Rapid Classification Research Articles

Rapid classification of coffee origin by combining mass spectrometry analysis of coffee aroma with deep learning

Mislabeling the geographical origin of coffee is a prevalent form of fraud. In this study, a rapid, nondestructive, and high-throughput method combining mass spectrometry (MS) analysis and intelligence algorithms to classify coffee origin was developed. Specifically, volatile compounds in coffee aroma were detected using self-aspiration corona discharge ionization mass spectrometry (SACDI–MS), and the acquired MS data were processed using a customized deep learning algorithm to perform origin authentication automatically. To facilitate high-throughput analysis, an air curtain sampling device was designed and coupled with SACDI–MS to prevent volatile mixing and signal overlap. An accuracy of 99.78% was achieved in the classification of coffee samples from six origins at a throughput of 1 s per sample. The proposed approach may be effective in preventing coffee fraud owing to its straightforward operation, rapidity, and high accuracy and thus benefit consumers.

Recognition of Mushrooms and Classification of Edible and Toxic Families using Hardware Implementation of CNN Algorithms on an Embedded system

In this work, new ideas in the realm of picture identification and classification are developed and implemented on hardware. This entails putting new algorithms into practice, whether for color, texture, or shape identification for AI (Artificial Intelligence) and picture recognition applications. We concentrate on identifying edible mushrooms in the harvesting and food manufacturing processes. Our proposal for an embedded system based on a Raspberry-Pi4 type microcomputer employing a combination of hardware and software components has helped with the recognition and classification of items in the image. Our object recognition system is built on a novel neighborhood topology and a cutting-edge kernel function that enables the effective embedding of image processing-related characteristics. We tested the suggested CNN-based object recognition system using a variety of challenging settings, including diverse fungus species, uncontrolled environments, and varying backdrop and illumination conditions. The outcomes were superior to various state-of-the-art outcomes. On the other hand, our contribution relating to the dynamic mode integrates a CNN network to accurately encode the temporal information with an attention mask allowing us to focus on the characteristics of an edible mushroom according to the state of the art, and guarantee the robustness of the recognition. We implemented our algorithm on a Raspberry Pi400-based embedded system connected to a CMOS camera-type image sensor plus an HMI human-machine interface for the instantaneous display of results for the rapid classification of edible and inedible mushrooms.

Fine Classification of Urban Tree Species Based on UAV-Based RGB Imagery and LiDAR Data

Rapid and accurate classification of urban tree species is crucial for the protection and management of urban ecology. However, tree species classification remains a great challenge because of the high spatial heterogeneity and biodiversity. Addressing this challenge, in this study, unmanned aerial vehicle (UAV)-based high-resolution RGB imagery and LiDAR data were utilized to extract seven types of features, including RGB spectral features, texture features, vegetation indexes, HSV spectral features, HSV texture features, height feature, and intensity feature. Seven experiments involving different feature combinations were conducted to classify 10 dominant tree species in urban areas with a Random Forest classifier. Additionally, Plurality Filling was applied to further enhance the accuracy of the results as a post-processing method. The aim was to explore the potential of UAV-based RGB imagery and LiDAR data for tree species classification in urban areas, as well as evaluate the effectiveness of the post-processing method. The results indicated that, compared to using RGB imagery alone, the integrated LiDAR and RGB data could improve the overall accuracy and the Kappa coefficient by 18.49% and 0.22, respectively. Notably, among the features based on RGB, the HSV and its texture features contribute most to the improvement of accuracy. The overall accuracy and Kappa coefficient of the optimal feature combination could achieve 73.74% and 0.70 with the Random Forest classifier, respectively. Additionally, the Plurality Filling method could increase the overall accuracy by 11.76%, which could reach 85.5%. The results of this study confirm the effectiveness of RGB imagery and LiDAR data for urban tree species classification. Consequently, these results could provide a valuable reference for the precise classification of tree species using UAV remote sensing data in urban areas.

A user-friendly AI-based clinical decision support system for rapid detection of pandemic diseases: Covid-19 and Monkeypox

Accurate and rapid diagnosis is a significant factor in reducing incidence rate; especially when the number of people inflicted with a disease is considerably high. In the healthcare sector, the decision-making process might be a complex and error-prone one due to excessive workload, negligence, time restrictions, incorrect or incomplete evaluation of medical reports and analyses, and lack of experience as well as insufficient knowledge and skills. Clinical decision support systems (CDSSs) are those developed to improve effectiveness of decisions by supporting physicians’ decision-making process regarding their patients. In this study, a new artificial intelligence-based CDSS and a user-friendly interface for this system were developed to ensure rapid and accurate detection of pandemic diseases. The proposed CDSS, which is called panCdss, uses hybrid models consisting of the Convolutional Neural Network (CNN) model and Machine Learning (ML) methods in order to detect covid-19 from lung computed tomography (CT) images. Transfer Learning (TL) models were used to detect monkeypox from skin lesion images and covid-19 from chest X-Ray images. The results obtained from these models were evaluated according to accuracy, precision, recall and F1-score performance metrics. Of these models, the ones with the highest classification performance were used in the panCdss. The highest classification values obtained for each dataset were as follows: % 91.71 accuracy, % 92.07 precision, % 90.29 recall and % 91.71 F1-score for covid-19 CT dataset by using CNN+RF hybrid model; % 99.56 accuracy, % 100 precision, % 99.12 recall and % 99.55 F1-score for covid-19 X-ray dataset by using VGG16 model; and % 90.38 accuracy, % 93.32 precision, % 88.11 recall and % 90.64 F1-score for monkeypox dataset by using MobileNetV2. It is believed that panCdss can be successfully employed for rapid and accurate classification of pandemic diseases and can help reduce physicians’ workload. Furthermore, the study showed that the proposed CDSS is an adaptable, flexible and dynamic system that can be practiced not only for the detection of pandemic diseases but also for other diseases. To the authors’ knowledge, this proposed CDSS is the first CDSS developed for pandemic disease detection.

Improved Detection Method for Micro-Targets in Remote Sensing Images

With the exponential growth of remote sensing images in recent years, there has been a significant increase in demand for micro-target detection. Recently, effective detection methods for small targets have emerged; however, for micro-targets (even fewer pixels than small targets), most existing methods are not fully competent in feature extraction, target positioning, and rapid classification. This study proposes an enhanced detection method, especially for micro-targets, in which a combined loss function (consisting of NWD and CIOU) is used instead of a singular CIOU loss function. In addition, the lightweight Content-Aware Reassembly of Features (CARAFE) replaces the original bilinear interpolation upsampling algorithm, and a spatial pyramid structure is added into the network model’s small target layer. The proposed algorithm undergoes training and validation utilizing the benchmark dataset known as AI-TOD. Compared to speed-oriented YOLOv7-tiny, the mAP0.5 and mAP0.5:0.95 of our improved algorithm increased from 42.0% and 16.8% to 48.7% and 18.9%, representing improvements of 6.7% and 2.1%, respectively, while the detection speed was almost equal to that of YOLOv7-tiny. Furthermore, our method was also tested on a dataset of multi-scale targets, which contains small targets, medium targets, and large targets. The results demonstrated that mAP0.5:0.95 increased from “9.8%, 54.8%, and 68.2%” to “12.6%, 55.6%, and 70.1%” for detection across different scales, indicating improvements of 2.8%, 0.8%, and 1.9%, respectively. In summary, the presented method improves detection metrics for micro-targets in various scenarios while satisfying the requirements of detection speed in a real-time system.

Rapid Classification of Serum from Patients with Paracoccidioidomycosis Using Infrared Spectroscopy, Univariate Statistics, and Linear Discriminant Analysis (LDA).

Paracoccidioidomycosis (PCM) is a systemic mycosis that is diagnosed by visualizing the fungus in clinical samples or by other methods, like serological techniques. However, all PCM diagnostic methods have limitations. The aim of this study was to develop a diagnostic tool for PCM based on Fourier transform infrared (FTIR) spectroscopy. A total of 224 serum samples were included: 132 from PCM patients and 92 constituting the control group (50 from healthy blood donors and 42 from patients with other systemic mycoses). Samples were analyzed by attenuated total reflection (ATR) and a t-test was performed to find differences in the spectra of the two groups. The wavenumbers that had p < 0.05 had their diagnostic potential evaluated using receiver operating characteristic (ROC) curves. The spectral region with the lowest p value was used for variable selection through principal component analysis (PCA). The selected variables were used in a linear discriminant analysis (LDA). In univariate analysis, the ROC curves with the best performance were obtained in the region 1551-1095 cm-1. The wavenumber that had the highest AUC value was 1264 cm-1, achieving a sensitivity of 97.73%, specificity of 76.01%, and accuracy of 94.22%. The total separation of groups was obtained in the PCA performed with a spectral range of 1551-1095 cm-1. LDA performed with the eight wavenumbers with the greatest weight from the group discrimination in the PCA obtained 100% accuracy. The methodology proposed here is simple, fast, and highly accurate, proving its potential to be applied in the diagnosis of PCM. The proposed method is more accurate than the currently known diagnostic methods, which is particularly relevant for a neglected tropical mycosis such as paracoccidioidomycosis.

Rapid classification and identification of chemical components in three different Zanthoxylum species by ultra-high-performance-liquid chromatography quadrupole-orbitrap-mass spectrometry.

Zanthoxylum, as a medicinal and edible herbal medicine, has a long history and complex chemical composition. There are many varieties of Zanthoxylum, and there are differences in composition between varieties. In this study, a rapid classification and identification method for the main components of Zanthoxylum was established using ultra-high-performance-liquid chromatography quadrupole-orbitrap-mass spectrometry. The components of Shandong Zanthoxylum bungeanum, Wudu Zanthoxylum bungeanum, and Zanthoxylum schinifolium were identified by studying the characteristic fragmentations and neutral losses of characteristic components. A total of 48 common components and 24 different components were identified and the fragmentation patterns of the main components, such as flavonoids, alkaloids, and organic acids were summarized. These findings provided a reference for the study of pharmacodynamic substance basis and quality control of different varieties of Zanthoxylum.

Analysis of electrode arrangements for brain stroke diagnosis via electrical impedance tomography through numerical computational models.

Objective.Rapid stroke-type classification is crucial for improved prognosis. However, current methods for classification are time-consuming, require expensive equipment, and can only be used in the hospital. One method that has demonstrated promise in a rapid, low-cost, non-invasive approach to stroke diagnosis is electrical impedance tomography (EIT). While EIT for stroke diagnosis has been the topic of several studies in recent years, to date, the impact of electrode placements and arrangements has rarely been analyzed or tested and only in limited scenarios. Optimizing the location and choice of electrodes can have the potential to improve performance and reduce hardware cost and complexity and, most importantly, diagnosis time.Approach.In this study, we analyzed the impact of electrodes in realistic numerical models by (1) investigating the effect of individual electrodes on the resulting simulated EIT boundary measurements and (2) testing the performance of different electrode arrangements using a machine learning classification model.Main results.We found that, as expected, the electrodes deemed most significant in detecting stroke depend on the location of the electrode relative to the stroke lesion, as well as the role of the electrode. Despite this dependence, there are notable electrodes used in the models that are consistently considered to be the most significant across the various stroke lesion locations and various head models. Moreover, we demonstrate that a reduction in the number of electrodes used for the EIT measurements is possible, given that the electrodes are approximately evenly distributed.Significance.In this way, electrode arrangement and location are important variables to consider when improving stroke diagnosis methods using EIT.

Gravity Spy: lessons learned and a path forward

The Gravity Spy project aims to uncover the origins of glitches, transient bursts of noise that hamper analysis of gravitational-wave data. By using both the work of citizen-science volunteers and machine learning algorithms, the Gravity Spy project enables reliable classification of glitches. Citizen science and machine learning are intrinsically coupled within the Gravity Spy framework, with machine learning classifications providing a rapid first-pass classification of the dataset and enabling tiered volunteer training, and volunteer-based classifications verifying the machine classifications, bolstering the machine learning training set and identifying new morphological classes of glitches. These classifications are now routinely used in studies characterizing the performance of the LIGO gravitational-wave detectors. Providing the volunteers with a training framework that teaches them to classify a wide range of glitches, as well as additional tools to aid their investigations of interesting glitches, empowers them to make discoveries of new classes of glitches. This demonstrates that, when giving suitable support, volunteers can go beyond simple classification tasks to identify new features in data at a level comparable to domain experts. The Gravity Spy project is now providing volunteers with more complicated data that includes auxiliary monitors of the detector to identify the root cause of glitches.

Screening radiation-differentially expressed circular RNAs and establishing dose classification models in the human lymphoblastoid cell line AHH-1

Purpose In the event of a large-scale radiological accident, rapid and high-throughput biodosimetry is the most vital basis in medical resource allocation for the prompt treatment of victims. However, the current biodosimeter is yet to be rapid and high-throughput. Studies have shown that ionizing radiation modulates expressions of circular RNAs (circRNAs) in healthy human cell lines and tumor tissue. circRNA expressions can be quantified rapidly and high-throughput. However, whether circRNAs are suitable for early radiation dose classification remains unclear. Methods We employed transcriptome sequencing and bioinformatics analysis to screen for radiation-differentially expressed circRNAs in the human lymphoblastoid cell line AHH-1 at 4 h following exposure to 0, 2, and 5 Gy 60Co γ-rays. The dose–response relationships between differentially expressed circRNA expressions and absorbed doses were investigated using real-time polymerase chain reaction and linear regression analysis at 4 h, 24 h, and 48 h post-exposure to 0, 2, 4, 6, and 8 Gy. Six distinct dose classification models of circRNA panels were established and validated by receiver operating characteristic (ROC) curve analysis. Results A total of 11 radiation-differentially expressed circRNAs were identified and validated. Based on dose–response effects, those circRNAs changed in a dose-responsive or dose-dependent manner were combined into panels A through F at 4 h, 24 h, and 48 h post-irradiation. ROC curve analysis showed that panels A through C had the potential to effectively classify exposed and non-exposed conditions, which area under the curve (AUC) of these three panels were all 1.000, and the associate p values were .009. Panels D through F excellently distinguished between different dose groups (AUC = 0.963–1.000, p < .05). The validation assay showed that panels A through F demonstrated consistent excellence in sensitivity and specificity in dose classification. Conclusions Ionizing radiation can indeed modulate the circRNA expression profile in the human lymphoblastoid cell line AHH-1. The differentially expressed circRNAs exhibit the potential for rapid and high-throughput dose classification.

A railway fastener inspection method based on lightweight network

To ensure the safety of train operations, regular inspection and maintenance of railway fastener is crucial. Currently, computer vision-based fastener inspection has become the primary method. However, to improve detection accuracy, model complexity has been continuously increasing, leading to a decrease in detection efficiency and seriously affecting real-time fastener inspection task. To address this issue, we propose a fastener inspection method based on lightweight network. First, we introduce YOLOv5n-Faster, where this model selects YOLOv5n as the base network for fastener localization and uses partial convolution (PConv) from FasterNet for lightweight design, while using CoordConv to improve localization accuracy, thereby achieving efficient fastener localization task. Then, we perform lightweight design on the overall network structure and Block module of the ConvNeXt-T to propose the fastener state inspection model, Light ConvNeXt, for rapid fastener classification. Experimental results show that our proposed fastener localization model achieves a mean average precision (mAP) of 84.7%, a detection speed of 161 FPS. The fastener state inspection model achieves an accuracy of 99.7%.

Deep learning analysis for rapid detection and classification of household plastics based on Raman spectroscopy

The overuse of plastics releases large amounts of microplastics. These tiny and complex pollutants may cause immeasurable damage to human social life. Raman spectroscopy detection technology is widely used in the detection, identification and analysis of microplastics due to its advantages of fast speed, high sensitivity and non-destructive. In this work, we first recorded the Raman spectra of eight common plastics in daily life. By adjusting parameters such as laser wavelength, laser power, and acquisition time, the Raman data under different acquisition conditions were diversified, and the corresponding Raman spectra were obtained, and a database of eight household plastics was established. Combined with deep learning algorithms, an accurate, fast and simple classification and identification method for 8 types of plastics is established. Firstly, the acquired spectral data were preprocessed for baseline correction and noise reduction, Then, four machine learning algorithms, linear discriminant analysis (LDA), decision tree, support vector machine (SVM) and one-dimensional convolutional neural network (1D-CNN), are used to classify and identify the preprocessed data. The results showed that the classification accuracy of the three machine learning models for the Raman spectra of standard plastic samples were 84%, 93% and 93% respectively. The 1D-CNN model has an accuracy rate of up to 97% for Raman spectroscopy. Our study shows that the combination of Raman spectroscopy detection techniques and deep learning algorithms is a very valuable approach for microplastic classification and identification.

An integrated digital microfluidic electrochemical impedimetric lipopolysaccharide sensor based on toll-like receptor-4 protein

The spread of infectious diseases poses a global threat to human health and the economy. Conventional laboratory-based pathogen detection analytical techniques are reliable, but are labour and time consuming. Decentralized, rapid pathogen detection and classification devices are essential to boost biosecurity efforts and can aid in the advancement of modern medicine. Here, we describe the development of an integrated digital microfluidic (DMF) electrochemical impedimetric sensor for rapid and on-site detection of lipopolysaccharide (LPS), a molecular signature of Gram-negative bacteria. The sensor was fabricated by immobilizing toll-like receptor protein (TLR4) onto a gold sensing electrode that was fabricated on an indium tin oxide (ITO) DMF top plate. The top plate also housed lithographically patterned ITO pseudo-reference and auxiliary electrodes for a three-electrode electrochemical impedance (EIS) detection system. We exploited the unique feature of DMF to manipulate droplets consisting of samples, buffers, wash solutions and reagents to perform automated EIS measurements due to the interaction of TLR4 with LPS. The integrated sensor platform showed a detection limit of 35 ng/mL LPS and a linear range of up to 400 ng/mL. The small size and ease of operation of the integrated system holds great prospect for the development of portable, and automated generic pathogen detection and classification platform for point-of-need applications.

Nondestructive and Rapid Screening of Aflatoxin-Contaminated Single Peanut Kernels Using Field-Portable Spectroscopy Instruments (FT-IR and Raman).

A nondestructive and rapid classification approach was developed for identifying aflatoxin-contaminated single peanut kernels using field-portable vibrational spectroscopy instruments (FT-IR and Raman). Single peanut kernels were either spiked with an aflatoxin solution (30 ppb-400 ppb) or hexane (control), and their spectra were collected via Raman and FT-IR. An uHPLC-MS/MS approach was used to verify the spiking accuracy via determining actual aflatoxin content on the surface of randomly selected peanut samples. Supervised classification using soft independent modeling of class analogies (SIMCA) showed better discrimination between aflatoxin-contaminated (30 ppb-400 ppb) and control peanuts with FT-IR compared with Raman, predicting the external validation samples with 100% accuracy. The accuracy, sensitivity, and specificity of SIMCA models generated with the portable FT-IR device outperformed the methods in other destructive studies reported in the literature, using a variety of vibrational spectroscopy benchtop systems. The discriminating power analysis showed that the bands corresponded to the C=C stretching vibrations of the ring structures of aflatoxins were most significant in explaining the variance in the model, which were also reported for Aspergillus-infected brown rice samples. Field-deployable vibrational spectroscopy devices can enable in situ identification of aflatoxin-contaminated peanuts to assure regulatory compliance as well as cost savings in the production of peanut products.

An interpretable and transferrable vision transformer model for rapid materials spectra classification

An interpretable and transferrable Vision Transformer (ViT) model was developed for classifying individual materials from their XRD and FTIR spectra.

Real-time detection and classification of PFAS using dynamic behaviors at liquid–liquid interfaces

Rapid detection and classification of per- and polyfluoroalkyl substances (PFAS) are important for monitoring their concentrations at potential contamination sites due to their severe impact on environmental and health safety.

Insect diversity estimation in polarimetric lidar.

Identifying flying insects is a significant challenge for biologists. Entomological lidar offers a unique solution, enabling rapid identification and classification in field settings. No other method can match its speed and efficiency in identifying insects in flight. This non-intrusive tool is invaluable for assessing insect biodiversity, informing conservation planning, and evaluating efforts to address declining insect populations. Although the species richness of co-existing insects can reach tens of thousands, current photonic sensors and lidars can differentiate roughly one hundred signal types. While the retrieved number of clusters correlate with Malaise trap diversity estimates, this taxonomic specificity, the number of discernible signal types is currently limited by instrumentation and algorithm sophistication. In this study, we report 32,533 observations of wild flying insects along a 500-meter transect. We report the benefits of lidar polarization bands for differentiating species and compare the performance of two unsupervised clustering algorithms, namely Hierarchical Cluster Analysis and Gaussian Mixture Model. Our analysis shows that polarimetric properties could be partially predicted even with unpolarized light, thus polarimetric lidar bands provide only a minor improvement in specificity. Finally, we use the physical properties of the clustered observations, such as wing beat frequency, daily activity patterns, and spatial distribution, to establish a lower bound for the number of species represented by the differentiated signal types.

Preliminary Study on Rapid Serial Visualization Presentation Multi-Class Target EEG Classification.

Brain-Computer Interface (BCI) provides a direct communication channel between the brain and external devices. After combining with the Rapid Serial Visualization Presentation (RSVP) paradigm, the RSVP-BCI system can be utilized for human vision-based fast information retrieval. Currently only binary classification of single-trial EEG can be achieved, also the research on the multi-class target RSVP is few, which limited information transfer rate and the application scenarios of the system. In this paper, we focus on the RSVP multi-class target image retrieval task that contains two classes of targets for achieving triple classification for RSVP-EEG. Designed two experiments, each containing two tasks with different task difficulties. We recruited 30 subjects to participate in the experiments, collected EEG data, and made the data publicly available. Moreover, we conducted behavioral analysis, ERP analysis, and proposed a model, MDCNet, for EEG classification to study the feasibility of multi-class target RSVP and the impact of task difficulty. The experimental results indicated that (1) RSVP-EEG classification that includes non-target and 2-class target is feasibility; (2) the different targets in the same task will evoke P300 with the same latency and different amplitudes, and the hit rate of the target in EEG classification is positively correlated with its amplitude; (3) the information hidden in the time dimension play an important role in EEG classification; (4) the harder the task is, the latency of P300 is longer. The experimental analysis obtained meaningful results, which provided a theoretical basis for subsequent research.

A Rapid Alloy Classification Method Based on Single-Frequency Eddy Current Sensing

A Rapid Alloy Classification Method Based on Single-Frequency Eddy Current Sensing

Vibrational spectroscopic profiling of biomolecular interactions between oak powdery mildew and oak leaves.

Oak powdery mildew, caused by the biotrophic fungus Erysiphe alphitoides, is a prevalent disease affecting oak trees, such as English oak (Quercus robur). While mature oak populations are generally less susceptible to this disease, it can endanger young oak seedlings and new leaves on mature trees. Although disruptions of photosynthate and carbohydrate translocation have been observed, accurately detecting and understanding the specific biomolecular interactions between the fungus and the leaves of oak trees is currently lacking. Herein, via hybrid Raman spectroscopy combined with an advanced artificial neural network algorithm, the underpinning biomolecular interactions between biological soft matter, i.e., Quercus robur leaves and Erysiphe alphitoides, are investigated and profiled, generating a spectral library and shedding light on the changes induced by fungal infection and the tree's defence response. The adaxial surfaces of oak leaves are categorised based on either the presence or absence of Erysiphe alphitoides mildew and further distinguishing between covered or not covered infected leaf tissues, yielding three disease classes including healthy controls, non-mildew covered and mildew-covered. By analysing spectral changes between each disease category per tissue type, we identified important biomolecular interactions including disruption of chlorophyll in the non-vein and venule tissues, pathogen-induced degradation of cellulose and pectin and tree-initiated lignification of cell walls in response, amongst others, in lateral vein and mid-vein tissues. Via our developed computational algorithm, the underlying biomolecular differences between classes were identified and allowed accurate and rapid classification of disease with high accuracy of 69.6% for non-vein, 73.5% for venule, 82.1% for lateral vein and 85.6% for mid-vein tissues. Interfacial wetting differences between non-mildew covered and mildew-covered tissue were further analysed on the surfaces of non-vein and venule tissue. The overall results demonstrated the ability of Raman spectroscopy, combined with advanced AI, to act as a powerful and specific tool to probe foliar interactions between forest pathogens and host trees with the simultaneous potential to probe and catalogue molecular interactions between biological soft matter, paving the way for exploring similar relations in broader forest tree-pathogen systems.