Spectral Features Research Articles

Time-frequency analysis of femtosecond CARS spectroscopy of N2 and O2 using the superlet transform.

Molecular dynamics plays a crucial role in understanding molecular interactions, rovibrational coupling mechanisms, and energy transfer processes. Femtosecond time-resolved coherent anti-Stokes Raman scattering spectroscopy was employed to study the molecular dynamics of N2 and O2 in air at room temperature. To reveal hidden spectral features, we have for the first time applied an analytical method that balances time resolution and frequency resolution, namely, the superlet transform (SLT), to perform time-frequency resolved spectral analysis of the complex molecular dynamics of N2 and O2 in air. A distinct evolution of the partial rotational modes of N2 and O2 outside the selective excitation region was observed, which is related to energy transfer collisions between N2 and O2 molecules during the rotational energy relaxation process in air. The SLT results accord well with the S-branch rotational spectra of N2 and O2 obtained from theoretical calculations, confirming the validity of SLT analysis. This method provides a valuable experimental analysis technique to deepen the understanding of the microscopic dynamic processes in molecular dynamics.

Detection of freely moving thoughts using SVM and EEG signals

Objective.Freely moving thought is a type of thinking that shifts from one topic to another without any overarching direction or aim. The ability to detect when freely moving thought occurs may help us promote its beneficial outcomes, such as for creative thinking and positive mood. Thus far, no studies have used machine learning to detect freely moving thought on the basis of 'objective' (e.g. neural or behavioral) data.Approach.Our study addresses this gap, using event-related potential (ERP) and spectral features of electroencephalogram (EEG) signals as well as behavioral measures during a simple attention task and machine learning to detect freely moving thought. EEG features were first examined with both inter-subject and intra-subject strategies. Specifically, the statistical and entropy features of the P3 ERP and alpha spectral measures were entered as inputs to the support vector machine. The best combination of EEG features achieving higher classification performance in both strategies were then selected to combine with behavioral features to further enhance classification performance.Main results.Our best performing model has a Matthew's correlation coefficient and area under the curve of 0.3105 and 0.6665 for inter-subject models and 0.2815 and 0.6407 for intra-subject models respectively.Significance.The above chance level performance in both strategies using EEG and behavioral features shows great promise for machine learning approaches to detect freely moving thought and highlights their potential for real-time prediction in the real world. This has important implications for enhancing creative processes and mood associated with freely moving thought.

Monitoring kidney microanatomy during ischemia-reperfusion using ANFIS optimized CNN.

Kidney disease is a dangerous disease that affects human health and causes various defects. Renal microbiological changes can be monitored using optical coherence tomography (OCT) images to identify the nature of the disease based on behavior during ischemia-reperfusion. Image analysis becomes the more sophisticated part of extracting information from feature dependencies from objects to identify the disease. Most methodologies use feature correlation dependencies in non-relation feature analysis-based disease identification with low precision and recall level. So, classification accuracy needs to be higher performance. To resolve this problem, we proposed the adaptive neuro-fuzzy inference system-based Resnet50 optimal convolutional neural network (ANFIS-CNN) method implemented using deep learning (DL) to monitor kidney disease. Initially, we analyze using OCT images collected from a standard repository. Furthermore, bidirectional filters can be used for preprocessing to reduce image noise. Gaussian filtering can be applied to identify the dependence of kidney structure. Afterward, the color density saturation can be analyzed through edge-based segmentation using the histogram equalization method, and the optimally extracted objects can be identified through edge-based segmentation. These spectral value-based relative feature detection thresholds are combined with texture point-based recursive spectral multiscale feature selection (RSMFS) to produce different entity contrasts. Then, spectral values are optimized with ANFIS-Resnet50 optimal CNN to classify the accuracy by selecting images. Moreover, the proposed method results in high classification accuracy up to 96.1 %, recall rate 95.18 % and precision up to 96.09 % well attained, enhancing their overall performance. The system develops high-performance image recognition for kidney disease monitoring.

Machine learning predicts spinal cord stimulation surgery outcomes and reveals novel neural markers for chronic pain

Spinal cord stimulation (SCS) is a well-accepted therapy for refractory chronic pain. However, predicting responders remain a challenge due to a lack of objective pain biomarkers. The present study applies machine learning to predict which patients will respond to SCS based on intraoperative electroencephalogram (EEG) data and recognized outcome measures. The study included 20 chronic pain patients who were undergoing SCS surgery. During intraoperative monitoring, EEG signals were recorded under SCS OFF (baseline) and ON conditions, including tonic and high density (HD) stimulation. Once spectral EEG features were extracted during offline analysis, principal component analysis (PCA) and a recursive feature elimination approach were used for feature selection. A subset of EEG features, clinical characteristics of the patients and preoperative patient reported outcome measures (PROMs) were used to build a predictive model. Responders and nonresponders were grouped based on 50% reduction in 3-month postoperative Numeric Rating Scale (NRS) scores. The two groups had no statistically significant differences with respect to demographics (including age, diagnosis, and pain location) or PROMs, except for the postoperative NRS (worst pain: p = 0.028; average pain: p < 0.001) and Oswestry Disability Index scores (ODI, p = 0.030). Alpha-theta peak power ratio differed significantly between CP3-CP4 and T3-T4 (p = 0.019), with the lowest activity in CP3-CP4 during tonic stimulation. The decision tree model performed best, achieving 88.2% accuracy, an F1 score of 0.857, and an area under the curve (AUC) of the receiver operating characteristic (ROC) of 0.879. Our findings suggest that combination of subjective self-reports, intraoperatively obtained EEGs, and well-designed machine learning algorithms might be potentially used to distinguish responders and nonresponders. Machine and deep learning hold enormous potential to predict patient responses to SCS therapy resulting in refined patient selection and improved patient outcomes.

DeepQuadrature: universal convolutional neural network enlarging space-bandwidth product in single-shot fringe pattern optical metrology

Abstract The representative of state-of-the-art full-field optical metrology techniques are fringe-pattern-based methods, where the measurand is encoded in the phase distribution of the recorded spatially-periodic intensity pattern. The phase demodulation process is a crucial part of a measurement that defines the overall achieved accuracy, and its algorithm imposes one of the main information throughput limitations on the entire experimental unit. In this paper, we propose a novel solution that increases the phase space-bandwidth product of an optical system in a purely numerical way. It is based on deep learning and a convolutional neural network (CNN) called DeepQuadrature. It uses two images as the input data: a prefiltered fringe pattern (denoised and detrended by, e.g., variational image decomposition) and a local orientation map (modulo 2π). The output of the network is the quadrature function (input fringe pattern shifted in phase by π/2). Finally, the phase distribution is calculated via the arctangent function of the input and output fringe patterns. The working principle is somewhat motivated by the state-of-the-art in high space-bandwidth product classical phase estimation called Hilbert spiral transform, however, with novel adaptive spectral features promoting DeepQuadrature as a versatile and universally applicable method. Phase decoding results for widespread simulated and experimental (testing technical and biological object) fringe data compare favorably with classical algorithms: learning-free Variational Hilbert Quantitative Phase Imaging and the reference multi-frame method. DeepQuadrature thus opens new possibilities in fringe-based high-content optical metrology, and is poised to advance biomedical and technical label-free microscopy.

Title: Identification of Redox State Based on the Difference in Solvation Dynamics.

Oxidation-reduction (Redox) reactions are crucial for many biological processes, yet there is no method available to evaluate redox states in a non-invasive, continuous manner. Here we introduce a novel approach to distinguish between reduced and oxidized states of glutathione (GSH and GSSG, respectively) using aquaphotomics near-infrared (NIR) spectroscopy and multivariate analysis. We identified clear differences in NIR spectra reflecting not only glutathione itself, but different redox states of glutathione based on the spectral features of water molecular conformations interacting with the reaction site. Molecular dynamic simulations also revealed the difference in water molecule coordination and hydration numbers around the reaction site. This approach not only sheds light on the significance of water molecules in redox reactions but also enables non-destructive, continuous assessment of redox states, with potential applications for bioreactor optimization.

Graph vertex and spectral features for EEG-based motor imagery classification.

Graph vertex and spectral features for EEG-based motor imagery classification.

Decoding intrinsic fluctuations of engagement from EEG signals during fingertip motor tasks.

Maintaining a high mental engagement is critical for motor rehabilitation interventions. Achieving a flow experience, often conceptualized as a highly engaged mental state, is an ideal goal for motor rehabilitation tasks. This paper proposes a virtual reality-based fine fingertip motor task in which the difficulty is maintained to match individual abilities. The aim of this study is to decode the intrinsic fluctuations of flow experience from electroencephalogram (EEG) signals during the execution of a motor task, addressing a gap in flow research that overlooks these fluctuations. To resolve the conflict between sparse self-reported flow sampling and the high dimensionality of neural signals, we use motor behavioral measures to represent flow and label the EEG data, thereby increasing the number of samples. A machine learning-based neural decoder is then established to classify each trial into high-flow or low-flow using spectral power and coherence features extracted from the EEG signals. Cross-validation reveals that the classification accuracy of the neural decoder can exceed 80%. Notably, we highlight the contributions of high-frequency bands in EEG activities to flow decoding. Additionally, EEG feature analyses reveal significant increases in the power of parietal-occipital electrodes and global coherence values, specifically in the alpha and beta bands, during high-flow durations. This study validates the feasibility of decoding the intrinsic flow fluctuations during fine motor task execution with a high accuracy. The methodology and findings in this work lay a foundation for future applications in manipulating flow experience and enhancing engagement levels in motor rehabilitation practice.

Conventional versus AI-based spectral data processing and classification approaches to enhance LIBS's analytical performance.

Laser-Induced Breakdown Spectroscopy (LIBS) combined with Artificial Intelligence (AI) offers a powerful method for analyzing and comparing spectral data. This study presents a comparative analysis of conventional and AI-developed methods for processing and interpreting LIBS data, especially in forensic applications, focusing on toner sample discrimination. We propose a novel AI-developed approach that combines normalization, interpolation, and peak detection techniques to simplify LIBS spectral analysis without user preprocessing and easily identify unique spectral features. This method was compared with conventional principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA), which are commonly used for LIBS data analysis. The AI-developed method demonstrated superior performance in discriminating between toner samples from various brands and models of printers and photocopiers. The quantitative evaluation of the performance of the AI-developed approach was performed using statistical analysis, including accuracy difference percentage, component-wise variance analysis, paired t-test, and cross-validation test. The results confirmed a significant improvement in accuracy with the AI-developed method compared to conventional approaches. This proposed work highlights the potential of AI in enhancing spectroscopic analysis for forensic applications, offering increased efficiency and accuracy in sample discrimination and classification. Additionally, it accelerates the analysis of LIBS data with no need for user preprocessing.

Multispectral Land Surface Reflectance Reconstruction Based on Non-Negative Matrix Factorization: Bridging Spectral Resolution Gaps for GRASP TROPOMI BRDF Product in Visible

In satellite remote sensing, mixed pixels commonly arise in medium- and low-resolution imagery, where surface reflectance is a combination of various land cover types. The widely adopted linear mixing model enables the decomposition of mixed pixels into constituent endmembers, effectively bridging spectral resolution gaps by retrieving the spectral properties of individual land cover types. This study introduces a method to enhance multispectral surface reflectance data by reconstructing additional spectral information, particularly in the visible spectral range, using the TROPOMI BRDF product generated by the Generalized Retrieval of Atmosphere and Surface Properties (GRASP) algorithm. Employing non-negative matrix factorization (NMF), the approach extracts spectral basis vectors from reference spectral libraries and reconstructs key spectral features using a limited number of wavelength bands. The comprehensive test results show that this method is particularly effective in supplementing surface reflectance information for specific wavelengths where gas absorption is strong or atmospheric correction errors are significant, demonstrating its applicability not only within the 400–800 nm range but also across the broader spectral range of 400–2400 nm. While not a substitute for hyperspectral observations, this approach provides a cost-effective means to address spectral resolution gaps in multispectral datasets, facilitating improved surface characterization and environmental monitoring. Future research will focus on refining spectral libraries, improving reconstruction accuracy, and expanding the spectral range to enhance the applicability and robustness of the method for diverse remote sensing applications.

From 1-D to 3-D: LIBS Pseudohyperspectral Data Cube Deep Learning Mechanism Used in Nuclear Metal Materials Classification.

In this paper, we propose a new spectral data mechanism called LIBS pseudohyperspectral data cube. This mechanism allows for the utilization of multidimensional information from laser-induced plasma, transforming 1-D LIBS spectra into 3-D data cube. Specifically, two additional dimensions are introduced to capture spectral variations information, allowing more features to be learned during pretraining. Proposed mechanism can make the LIBS system more robust when handling unstable spectra acquired onsite, and can also allow LIBS take full advantage of deep learning algorithms. In the context of nuclear power plants, traditional LIBS classification faces significant challenges due to unstable spectra, which reduce the accuracy of classifying similar or tiny extreme condition materials. By combining deep learning algorithms with LIBS pseudohyperspectral data cube, we can capture spectral and other dimensional features to enhance classification accuracy. Experimental results show that, compared to traditional 1-D data processing, the new method significantly improves the classification accuracy of unstable spectra. Moreover, by incorporating an attention mechanism, the model can adaptively adjust the weights of different features, further improving classification accuracy to over 99%. Visualizing the attention mechanism's weight matrix allows us to identify the importance of different features in classification. Additionally, t-SNE visualizations demonstrate the clustering of different categories in the feature space, further validating the performance of the new method. We believe this data cube mechanism offers an effective new approach for applying deep learning algorithms and enhancing data dimensionality in the LIBS field.

Safe-by-Design Strategies for Intranasal Drug Delivery Systems: Machine and Deep Learning Solutions to Differentiate Epithelial Tissues via Attenuated Total Reflection Fourier Transform Infrared Spectroscopy.

The development of nasal drug delivery systems requires advanced analytical techniques and tools that allow for distinguishing between the nose-to-brain epithelial tissues with better precision, where traditional bioanalytical methods frequently fail. In this study, attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy is coupled to machine learning (ML) and deep learning (DL) techniques to discriminate effectively between epithelial tissues. The primary goal of this work was to develop Safe-by-Design models for intranasal drug delivery using ex vivo pig tissues experiment, which were analyzed by way of ML modeling. We compiled an ATR-FTIR spectral data set from olfactory epithelium (OE), respiratory epithelium (RE), and tracheal tissues. The data set was used to train and test different ML algorithms. Accuracy, sensitivity, specificity, and F1 score metrics were used to evaluate optimized model performance and their abilities to identify specific spectral signatures relevant to each tissue type. The used feedforward neural network (FNN) has shown 0.99 accuracy, indicating that it had performed a discrimination with a high level of trueness estimates, without overfitting, unlike the built support vector machine (SVM) model. Important spectral features detailing the assignment and site of two-dimensional (2D) protein structures per tissue type were determined by the SHapley Additive exPlanations (SHAP) value analysis of the FNN model. Furthermore, a denoising autoencoder was built to improve spectral quality by reducing noise, as confirmed by higher Pearson correlation coefficients for denoised spectra. The combination of spectroscopic analysis with ML modeling offers a promising strategy called, Safe-by-Design, as a monitoring strategy for intranasal drug delivery systems, also for designing the analysis of tissue for diagnosis purposes.

Many-Body Configurational Spectral Splitting between a Trion and a Charged Exciton in a Monolayer Semiconductor.

Many-body complexes in semiconductors are important for both fundamental physics and practical device applications. A three-body system of two electrons (e) and one hole (h) or one electron and two holes (2e1h or 1e2h) is commonly believed to form a trion (or a charged exciton) with a spectral peak red-shifted from an exciton. However, both the validity of this understanding and the physical meaning of a trion or charged exciton have not been thoroughly examined. In general, there are two different configurations for a three-body system, <e><eh> or <eeh> (alternatively <eh><h> or <ehh>), which could be considered a charged exciton and trion, respectively. Here, <···> represents an irreducible cluster with respect to Coulomb interactions. In this article, we consider these issues theoretically and experimentally using monolayer MoTe2 as an example. Experimentally, the photoluminescence spectrum showed two spectral peaks that are 21 and 4 meV below the exciton peak, in contrast to the single "trion" peak from the conventional understanding. Theoretically, the three-body Bethe-Salpeter equation in a two-band model reproduced both spectral features, while the cluster expansion technique allows us to further identify the two peaks with the charged exciton <e><eh> (<eh><h>) and the trion <eeh> (<ehh>). Importantly, the spectral splitting is a pure many-body splitting and should not be confused with the fine structure of the trion due to spin-split. Additionally, our theory could also explain similar spectral features in previous experiments on MoSe2, demonstrating the universality of the many-body configurational splitting. Our results provide a more complete understanding of many-body systems.

MGHT: a multi-granularity hyperspectral image classification method based on hybrid Transformer

Convolutional neural network is widely used in hyperspectral image classification due to its excellent feature extraction ability, but it also suffers from the problems of limited global receptive field and insufficient utilization of fine-grained spectral information. This study proposes a multi-granularity hyperspectral image classification model based on hybrid Transformer to address the identified challenge by using the Transformer’s ability to capture global dependencies in images. The model first extracts image features with different granularity sizes by the lightweight multi-granularity spectral spatial feature extraction module and then combines with the multi-head spatial–spectral attention module in Transformer to effectively capture global spectral and spatial feature relationships of the images, generating deep fusion features, thereby expanding the global receptive field and utilizing the rich spectral information in the hyperspectral images to improve the classification performance of the model. Experimental validation was conducted on three representative datasets, Indian Pines, Pavia, and Houston 2013, demonstrating the success of our approach in terms of performance superiority over other methods.

Enhancement of transverse Kerr effect in a BiYIG based magneto optical device with metal nanopyramid arrays

The magneto-optical device based on a periodic array of metal nanopyramids embedded in a film of Bi:YIG coated on the surface of a silver substrate, is optimized utilizing a computational simulation technique (CST Microwave Studio). Under coupling conditions through the two-dimensional grating, the TM-guided modes with narrow resonance are excited in the Bi:YIG film by the incident light, increasing hence the light-matter interaction. Such coupling results as the dips in the reflectance spectrum. The transverse magneto-optical Kerr effect (TMOKE) for a similar structure with a periodic array of rectangular nano-silver is around 4%, hence an enhancement with one order of magnitude was achieved through the proposed device. The surface plasmon polariton modes with the asymmetric narrowband spectral feature are presented in the diffraction of light by a grating in the near-wavelength regime for the TMOKE case, particularly the high reflectivity () and the high-quality factor resonances (). Hence, it would be a promising device for magnetic field sensing and even biosensing applications.

A Fusion XGBoost Approach for Large-Scale Monitoring of Soil Heavy Metal in Farmland Using Hyperspectral Imagery

Heavy metal pollution of farmland is worsened by the excessive introduction of heavy metal elements into soil systems, posing a substantial threat for global food security and human health. The traditional laboratory-based methods for monitoring soil heavy metals are limited for large-scale applications, while hyperspectral imagery data-based methods still face accuracy challenges. Therefore, a fusion XGBoost model based on the superposition of ensemble learning and packaging methods is proposed for large-scale monitoring with high accuracy of soil heavy metal using hyperspectral imagery. We took Xiong’an New Area, Hebei Province, as the study area, and acquired heavy metal content using chemical analysis. The XGB-Boruta-PCC algorithm was used for precise feature selection to obtain the final modeled spectral response features. On this basis, the performance indicators of the Optuna-optimized XGBoost model were compared with traditional linear and nonlinear models. The optimal model was extended to the entire region for drawing the spatial distribution map of soil heavy metal content. The results suggested that the XGB-Boruta-PCC method effectively achieved double dimensionality reduction of high-dimensional hyperspectral data, extracting spectral response features with a high contribution, which, combined with the XGBoost model, exhibited greater general estimation accuracies for heavy metal (Pb) in farmland (i.e., Pb: R2 = 0.82, RMSE = 11.58, MAE = 9.89). The results of the mapping indicated that there were exceedances for the southwest and parts of the west over the research region. Factories and human activities were the potential causes of heavy metal Pb contamination in farmland. In conclusion, this innovative method can quickly and accurately achieve monitoring large-scale soil heavy metal contamination in farmland, with ZY-1-02E spaceborne hyperspectral imagery proving to be a reliable tool for mapping soil heavy metal in farmland.

Ecological risk evaluation of heavy metals based on hyperspectral: a case study of rice paddy soil in Xiangtan County, China

Heavy metal pollution in agricultural soils pose significant risks to food safety and human health, necessitating effective monitoring and assessment techniques. Recognizing characteristic spectral features is essential for accurately assessing the ecological risks associated with heavy metal pollution. To explore the potential of utilizing the hyperspectrum to evaluate ecologial risks of heavy metals pollution in paddy soil, the study investigated ecological risks and spectral characteristics (350 ~ 2500 nm) of heavy metals (Cd, Pb, Cr, Cu) in rice paddy soil in Xiangtan County, Hunan Province, China. Potential ecological risk index (RI) was employed to evaluate pollution level, and derivative transformations (first derivative R′ and second derivative R″) were applied to eliminate the hyperspectral noise. Results identified Cd as the primary pollutant in the paddy soil of Xiangtan County, followed by Pb, Cu, and Cr. Serious heavy metal pollution concentrated in riverine and cropland areas. After applying spectral transformations, the correlation between spectral reflectance and ecological risks significantly improved. The R″ transformation demonstrated the highest correlations for Cd (r = 0.95) at 2046 nm in high ecological risk category, while the R′ transformation exhibited the strongest negative correlation for Cd (r = − 0.94) at 2059 nm. This study highlights the potential of hyperspectral analysis as an effective tool for estimating ecological risks posed by heavy metal pollution in rice paddy soils. The findings provide valuable insights for improving monitoring practices and informing risk management strategies in agricultural environments.

Investigation of Peanut Leaf Spot Detection Using Superpixel Unmixing Technology for Hyperspectral UAV Images

Leaf spot disease significantly impacts peanut growth. Timely, effective, and accurate monitoring of leaf spot severity is crucial for high-yield and high-quality peanut production. Hyperspectral technology from unmanned aerial vehicles (UAVs) is widely employed for disease detection in agricultural fields, but the low spatial resolution of imagery affects accuracy. In this study, peanuts with varying levels of leaf spot disease were detected using hyperspectral images from UAVs. Spectral features of crops and backgrounds were extracted using simple linear iterative clustering (SLIC), the homogeneity index, and k-means clustering. Abundance estimation was conducted using fully constrained least squares based on a distance strategy (D-FCLS), and crop regions were extracted through threshold segmentation. Disease severity was determined based on the average spectral reflectance of crop regions, utilizing classifiers such as XGBoost, the MLP, and the GA-SVM. Results indicate that crop spectra extracted using the superpixel-based unmixing method effectively captured spectral variability, leading to more accurate disease detection. By optimizing threshold values, a better balance between completeness and the internal variability of crop regions was achieved, allowing for the precise extraction of crop regions. Compared to other unmixing methods and manual visual interpretation techniques, the proposed method achieved excellent results, with an overall accuracy of 89.08% and a Kappa coefficient of 85.42% for the GA-SVM classifier. This method provides an objective, efficient, and accurate solution for detecting peanut leaf spot disease, offering technical support for field management with promising practical applications.

Lung Cancer Detection: Advancing CT Image Analysis Through Hybrid Bidirectional Long Short-Term Memory and Recurrent Neural Network

Globally, lung cancer (LC) is the leading cause of death from cancer. Medical image analysis based on deep learning (DL) has strong potential for detecting and diagnosing lung cancer by identifying early symptoms with image aid from positron emission tomography (PET) and computed tomography (CT). The majority of DL models created for LC detection are very resource-intensive, requiring a great deal of computational capacity; hence, they pose a challenge to deployment on a standard clinical system and are therefore significantly less accessible in resource-constrained settings. This additional computational load may delay diagnosis and treatment, thus affecting the outcome of the patients. Therein lies the critical need for developing more lightweight and efficient deep learning models that ensure high accuracy while reducing computational requirements. This manuscript presents a Lung Cancer Detection technique, LCD-CT-BiLSTM-RNN, based on advanced CT image analysis. First, noise reduction in the lung CT images by anisotropic guided filtering (AGF) is performed. Then, adaptive fuzzy K-means clustering (AFKMC) separates the affected areas of cancer, and Synchroextracting Transform (SET) adds the spectral features. Finally, a hybrid BiLSTM and RNN architecture runs the classification task with an improved overall accuracy. Hybrid optimization using Slime Mould Optimization (SMO) and Golden Eagle Optimization (GEO) fine-tunes the model. The performance of the methods is assessed using MATLAB's accuracy, precision, recall, F1-score, specificity, Matthews Correlation Coefficient (MCC), and ROC to compare the acquired findings with the existing approaches. The performance of the proposed method provides 2.03%, 3.45%, and 2.36% higher accuracy compared with existing techniques like Fuzzy Particle Swarm Optimization with Convolutional Neural Network for Detection of LC (FPSO-CNN), Deep learning Instantaneously Accomplished Neural Network with Improved Profuse Clustering Technique for LC detection (DITNN-IPCT), and Residual Learning Denoising Model with Convolutional Neural Network (DR-Net-CNN) for the Detection of LC.

JCMT Observations of CO, SO2, and a U-line in the Circumstellar Envelopes of Four O-rich AGB Stars

Abstract The circumstellar envelopes (CSE) of asymptotic giant branch (AGB) stars are abundant in molecular emissions, offering valuable insights into the physical and chemical conditions of these evolving stars. In this paper, we report observations of two molecules (CO and SO2) toward four O-rich AGB stars using the James Clerk Maxwell Telescope (JCMT). We detected an unusual SO2 spectral feature comprising both broad and narrow components in IK Tau and Ap Lyn. The broad line profiles may originate from thermal molecular emission, while the narrow profiles could come from other species (or masers) or astrophysical phenomena occurring within the CSEs of the AGB stars, such as episodic mass loss, bipolar outflows, or emissions associated with the complex physical processes near the central star. The narrow lines of SO2 may also arise from vibrationally excited emissions. Additionally, we observed the same U-line in both TX Cam and IK Tau, which may originate from the molecule N17O. We analyzed the identified molecular lines using rotational diagrams to determine their excitation temperatures, column densities, and fractional abundances. This information aids in the constructing of reliable astrochemical models for a more detailed examination of the target stars. The narrow component of the SO2 line suggests unusual astrophysical phenomena, making IK Tau and Ap Lyn particularly intriguing for further investigation to fully understand the physical processes at play in these sources.