Visible-short wavelength near infrared hyperspectral imaging coupled with multivariate curve resolution-alternating least squares for diagnosis of breast cancer.

This study investigates the application of visible-short wavelength near-infrared hyperspectral imaging (Vis-SWNIR HSI) in the wavelength range of 400-950 nm and advanced chemometric techniques for diagnosing breast cancer (BC). The research involved 56 ex-vivo samples encompassing both cancerous and non-cancerous breast tissue from females. First, HSI images were analyzed using multivariate curve resolution-alternating least squares (MCR-ALS) to exploit pure spatial and spectral profiles of active components. Then, the MCR-ALS resolved spatial profiles were arranged in a new data matrix for exploration and discrimination between benign and cancerous tissue samples using principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA). The PLS-DA classification accuracy of 82.1 % showed the potential of HSI and chemometrics for non-invasive detection of BC. Additionally, the resolved spectral profiles by MCR-ALS can be used to track the changes in the breast tissue during cancer and treatment. It is concluded that the proposed strategy in this work can effectively differentiate between cancerous and non-cancerous breast tissue and pave the way for further studies and potential clinical implementation of this innovative approach, offering a promising avenue for improving early detection and treatment outcomes in BC patients.

Smartphone video imaging: A versatile, low-cost technology for food authentication.

This study presents a low-cost smartphone-based imaging technique called smartphone video imaging (SVI) to capture short videos of samples that are illuminated by a colour-changing screen. Assisted by artificial intelligence, the study develops new capabilities to make SVI a versatile imaging technique such as the hyperspectral imaging (HSI). SVI enables classification of samples with heterogeneous contents, spatial representation of analyte contents and reconstruction of hyperspectral images from videos. When integrated with a residual neural network, SVI outperforms traditional computer vision methods for ginseng classification. Moreover, the technique effectively maps the spatial distribution of saffron purity in powder mixtures with predictive performance that is comparable to that of HSI. In addition, SVI combined with the U-Net deep learning module can produce high-quality images that closely resemble the target images acquired by HSI. These results suggest that SVI can serve as a consumer-oriented solution for food authentication.

Rapid detection of colored and colorless macro- and micro-plastics in complex environment via near-infrared spectroscopy and machine learning.

To better understand the migration behavior of plastic fragments in the environment, development of rapid non-destructive methods for in-situ identification and characterization of plastic fragments is necessary. However, most of the studies had focused only on colored plastic fragments, ignoring colorless plastic fragments and the effects of different environmental media (backgrounds), thus underestimating their abundance. To address this issue, the present study used near-infrared spectroscopy to compare the identification of colored and colorless plastic fragments based on partial least squares-discriminant analysis (PLS-DA), extreme gradient boost, support vector machine and random forest classifier. The effects of polymer color, type, thickness, and background on the plastic fragments classification were evaluated. PLS-DA presented the best and most stable outcome, with higher robustness and lower misclassification rate. All models frequently misinterpreted colorless plastic fragments and its background when the fragment thickness was less than 0.1mm. A two-stage modeling method, which first distinguishes the plastic types and then identifies colorless plastic fragments that had been misclassified as background, was proposed. The method presented an accuracy higher than 99% in different backgrounds. In summary, this study developed a novel method for rapid and synchronous identification of colored and colorless plastic fragments under complex environmental backgrounds.

Production monitoring and quality characterization of black garlic using Vis-NIR hyperspectral imaging integrated with chemometrics strategies.

As a new deep-processing garlic product with notable health benefits, the accurate discrimination of processing stages and prediction of key physicochemical constituents in black garlic are vital for maintaining product quality. This study proposed a novel method utilizing hyperspectral imaging technology to both rapidly monitor the processing stages and quantitatively predict changes in the key physicochemical constituents during black garlic processing. Multiple methods of noise reduction and feature screening were used to process the acquired hyperspectral information. To differentiate processing stages, pattern recognition methods including linear discriminant analysis (LDA), K-nearest neighbor (KNN), support vector machine classification (SVC) analysis were utilized, achieving a discriminant accuracy of up to 98.46 %. Furthermore, partial least squares regression (PLSR) and support vector machine regression (SVR) analysis were performed to achieve quantitative prediction of the key physicochemical constituents including moisture and 5-HMF. PLSR models outperformed SVR models, with correlation coefficient of prediction of 0.9762 and 0.9744 for moisture and 5-HMF content, respectively. The current study can not only offer an effective approach for quality detection and assessment during black garlic processing, but also have a positive significance for the advancement of black garlic related industries.

Efficient microplastic identification by hyperspectral imaging: A comparative study of spatial resolutions, spectral ranges and classification models to define an optimal analytical protocol.

Microplastics (MPs) pollution is a global and challenging issue, necessitating the development of efficient analytical strategies for their detection to monitor their environmental impact. This study aims to define an optimal analytical protocol for characterizing MPs by hyperspectral imaging (HSI), comparing different setups based on spatial resolution, spectral range and classification models. The investigated MPs include polymers commonly found in the environment, such as polystyrene (PS), polypropylene (PP) and high-density polyethylene (HDPE), subdivided in three size classes (1000-2000 µm, 500-1000 µm, 250-500 µm). Furthermore, MP particles with diameters ranging from 30 to 250 µm were assessed to determine the limit of detection (LOD) in the different configurations. Hyperspectral images were acquired with two spatial resolutions, 150 and 30 µm/pixel, and two spectral ranges, 1000-1700 nm (NIR) and 1000-2500 nm (SWIR). Three classification models, Partial Least Square-Discriminant Analysis (PLS-DA), Error Correction Output Coding-Support Vector Machine (ECOC-SVM) and Neural Network Pattern Recognition (NNPR) were tested on the acquired images. The correctness of these models was evaluated by prediction maps and statistical parameters (Recall, Specificity and Accuracy). The results demonstrated that for MP particles larger than 250 µm, the optimal setup is a spatial resolution of 150 µm/pixel and a spectral range of 1000-1700 nm, utilizing a linear classification model like PLS-DA. This approach offers accurate predictions while being time- and cost-efficient. For MPs smaller than 250 µm, a higher spatial resolution of 30 µm/pixel with a spectral range of 1000-2500 nm and a non-linear classification method like ECOC-SVM is preferable. The LOD is 250 µm for the 150 µm/pixel resolution and ranges from 100 to 200 µm for the 30 µm/pixel resolution. These findings provide a valuable guide for selecting the appropriate HSI acquisition conditions and data processing methods to optimally characterize MPs of different sizes.

Hyperspectral retinal imaging in Alzheimer's disease and age-related macular degeneration: a review.

While Alzheimer's disease and other neurodegenerative diseases have traditionally been viewed as brain disorders, there is growing evidence indicating their manifestation in the eyes as well. The retina, being a developmental extension of the brain, represents the only part of the central nervous system that can be noninvasively imaged at a high spatial resolution. The discovery of the specific pathological hallmarks of Alzheimer's disease in the retina of patients holds great promise for disease diagnosis and monitoring, particularly in the early stages where disease progression can potentially be slowed. Among various retinal imaging methods, hyperspectral imaging has garnered significant attention in this field. It offers a label-free approach to detect disease biomarkers, making it especially valuable for large-scale population screening efforts. In this review, we discuss recent advances in the field and outline the current bottlenecks and enabling technologies that could propel this field toward clinical translation.

Unveiling Thymoma Typing Through Hyperspectral Imaging and Deep Learning.

Thymoma, a rare tumor from thymic epithelial cells, presents diagnostic challenges because of the subjective nature of traditional methods, leading to high false-negative rates and long diagnosis times. This study introduces a thymoma classification technique that integrates hyperspectral imaging with deep learning. We initially capture pathological slice images of thymoma using a hyperspectral camera and delineate regions of interest to extract spectral data. This data undergoes reflectance calibration and noise reduction. Subsequently, we transform the spectral data into two-dimensional images via the Gramian Angular Field (GAF) method. A variant residual network is then utilized to extract features and classify these images. Our results demonstrate that this model significantly enhances classification accuracy and efficiency, achieving an average accuracy of 95%. The method proves highly effective in automated thymoma diagnosis, optimizing data utilization, and feature representation learning.

Rapid identification of moxa wool storage period based on hyperspectral imaging technology and machine learning.

Moxa wool (MW), derived from the dried leaves of A r t e m i s i a a r g y i , plays a significant role in traditional Chinese medicine. However, the quality of MW varies with its storage period, impacting its therapeutic efficacy. Traditional methods for quality detection are limited and destructive. To address this, we propose a non-destructive detection method using hyperspectral imaging technology and machine learning algorithms to accurately identify the storage period of MW. Nevertheless, hyperspectral data poses challenges due to its high dimensionality and redundancy, leading to increased computational complexity. To overcome this, we employed principal component analysis (PCA), competitive adaptive reweighted sampling (CARS), and successive projection algorithm (SPA) for data dimensionality reduction and wavelength selection. The results demonstrate that these techniques significantly enhance the accuracy of MW storage year identification. For Nanyang MW, the CARS+SVM model achieved the highest accuracy rates of 99.8% in the visible-near-infrared (VNIR) range and 99.55% in the shortwave infrared (SWIR) range. Similarly, for Qichun MW, the SPA+SVM model achieved identification accuracies of 99.78% and 99.47% in the VNIR and SWIR ranges, respectively. This research provides valuable insights into the rapid detection of MW quality by indication of storage years and presents a novel approach for quality control of MW in the field of traditional Chinese medicine. The combination of hyperspectral imaging and machine learning offers a promising solution for efficient and accurate MW identification, contributing to the advancement of traditional medicine practices.

Hyperspectral reflectance imaging for visualizing reducing sugar content, moisture, and hollow rate in red ginseng.

Red ginseng (RG) has been traditionally valued in Northeast Asia for its health-enhancing properties. Recent advancements in hyperspectral imaging (HSI) offer a non-destructive, efficient, and reliable method to assess critical quality indicators of RG, such as reducing sugar content (RSC), water content (WC), and hollow rate (HR). This study developed predictive models using HSI technology to monitor these quality indicators over the spectral range of 400-1700 nm. Image features were enhanced using Principal Component Analysis (PCA) and Minimum Noise Fraction (MNF), followed by classification through Spectral Angle Mapping (SAM). The best-performing model for RSC achieved an R2 value of 0.6198 and a root mean square error (RMSE) of 0.013. For WC, the optimal model obtained an R2 value of 0.6555 and an RMSE of 0.014. The spatial distribution of RSC, WC, and HR was effectively visualized, demonstrating the potential of HSI for on-site quality control of RG. This study provides a foundation for real-time, non-invasive monitoring of RG quality, addressing industry needs for rapid and reliable assessment methods.

A new method for the evaluation of makeup coverage using hyperspectral imaging.

Introduction: The coverage of a makeup foundation is a perceived attribute which is not captured by opacity or any other single optical property. As previous instrumental measurements do not allow us to consistently compare one product to another, we have begun exploring new parameters and analysis methods made available by hyperspectral imaging. Presumably, the coverage of makeup comes from the change in color, homogeneity, and evenness over the face after application, and the ability of the product to hide spots and other blemishes. Methods: As a starting point to unravelling this complex topic, we define a homogeneity factor α H F which measures the change in the homogeneity of the spectra using the distribution of spectral angles in the face. We likewise define a spectral shift factor ß S F which indicates the degree of spectral change after product application. To test these new parameters and the overall analysis method, we applied them to the HSI validation dataset which contains data for three makeup foundation products of different coverage levels applied to 9 models. Results: We find that α H F correlates with the sensory ranking of coverage. Similarly, the parameter ß S F correlates with the visible color change induced by the product, and we can map the three products into distinct categories based on their effect on α H F and ß S F . Discussion: Nevertheless, the homogeneity factor α H F does not fully describe coverage, and in the variability in the product effect from model to model we find evidence that we must also account for the relative color difference between the model's skin tone and the product shade among other factors.

Transportable hyperspectral imaging setup based on fast, high-density spectral scanning for in situ quantitative biochemical mapping of fresh tissue biopsies.

Significance: Histopathological examination of surgical biopsies, such as in glioma and glioblastoma resection, is hindered in current clinical practice by the long time required for the laboratory analysis and pathological screening, typically taking several days or even weeks to be completed. Aim: We propose here a transportable, high-density, spectral scanning-based hyperspectral imaging (HSI) setup, named HyperProbe1, that can provide in situ, fast biochemical analysis, and mapping of fresh surgical tissue samples, right after excision, and without the need for fixing, staining nor compromising the integrity of the tissue properties. Approach: HyperProbe1 is based on spectral scanning via supercontinuum laser illumination filtered with acousto-optic tunable filters. Such methodology allows the user to select any number and type of wavelength bands in the visible and near-infrared range between 510 and 900 nm (up to a maximum of 79) and to reconstruct 3D hypercubes composed of high-resolution (4 to 5 µ m ), widefield images ( 0.9 × 0.9 mm 2 ) of the surgical samples, where each pixel is associated with a complete spectrum. Results: The HyperProbe1 setup is here presented and characterized. The system is applied to 11 fresh surgical biopsies of glioma from routine patients, including different grades of tumor classification. Quantitative analysis of the composition of the tissue is performed via fast spectral unmixing to reconstruct the mapping of major biomarkers, such as oxy-( HbO 2 ) and deoxyhemoglobin (HHb), as well as cytochrome-c-oxidase (CCO). We also provided a preliminary attempt to infer tumor classification based on differences in composition in the samples, suggesting the possibility of using lipid content and differential CCO concentrations to distinguish between lower and higher-grade gliomas. Conclusions: A proof of concept of the performances of HyperProbe1 for quantitative, biochemical mapping of surgical biopsies is demonstrated, paving the way for improving current post-surgical, histopathological practice via non-destructive, in situ streamlined screening of fresh tissue samples in a matter of minutes after excision.

A Microscope Setup and Methodology for Capturing Hyperspectral and RGB Histopathological Imaging Databases.

The digitization of pathology departments in hospitals around the world is now a reality. The current commercial solutions applied to digitize histopathological samples consist of a robotic microscope with an RGB-type camera attached to it. This technology is very limited in terms of information captured, as it only works with three spectral bands of the visible electromagnetic spectrum. Therefore, we present an automated system that combines RGB and hyperspectral technology. Throughout this work, the hardware of the system and its components are described along with the developed software and a working methodology to ensure the correct capture of histopathological samples. The software is integrated by the controller of the microscope, which features an autofocus functionality, whole slide scanning with a stitching algorithm, and hyperspectral scanning functionality. As a reference, the time to capture and process a complete sample with 20 regions of high biological interest using the proposed method is estimated at a maximum of 79 min, reducing the time required by a manual operator by at least three times. Both hardware and software can be easily adapted to other systems that might benefit from the advantages of hyperspectral technology.

Innovations in forensic science: Comprehensive review of hyperspectral imaging for bodily fluid analysis.

Hyperspectral imaging (HSI) has become a crucial innovation in forensic science, particularly for analysing bodily fluids. This advanced technology captures both spectral and spatial data across a wide spectrum of wavelengths, offering comprehensive insights into the composition and distribution of bodily fluids found at crime scenes. In this review, we delve into the forensic applications of HSI, emphasizing its role in detecting, identifying, and distinguishing various bodily fluids such as blood, saliva, urine, vaginal fluid, semen, and menstrual blood. We examine the benefits of HSI compared to traditional methods, noting its non-destructive approach, high sensitivity, and capability to differentiate fluids even in complex mixtures. Additionally, we discuss recent advancements in HSI technology and their potential to enhance forensic investigations. This review highlights the importance of HSI as a valuable tool in forensic science, opening new pathways for improving the accuracy and efficiency of crime scene analyses.

Hyperspectral Imaging Database of Human Facial Skin.

The perceived color of human skin is the result of the interaction of environmental lighting with the skin. Only by resorting to human skin spectral reflectance, it is possible to obtain physical outcomes of this interaction. The purpose of this work was to provide a cured and validated database of hyperspectral images of human faces, useful for several applications, such as psychophysics-based research, object recognition, and material modeling. The hyperspectral imaging data from 29 human faces with different skin tones and sexes, under constant lighting and controlled movements, were described and characterized. Each hyperspectral image, which comprised spectral reflectance of the whole face from 400 to 720 nm in 10 nm steps at each pixel, was analyzed between and within nine facial positions located at different areas of the face. Simultaneously, spectral measurements at the same nine facial positions using conventional local point and/or contact devices were used to ascertain the data. It was found that the spectral reflectance profile changed between skin tones, subjects, and facial locations. Important local variations of the spectral reflectance profile showed that extra care is needed when considering average values from conventional devices at the same area of measurement.

Tracing pistachio nuts' origin and irrigation practices through hyperspectral imaging.

Pistachio trees have become a significant global agricultural commodity because their nuts are renowned for their unique flavour and numerous health benefits, contributing to their high demand worldwide. This study explores the application of Hyperspectral Imaging (HSI) and Machine Learning (ML) to determine pistachio nuts' geographic origin and irrigation practices, alongside predicting essential commercial quality and yield parameters. The study was conducted in two Spanish orchards and employed HSI technology to capture spectral data. It used ML models like Partial Least Squares (PLS), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost) for analysis. The results demonstrated high accuracy in classifying pistachios based on origin, with accuracies exceeding 94%, and in assessing water content and colour pigments, where both PLS and SVM models achieved 99% accuracy. The research highlighted distinct spectral signatures associated with different irrigation treatments, particularly in the Near-Infrared (NIR) region, with PLS showing an accuracy of 92%. However, challenges were noted in predicting fruit orientation, while predicting height location within the tree was more successful, reflecting clearer spectral distinctions. Regression models also showed promise, particularly in predicting yield (R2 = 0.89 with PLS) and percentage of blank nuts (R2 = 0.71 with PLS). The correlation analysis revealed key insights, such as an inverse relationship between blank nuts and yield, and a strong correlation between yield and split nuts. Despite challenges in predicting fruit orientation, the research showed promising results in forecasting yield and commercial quality factors, indicating the effectiveness of spectral analysis in optimising pistachio production and sustainability.

Spectral analysis comparison of pushbroom and snapshot hyperspectral cameras for in vivo brain tissues and chromophore identification.

Significance: Hyperspectral imaging sensors have rapidly advanced, aiding in tumor diagnostics for in vivo brain tumors. Linescan cameras effectively distinguish between pathological and healthy tissue, whereas snapshot cameras offer a potential alternative to reduce acquisition time. Aim: Our research compares linescan and snapshot hyperspectral cameras for in vivo brain tissues and chromophore identification. Approach: We compared a linescan pushbroom camera and a snapshot camera using images from 10 patients with various pathologies. Objective comparisons were made using unnormalized and normalized data for healthy and pathological tissues. We utilized the interquartile range (IQR) for the spectral angle mapping (SAM), the goodness-of-fit coefficient (GFC), and the root mean square error (RMSE) within the 659.95 to 951.42 nm range. In addition, we assessed the ability of both cameras to capture tissue chromophores by analyzing absorbance from reflectance information. Results: The SAM metric indicates reduced dispersion and high similarity between cameras for pathological samples, with a 9.68% IQR for normalized data compared with 2.38% for unnormalized data. This pattern is consistent across GFC and RMSE metrics, regardless of tissue type. Moreover, both cameras could identify absorption peaks of certain chromophores. For instance, using the absorbance measurements of the linescan camera, we obtained SAM values below 0.235 for four peaks, regardless of the tissue and type of data under inspection. These peaks are one for cytochrome b in its oxidized form at λ = 422 nm , two for HbO 2 at λ = 542 nm and λ = 576 nm , and one for water at λ = 976 nm . Conclusion: The spectral signatures of the cameras show more similarity with unnormalized data, likely due to snapshot sensor noise, resulting in noisier signatures post-normalization. Comparisons in this study suggest that snapshot cameras might be viable alternatives to linescan cameras for real-time brain tissue identification.

Hyperspectral imaging with deep learning for quantification of tissue hemoglobin, melanin, and scattering.

Significance: Hyperspectral cameras capture spectral information at each pixel in an image. Acquired spectra can be analyzed to estimate quantities of absorbing and scattering components, but the use of traditional fitting algorithms over megapixel images can be computationally intensive. Deep learning algorithms can be trained to rapidly analyze spectral data and can potentially process hyperspectral camera data in real time. Aim: A hyperspectral camera was used to capture 1216 × 1936 pixel wide-field reflectance images of in vivo human tissue at 205 wavelength bands from 420 to 830 nm. Approach: The optical properties of oxyhemoglobin, deoxyhemoglobin, melanin, and scattering were used with multi-layer Monte Carlo models to generate simulated diffuse reflectance spectra for 24,000 random combinations of physiologically relevant tissue components. These spectra were then used to train an artificial neural network (ANN) to predict tissue component concentrations from an input reflectance spectrum. Results: The ANN achieved low root mean square errors in a test set of 6000 independent simulated diffuse reflectance spectra while calculating concentration values more than 4000× faster than a conventional iterative least squares approach. Conclusions: In vivo finger occlusion and gingival abrasion studies demonstrate the ability of this approach to rapidly generate high-resolution images of tissue component concentrations from a hyperspectral dataset acquired from human subjects.

WeedCube: Proximal hyperspectral image dataset of crops and weeds for machine learning applications.

WeedCube dataset consists of hyperspectral images of three crops (canola, soybean, and sugarbeet) and four invasive weeds species (kochia, common waterhemp, redroot pigweed, and common ragweed). Plants were grown in two separate greenhouses and plant canopies were captured from a top-down camera angle. A push-broom hyperspectral sensor in the visible near infrared region of 400-1000 nm was used for data collection. The dataset includes 160 calibrated images. The number of images can be further increased by selection of smaller region of interests (ROIs). Dataset is supplemented by Jupyter Notebook scripts that help in data augmentation, spectral pre-processing, ROI selection for points and images, and data visualization. The primary purpose of this dataset is to support weed classification or identification studies by enhancing existing training datasets and validating the generalization capabilities of existing models. Owing to the three-dimensional (3D) nature of hyperspectral images, this dataset can also be utilized by researchers and educators across various domains for the development and testing of deep learning algorithms, the creation of automated data processing pipelines effective for 3D data, the development of tools for 3D data visualization, the creation of innovative solutions for data compression, and addressing system memory issues associated with high-dimensional data.

Hyperspectral imaging for chloroplast movement detection.

We employed hyperspectral imaging to detect chloroplast positioning and assess its influence on common vegetation indices. In low blue light, chloroplasts move to cell walls perpendicular to the direction of the incident light. In high blue light, chloroplasts exhibit the avoidance response, moving to cell walls parallel to the light direction. Irradiation with high light results in significant changes in leaf reflectance and the shape of the reflectance spectrum. Using mutants with disrupted chloroplast movements, we found that blue-light-induced changes in the reflectance spectrum are mostly due to chloroplast relocations. We trained machine learning methods in the classification of leaves according to the chloroplast positioning, based on the reflectance spectra. The convolutional network showed low levels of misclassification of leaves irradiated with high light even when different species were used for training and testing, suggesting that reflectance spectra may be used to detect chloroplast avoidance in heterogeneous vegetation. We also examined the correlation between chloroplast positioning and values of indices of normalized-difference type for various combinations of wavelengths and identified an index sensitive to chloroplast positioning. We found that values of some of the vegetation indices, including those sensitive to the carotenoid levels, may be altered due to chloroplast rearrangements.

Harnessing UAVs and deep learning for accurate grass weed detection in wheat fields: a study on biomass and yield implications.

Weeds are undesired plants competing with crops for light, nutrients, and water, negatively impacting crop growth. Identifying weeds in wheat fields accurately is important for precise pesticide spraying and targeted weed control. Grass weeds in their early growth stages look very similar to wheat seedlings, making them difficult to identify. In this study, we focused on wheat fields with varying levels of grass weed infestation and used unmanned aerial vehicles (UAVs) to obtain images. By utilizing deep learning algorithms and spectral analysis technology, the weeds were identified and extracted accurately from wheat fields. Our results showed that the precision of weed detection in scattered wheat fields was 91.27% and 87.51% in drilled wheat fields. Compared to areas without weeds, the increase in weed density led to a decrease in wheat biomass, with the maximum biomass decreasing by 71%. The effect of weed density on yield was similar, with the maximum yield decreasing by 4320 kg·ha- 1, a drop of 60%. In this study, a method for monitoring weed occurrence in wheat fields was established, and the effects of weeds on wheat growth in different growth periods and weed densities were studied by accurately extracting weeds from wheat fields. The results can provide a reference for weed control and hazard assessment research.