productivity is the Leaf Area Index (LAI), which can be effectively monitored using remote sensing data. Hyperspectral remote sensing provides contiguous bands at narrow wavelengths for mapping LAI at various rice phenological stages, and it is functionally related to canopy spectral reflectance. Hyperspectral signatures for different phases of rice crop growth was recorded using Airborne Visible Near-Infrared Imaging Spectrometer-Next Generation (AVIRIS-NG) along with corresponding ground based observations. Ground-based hyperspectral canopy spectral reflectance measurements were recorded with FieldSpec 3 Hi-Res spectroradiometer (ASD Inc., Forsyth County, GA, USA; spectral range: 350–2500 nm) and LAI data from 132 farmer’s fields in Southern India. Among 29 hyperspectral vegetation indices tested, 8 were found promising for mapping rice LAI at various phenological stages. Among all the growth stages, the elongation stage was the most accurately estimated using vegetation indices that exhibited a significant correlation with the airborne hyperspectral reflectance. The validation of hyperspectral vegetation indices revealed that the best fit model for estimating rice LAI was mND705 (red-edge, blue, and NIR bands) at seedling and elongation, SAVI (red and NIR bands) at tillering and WDRVI (red and NIR bands) at booting stage.

Space object identification and classification from hyperspectral material analysis

This paper presents a data processing pipeline designed to extract information from the hyperspectral signature of unknown space objects. The methodology proposed in this paper determines the material composition of space objects from single pixel images. Two techniques are used for material identification and classification: one based on machine learning and the other based on a least square match with a library of known spectra. From this information, a supervised machine learning algorithm is used to classify the object into one of several categories based on the detection of materials on the object. The behaviour of the material classification methods is investigated under non-ideal circumstances, to determine the effect of weathered materials, and the behaviour when the training library is missing a material that is present in the object being observed. Finally the paper will present some preliminary results on the identification and classification of space objects.

Noisy band selection based on the integration of the Stacked-Autoencoder and Convolutional Neural Network approaches for hyperspectral data.

The presence of noise on hyperspectral images causes degradation and hinders efficiency of processing for land cover classification. In this sense, removing noise or detecting noisy bands automatically on hyperspectral images becomes a challenge for research in remote sensing. To cope this problem, an integrated model (SAE-1DCNN) is presented in this study, based on Stacked-Autoencoders (SAE) and Convolutional Neural Networks (CNN) algorithms for the selection and exclusion of noisy bands. The proposed model employs convolutional layers to improve the performance of autoencoders focused on discriminating the training data by analyzing the hyperspectral signature of the pixel. Thus, in the SAE-1DCNN model, information can be compressed, and then redundant information can be detected and extracted by taking advantage of the efficiency of the deep architecture based on the convolutional and pooling layers. Hyperspectral data from the AVIRIS (Airborne Visible/Infrared Imaging Spectrometer) sensor were used to evaluate the performance of the proposed automatic method based on feature selection. The results showed effectiveness to identify noisy bands automatically, suggesting that the proposed methodology was found to be promising and can be an alternative to identify noisy bands within the scope of hyperspectral data pre-processing.
 Keywords: noisy bands; feature selection; convolutional neural network; stacked-autoencoders; hyperspectral data

A deep scalable neural architecture for soil properties estimation from spectral information

In this paper we propose an adaptive deep neural architecture for the prediction of multiple soil characteristics from the analysis of hyperspectral signatures. The proposed method overcomes the limitations of previous methods in the state of art: (i) it allows to predict multiple soil variables at once; (ii) it is based on a flexible neural architecture capable of automatically adapting to the spectral library under analysis. As an additional feature, our method permits to backtrace the spectral bands that most contribute to the estimation of a given variable. The proposed architecture is experimented on LUCAS, a large laboratory dataset and on a dataset achieved by simulating PRISMA hyperspectral sensor. Results, compared with other state-of-the-art methods confirm the effectiveness of the proposed solution. In fact, the proposed method achieves an overall R2 of about 0.75. In detail, pHCaCl2 and CaCO3 are predicted with an R2 above 0.90, pHH2O, OC, N, clay, sand, CEC, silt are predicted with an R2 between 0.89 and 0.70, K and P with an R2 respectively of 0.54 and 0.37. The proposed method overcomes, in terms of R2, the state of the art of about 15% on the LUCAS dataset and of about 24% on the PRISMA simulated dataset.

SERS of Human Red Blood Cells in Non-Resonant Conditions: Benefits, Limitations, and Complementary Tools (CytoViva and GFAAS)

Herein, Raman and surface-enhanced Raman spectroscopies (SERS) were successfully employed to establish the chemical interactions of citrate-capped silver nanoparticles (AgNPs, 10–15 nm) with human red blood cells (RBCs). The Raman/SERS spectra offered spectral evidence for the cellular uptake of AgNPs and the subsequent change in the conformation of the most abundant component, hemoglobin (Hb), from oxyhemoglobin to deoxyhemoglobin. The spectral characterization of AgNPs’ interactions with other RBC biomarkers (membrane proteins and lipids) was impeded by the dominant Hb bands, even for non-resonant Hb conditions. CytoViva hyperspectral imaging and graphite furnace atomic absorption spectroscopy (GFAAS) served as complementary tools to effectively address the challenges related to a single excitation line (632.8 nm) and the resolution of the confocal Raman microscope (0.5–1.0 µm). CytoViva confirmed the RBC-AgNP interactions through hyperspectral signatures and facilitated the label-free localization of AgNPs extracellularly and intracellularly. Irreversible agglutination of RBCs was noted after 24 h of exposure, raising concerns about the toxicity of AgNPs of biocompatible citrate coatings. GFAAS validated the Raman/SERS results by quantifying the proportion of AgNPs absorbed by RBCs, which was significant (~48% AgNPs by mass), mostly at the membrane (60% RBCs), and size dependent (no large AgNPs or AgNP-aggregates in RBCs, after 12–24 h).

Denoising of Hyperspectral Signal from Drone for Ganoderma Disease Detection in Oil Palm

Oil palm is an important crop that generates high income to Malaysia. However, the oil palm is susceptible to Ganoderma infection that reduces the productivity of the oil palm. Conventional ground-based disease detection is laborious and costly. Therefore, airborne remote sensing technology coupled with ground detection provides a more effective control of the disease. Airborne hyperspectral remote sensing utilizes narrow and contiguous bands to assist in detection of diseases in crops. Spectral responses recorded by the camera tend to suffer from interference and these noises could reduce the quality of the data. Therefore, this study presents the application of Savitzky-Golay and wavelet spectral denoising technique to improve the hyperspectral signatures for Ganoderma disease detection in oil palm

Coating Condition Detection and Assessment on the Steel Girder of a Bridge through Hyperspectral Imaging

The organic coating of bridge steel girders is subjected to physical scratches, corrosion, and aging in natural weathering. The breakdown of the coating may cause serviceability and safety problems if left unnoticed. Conventional coating inspection is time-consuming and lacks information about the coating’s chemical integrity. A hyperspectral imaging method is proposed to detect the condition of steel coatings based on coating-responsive features in reflectance spectra. A field test was conducted on the real-world bridge, which shows obvious signs of degradation. The hyperspectral signature enables an assessment of the coating’s health and defect severity. The results indicated that the coating scratch can be effectively located in the domain of a hyperspectral image and the scratch depth can be determined by mapping a scratch depth indicator (SDI = R532 nm/R641 nm). Rust sources and products in steel girders can be identified by the unique spectral signatures in the VNIR range, and the rust stains (and thus stain areas) scattered on the coating can be pinpointed at pixel level by the chloride rust (CR) indicators >1.11 (CR = R733 nm/R841 nm). The chemical integrity of a topcoat is demonstrated by the short-wave infrared spectroscopy and the topcoat degradation can be evaluated by the decreased absorption at 8000 cm−1 and 5850 cm−1. Hyperspectral imaging enables faster and more reliable coating condition detection by the spectral features and provides an alternative for multi-object coating detection.

Hyperspectral Signatures and Volumetric Study of Iron Ore from Kemmanagundi, Bababudan Hill Ranges of Karnataka State, India

Iron is one among the oldest metals known to man and is quite rich in our country. 95% of the total iron reserves of India were accounted from few states such as Jharkhand, Orissa, Madhya Pradesh, Chhattisgarh, Karnataka and Goa. Kudremukha and Donimalai is known for producing high grades of iron in Karnataka state. The present study aims to analyze the spectral signatures and volumetric determinations of iron ore deposits which were well developed and was being mined in Kemmanagundi of Bababudan hills in Chikmagalur district. Field random samples were collected for laboratory work and analysis in determining the spectral signature curves and volumetric content. The spectral signatures of any mineral will discriminated based on their characteristics features of absorption and reflectance curves. Volumetric analysis is generally one of the quantitative methods that deal with determination of amount of particular elements in tested sample. As demand of iron in India is expected to increase day by day this deposit has to be considered as an important one for its rich iron content of 69.8 percent of iron content occurring around Kemmanagundi of Bababudanhill range, Chikmagalur district, Karnataka. The final results show the presence of rich iron content around the study area.

Hyperspectral Non-Imaging Measurements and Perceptron Neural Network for Pre-Harvesting Assessment of Damage Degree Caused by Septoria/Stagonospora Blotch Diseases of Wheat

The detection and identification of plant diseases is a fundamental task for sustainable crop production. Septoria tritici and Stagonospora nodorum blotch (STB and SNB) are two of the most common diseases of cereal crops that cause significant economic damage. Both pathogens are difficult to identify at early stages of infection. Determining the degree of the disease at a late infection stage is useful for assessing cereal crops before harvesting, as it allows the assessment of potential yield losses. Hyperspectral sensing could allow for automatic recognition of Septoria harmfulness on wheat in field conditions. In this research, we aimed to collect information on the hyperspectral data on wheat plants with different lesion degrees of STB&SNB and to create and train a neural network for the detection of lesions on leaves and ears caused by STB&SNB infection at the late stage of disease development. Spring wheat was artificially infected twice with Septoria pathogens in the stem elongation stage and in the heading stage. Hyperspectral reflections and brightness measurements were collected in the field on wheat leaves and ears on the 37th day after STB and the 30th day after SNB pathogen inoculation using an Ocean Insight “Flame” VIS-NIR hyperspectrometer. Obtained non-imaging data were pre-treated, and the perceptron model neural network (PNN) was created and trained based on a pairwise comparison of datasets for healthy and diseased plants. Both statistical and neural network approaches showed the high quality of the differentiation between healthy and damaged wheat plants by the hyperspectral signature. A comparison of the results of visual recognition and automatic STB&SNB estimation showed that the neural network was equally effective in the quality of the disease definition. The PNN, based on a neuron model of hyperspectral signature with a spectral step of 6 nm and 2000–4000 value datasets, showed a high quality of detection of the STB&SNB severity. There were 0.99 accuracy, 0.94 precision, 0.89 recall and 0.91 F-score metrics of the PNN model after 10,000 learning epochs. The estimation accuracy of diseased/healthy leaves ranged from 88.1 to 97.7% for different datasets. The accuracy of detection of a light and medium degree of disease was lower (38–66%). This method of non-imaging hyperspectral signature classification could be useful for the identification of the STB and SNB lesion degree identification in field conditions for pre-harvesting crop estimation.

Genome-wide association analysis of hyperspectral reflectance data to dissect the genetic architecture of growth-related traits in maize under plant growth-promoting bacteria inoculation.

Plant growth-promoting bacteria (PGPB) may be of use for increasing crop yield and plant resilience to biotic and abiotic stressors. Using hyperspectral reflectance data to assess growth-related traits may shed light on the underlying genetics as such data can help assess biochemical and physiological traits. This study aimed to integrate hyperspectral reflectance data with genome-wide association analyses to examine maize growth-related traits under PGPB inoculation. A total of 360 inbred maize lines with 13,826 single nucleotide polymorphisms (SNPs) were evaluated with and without PGPB inoculation; 150 hyperspectral wavelength reflectances at 386-1021 nm and 131 hyperspectral indices were used in the analysis. Plant height, stalk diameter, and shoot dry mass were measured manually. Overall, hyperspectral signatures produced similar or higher genomic heritability estimates than those of manually measured phenotypes, and they were genetically correlated with manually measured phenotypes. Furthermore, several hyperspectral reflectance values and spectral indices were identified by genome-wide association analysis as potential markers for growth-related traits under PGPB inoculation. Eight SNPs were detected, which were commonly associated with manually measured and hyperspectral phenotypes. Different genomic regions were found for plant growth and hyperspectral phenotypes between with and without PGPB inoculation. Moreover, the hyperspectral phenotypes were associated with genes previously reported as candidates for nitrogen uptake efficiency, tolerance to abiotic stressors, and kernel size. In addition, a Shiny web application was developed to explore multiphenotype genome-wide association results interactively. Taken together, our results demonstrate the usefulness of hyperspectral-based phenotyping for studying maize growth-related traits in response to PGPB inoculation.