Main Dataset Research Articles

The association of impulsivity with depression and anxiety symptoms: A transdiagnostic network analysis and replication

BackgroundImpulsivity increases the risk for depression and anxiety. However, the granular pathways among them remain unknown. A network approach that moves from disorder-level analysis to symptom-level analysis can provide further understanding of psychopathological mechanisms. In this study, we examined the network structure of impulsivity and separate and comorbid symptoms of depression and anxiety. MethodsRegularized partial-correlation networks were estimated using cross-sectional data from 1047 Chinese participants aged 18–26 years (main dataset, mean age = 21.45 ± 2.01 years) and 325 Chinese participants aged 18–36 years (an independent replication dataset, mean age = 21.49 ± 3.73 years), including impulsivity-depression, impulsivity-anxiety, and impulsivity-depression-anxiety networks. The datasets were collected from 1 June 2023 to 4 August 2023 and from 27 April 2022 to 16 May 2022, respectively. Impulsivity, depression, and anxiety were assessed using Barratt Impulsiveness Scale Version 11, Patient Health Questionnaire-9, and Generalized Anxiety Disorder-7, respectively. Bridge centrality was analyzed, and a network comparison test (NCT) was conducted to investigate the differences between the main dataset and replication dataset. ResultsThe motor impulsivity dimension was revealed to be closely connected with individual symptoms of depression and anxiety regardless of whether they were in separate disorder forms or comorbid forms. In all the networks, motor impulsivity was the most important bridge node. The NCT showed comparable network connectivity and network structure between the main and replication datasets. LimitationsThe use of cross-sectional data limited the inferences about the direction of causality between variables. ConclusionsThese findings elucidate the psychopathological mechanisms underlying how impulsivity functions within depression, anxiety, and comorbidity and support that motor impulsivity is an important risk factor across different mental disorders and is responsible for comorbidity. The implications of these findings are discussed.

Classification of Periapical and Bitewing Radiographs as Periodontally Healthy or Diseased by Deep Learning Algorithms.

Objectives The aim of this artificial intelligence (AI) study was to develop a deep learning algorithm capable of automatically classifying periapical and bitewing radiography images as either periodontally healthy or unhealthy and to assess the algorithm's diagnostic success. Materials and methods The sample of the study consisted of 1120 periapical radiographs (560 periodontally healthy, 560 periodontally unhealthy) and 1498 bitewing radiographs (749 periodontally healthy, 749 periodontally ill). From the main datasets of both radiography types, three sub-datasets were randomly created: a training set (80%), a validation set (10%), and a test set (10%). Using these sub-datasets, a deep learning algorithm was developed with the YOLOv8-cls model (Ultralytics, Los Angeles, California, United States) and trained over 300 epochs. The success of the developed algorithm was evaluated using the confusion matrix method. Results The AI algorithm achieved classification accuracies of 75% or higher for both radiograph types. For bitewing radiographs, the sensitivity, specificity, precision, accuracy, and F1 score values were 0.8243, 0.7162, 0.7439, 0.7703, and 0.7821, respectively. For periapical radiographs, the sensitivity, specificity, precision, accuracy, and F1 score were 0.7500, 0.7500, 0.7500, 0.7500, and 0.7500, respectively. Conclusion The AI models developed in this study demonstrated considerable success in classifying periodontal disease.Future applications may involve employing AI algorithms for assessing periodontal status across various types of radiography images and for automated disease detection.

Identification of spatial clusters of potentially toxic elements in different soil types using unsupervised machine learning and compositional data analysis

Soil health is important, with soil chemical composition data, including potentially toxic elements (PTEs) being one of its conceptual components. This study aims to reveal the spatial distribution patterns of soil PTEs contents, identify their potential sources, and unveil their geochemical associations in Aragatsotn region, Armenia. For that purpose, the contents of Cr, V, Ti, As, Zn, Cu, Co, Fe, Mn, Ba, and Pb were determined using an X-ray fluorescence spectrometer. The mean contents of Cr and As exceeded their upper continental crust by 1.5 and 3.1 times and their maximum acceptable values by 1.5 and 1.5 times. The analysis demonstrated the presence of sites where all these elements displayed comparatively higher values. The combined application of compositional data analysis and geospatial mapping revealed multivariate outliers, which were located in structural-metallogenic zones with active ore exploitation . The application of unsupervised machine learning algorithm unveiled three groups within the main dataset and the clr-biplot identified the source-specific elements. Particularly, Group I included Cu and displayed the highest mean value among the identified groups. The soil samples included in Group I were in areas where Calcisols were developed and comparatively high Cu contents were attributed to agricultural activities and vehicle emissions. Group II is represented by the geochemical association of Fe, Co, Cr, Mn, Zn, and As. The formation of this group is conditioned by volcanic rocks of the local geological origin. However, no spatial pattern was identified in Group II distribution aligned with soil types. Group III included Ti, V, Pb, and Ba, which may have a mixed origin as it is spatially distributed in areas where regional highways pass through and where Group II elements also exhibit their higher values.

Cold Chain Logistics Projects and Performance of Horticulture Export Companies in Rwanda: A Case of Horticulture Exporters Association of Rwanda (HEAR)

Globally, the top three markets for horticulture imports are Europe, the US, and the UAE. The general objective of this study aims to assess the effects of cold chain logistics projects on performance of horticulture exports companies in Rwanda: a case study of Horticulture Exporters association of Rwanda. The specific objectives of this study aim: to assess the effects associated with costs incurred in cold chain logistics on performance of horticulture exports companies in Rwanda, to evaluate the contributions of technology used in cold chain logistics on performance of horticulture exports companies in Rwanda and to determine the effects of geographic location of cold chain stores on performance of horticulture exports companies in Rwanda. Research focused on system theory and resource-based theory. By combining quantitative and qualitative analytic techniques, the researchers used a correlational approach to look for patterns in the interactions between the study's components. The research was based on responses from 92 out of 120 respondents from Rwanda’s horticulture export company. Statistical sampling and other methods were part of the "mixed method" approach utilized in the study. Stratified and random sampling was used to pick the exporters, whilst convenient and purposive sampling were used to select the participants from the cold chain logistics management group. Using both personal and secondary sources, the research aimed to determine the impact of cold chain logistics on the growth of Rwandan horticultural export businesses. For the main data set, which covered the years 2021–2023, researchers utilized closed-ended questions to survey participants. For the secondary data set, they combed through financial records and relevant documents from cold chain facilities. The data was analyzed using version 24 of the Statistical Package for the Social Sciences, a widely used statistical tool in the field of social science. Descriptive data were presented in tables displaying percentages and frequencies. Additionally, regression analysis was conducted to delve deeper into the relationship between the study variables. The findings suggest that there are significant associations between certain factors and performance. Firstly, the cost associated with cold trucks and cold rooms demonstrated a strong positive relationship with performance (β = 0.427, p <0.05). This implies that higher expenditures in maintaining cold trucks and rooms tend to correspond with better performance among horticulture exports companies. Secondly, technology used in cold chain logistics also exhibited a positive relationship with performance, although the effect size was smaller and the p-value slightly higher (β = 0.395, p<0.05). This indicates that employing advanced technology in managing the cold chain logistics process may contribute positively to the performance of these companies. Finally, the geographic location of cold chain stores displayed a significant positive relationship with performance (β = 0.879, p<0.05). The results show that horticultural export companies may boost their performance and profits by investing in cold chain logistics, refrigeration, and holding facilities. If Rwanda wants to make its horticulture exports more competitive, HEAR should weigh the pros and cons of various cold truck and cold room solutions before making a final decision.

СУПУТНИКОВИЙ МОНІТОРИНГУ ОЦІНКИ СТАНУ ЗАБРУДНЕННЯ ВОДНИХ ОБ’ЄКТІВ НА ОСНОВІ МАШИННОГО НАВЧАННЯ

Satellite monitoring is a reliable tool for obtaining operational information about the state of water bodies and their changes with a certain periodicity, thanks to the use of a large set of heterogeneous data. And the availability of archival data becomes a significant advantage, opening up the possibility of comparing the obtained results and analyzing factors affecting the state of water bodies and the environment. The paper describes the main methods and datasets that can be used to apply remote monitoring methods based on machine learning. The use of medium spatial resolution data is proposed as the main data set. Examples of the application of the remote sensing approach to the assessment of water pollution in large areas are analyzed. Possible machine learning methods for solving an applied problem are presented, the new artificial intelligence (AI) approach of convolutional long-short-term memory (ConvLSTM) model significantly outperforms classical methods.

Relationship between monthly surface water derived from Sentinel-2 imagery and meteorological data (precipitation and evaporation) at Baghdad, Iraq

ABSTRACT This study was carried out using two main monthly datasets: satellite images and some meteorological parameters (like precipitation and evaporation) for 2 years (2018 and 2021). Using the fine spatial resolution images acquired from Sentinel-2, digital maps of surface water areas were produced by calculating the Modified Normalized Difference Water Index for all months in QGIS and then classifying them into two main categories (water and non-water bodies). The results show that the largest areas were in January 2021 (77.2 km2) and 40.7 km2 in November 2018, while the lowest (18.7 km2) were in June 2021 and July 2018. Monthly mean values of cumulative precipitation and evaporation were used to find their relationships with surface water areas. The results show that the year 2021 was a severe dry year, whereas the lowest precipitation (25 mm) in that year was associated with higher evaporation losses (3,021.8 mm) during this year. When combined with surface water areas and meteorological parameters, a non-linear relationship between water areas and evaporation was found for both years, and a linear relationship between precipitation and water areas for the wet year of 2018.

Multi-scale terahertz image reconstruction

This paper proposes a novel terahertz (THz) image recovery algorithm and a new THz image dataset is publicly available. Because of transmission noise, artificial errors and problems with diffraction phenomena are among the main problems that cause degradation of THz images. The point spread function (PSF) in real transmission is first obtained using a terahertz imaging system. Experiments can generate degraded images and clean image datasets for supervised training of neural networks via terahertz PSF. A Depthwise Dense Instantiation Normalization Block (DwDIN Block) has been proposed to recover the datasets. We constructed a powerful multi-stage network (DwDINet) at each scale by utilizing the DwDIN module. For input and output, the results of each stage of the network and the original images are taken as input into the next layer in order to enhance the background information. DwDINet achieves state-of-the-art (SOTA) results on the two main THz datasets. To verify the generality of the model, we performed experiments on various image recovery tasks. According to the experimental results, the proposed model is of great potential for the field of low-level image clarification.

Analysis of methods for assessing the degree of water pollution

The most important modern methods for the monitoring of water bodies, their state of pollution and the assessment of water quality are considered. Commonly accepted, standard-ized methods for assessing water quality in water bodies include ground-based sampling, which is untimely and expensive under certain conditions. Best suited for practical use in large areas is the remote approach, which combines ground and satellite measurements to classify changes and detect pollution in closed water bodies. Satellite monitoring is a reliable tool for obtaining operational information on the state of water bodies and their changes with a certain periodicity, thanks to the use of a large set of heterogeneous data. And the availabil-ity of archived data becomes a significant advantage, opening up the possibility of comparing the results obtained and analyzing the factors influencing the state of water bodies and the environment. The paper describes the main methods and datasets that can be used for the ap-plication of remote monitoring methods based on machine learning. The use of Landsat mul-tispectral sensors, including TM (Thematic Mapper), MSS (Multispectral Scanner), ETM (Enhanced Thematic Mapper), and OLI (Operational Land Imager), is proposed as the main data set. Examples of the application of the remote sensing approach to the assessment of wa-ter pollution over large areas will be analyzed. Possible machine learning methods for solv-ing an applied problem are presented.

The deep convolutional networks for the classification of multi-class arrhythmia

An arrhythmia is an irregular heartbeat. Many researchers in the AI field have carried out the automatic classification of arrhythmias, and the issue that has been widely discussed is imbalanced data. A popular technique for overcoming this problem is the synthetic minority oversampling technique (SMOTE) technique. In this paper, the author adds some sampling of data obtained from other datasets into the primary dataset. In this case, the main dataset is the Massachusetts Institute of Technology–Beth Israel Hospital (MIT-BIH) arrhythmia database and an additional dataset from the MIT-BIH supraventricular arrhythmia database. The classification process is carried out with one-dimensional convolutional neural network model (1D-CNN) to perform multiclass and subject-class advancement of medical instrumentation (AAMII) classifications. The results obtained from this study are an accuracy of 99.10% for multiclass and 99.25% for subject-class.

A JWST Project on 47 Tucanae. Overview, Photometry, and Early Spectroscopic Results of M Dwarfs and Observations of Brown Dwarfs* *The main data set presented in this article was obtained from the Mikulski Archive for Space Telescopes (MAST) at the Space Telescope Science Institute. The specific observations analyzed can be accessed via doi:10.17909/6t82-4360.

James Webb Space Telescope (JWST) observations have been demonstrated to be efficient in detecting multiple stellar populations in globular clusters (GCs) in the low-mass regime of M dwarfs. We present an overview, and first results, of different projects that can be explored by using the JWST observations gathered under program GO2560 for 47 Tucanae, the first program entirely devoted to the investigation of multiple populations in very-low-mass stars, which includes spectroscopic data for the faintest GC stars for which spectra are available. Our color–magnitude diagram (CMD) shows some substructures for ultracool stars, including gaps and breaks in slope. In particular, we observe both a gap and a minimum in the F322W2 luminosity function less than 1 mag apart, and discuss which it could be associated with the H-burning limit. We detect stars fainter than this minimum, very likely brown dwarfs. We corroborate the ubiquity of the multiple populations across different masses, from ∼0.1 M ⊙ up to red giants (∼0.8 M ⊙). The oxygen range inferred for the M dwarfs, both from the CMD and from the spectra of two M dwarfs associated with different populations, is similar to that observed for giants. We have not detected any difference between the fractions of stars in distinct populations across stellar masses ≳ 0.1 M ⊙. This work demonstrates the JWST's capability in uncovering multiple populations within M dwarfs and illustrates the possibility to analyze very-low-mass stars in GCs approaching the H-burning limit and the brown-dwarf sequence.

Offline supervised learning v.s. online direct policy optimization: A comparative study and a unified training paradigm for neural network-based optimal feedback control

This work is concerned with solving neural network-based feedback controllers efficiently for optimal control problems. We first conduct a comparative study of two prevalent approaches: offline supervised learning and online direct policy optimization. Albeit the training part of the supervised learning approach is relatively easy, the success of the method heavily depends on the optimal control dataset generated by open-loop optimal control solvers. In contrast, direct policy optimization turns the optimal control problem into an optimization problem directly without any requirement of pre-computing, but the dynamics-related objective can be hard to optimize when the problem is complicated. Our results underscore the superiority of offline supervised learning in terms of both optimality and training time. To overcome the main challenges, dataset and optimization, in the two approaches respectively, we complement them and propose the Pre-train and Fine-tune strategy as a unified training paradigm for optimal feedback control, which further improves the performance and robustness significantly. Our code is accessible at https://github.com/yzhao98/DeepOptimalControl.

Optimization in University Student Timetables: A Comprehensive Literature Review

The University Course Timetabling Problem presents a complex optimization problem as it seeks to meet the diverse requirements and constraints of a university's academic program through efficient timetabling. The objective of this research is to conduct comprehensive review of existing studies on optimization approaches in optimizing university timetable problems. The aim of the comprehensive studies is to further analyse the existing studies to discuss comparative studies to compare the performance of the approach methods in satisfying the constraints involved in university timetabling. The methodology used for this paper is systematic literature review which brought three main themes such constraints, optimization approaches, and datasets involved in University Course Timetabling Problem. The result shown that the majority of approaches involved hybrid approaches. In conclusion, this comprehensive review summarizes the constraints, optimization approach and datasets involved University Course Timetabling Problem in higher educational institutions of ASEAN countries.

Deep learning for automated segmentation and counting of hypocotyl and cotyledon regions in mature Pinus radiata D. Don. somatic embryo images.

In commercial forestry and large-scale plant propagation, the utilization of artificial intelligence techniques for automated somatic embryo analysis has emerged as a highly valuable tool. Notably, image segmentation plays a key role in the automated assessment of mature somatic embryos. However, to date, the application of Convolutional Neural Networks (CNNs) for segmentation of mature somatic embryos remains unexplored. In this study, we present a novel application of CNNs for delineating mature somatic conifer embryos from background and residual proliferating embryogenic tissue and differentiating various morphological regions within the embryos. A semantic segmentation CNN was trained to assign pixels to cotyledon, hypocotyl, and background regions, while an instance segmentation network was trained to detect individual cotyledons for automated counting. The main dataset comprised 275 high-resolution microscopic images of mature Pinus radiata somatic embryos, with 42 images reserved for testing and validation sets. The evaluation of different segmentation methods revealed that semantic segmentation achieved the highest performance averaged across classes, achieving F1 scores of 0.929 and 0.932, with IoU scores of 0.867 and 0.872 for the cotyledon and hypocotyl regions respectively. The instance segmentation approach demonstrated proficiency in accurate detection and counting of the number of cotyledons, as indicated by a mean squared error (MSE) of 0.79 and mean absolute error (MAE) of 0.60. The findings highlight the efficacy of neural network-based methods in accurately segmenting somatic embryos and delineating individual morphological parts, providing additional information compared to previous segmentation techniques. This opens avenues for further analysis, including quantification of morphological characteristics in each region, enabling the identification of features of desirable embryos in large-scale production systems. These advancements contribute to the improvement of automated somatic embryogenesis systems, facilitating efficient and reliable plant propagation for commercial forestry applications.

Large-scale spatial reconstitution of groundwater fluxes in a karst aquifer on the basis of hydrodynamic and hydrodispersive measurements

This article presents a multiphysics-based approach for a large-scale spatial reconstitution of head and tracer tests responses measured in a regional scale karst aquifer in Southern France. The main dataset consists in hydraulic heads measured in 11 wells during 6 weeks and 4 tracer tests. A pump station withdraws groundwater from this aquifer at varying flow rates during days and nights to supply water to Montpellier agglomeration, generating a periodic signal which propagates in the whole aquifer. Some of the measurements show a periodic response, some others not. As a first step, this specificity is exploited to localize spatially the preferential flow paths network with a structural deterministic inversion approach. Then, inversions of the weekly averaged head variations are performed on the first and the last weeks of the dataset (showing different drawdown trends) to assess the hydraulic diffusivity field. Finally, the tracer restitutions at the spring are inverted in order to optimize the conduits diameters associated to the fastest flow paths.The proposed approach permits a satisfying reproduction of the hydraulic heads and tracer measurements. Simulations with the optimized model on a week of the dataset not used in the inversion allows to validate the inversion result. The inverted model suggests that daily periodic responses represent efficient constraints for the localization of the preferential flows, while the integration of weekly head variations provides a characterization of the hydraulic properties between each well and the closest preferential flow paths, and tracer restitutions permit to characterize the velocities in the preferential paths.

An overview of the emotional brain-computer interface

This paper provides a comprehensive review of current research advances in emotional brain-computer interfaces. We introduce an approach to classifying emotions and highlight the two main datasets used for emotion recognition (DEAP and SEED). Subsequently, an extensive analysis of existing emotion recognition methods, both traditional and deep neural network methods, is presented. Finally, we explore the potential benefits of using transfer learning techniques to improve the performance of emotion recognition methods. Various deep neural network models exhibit redundant neural units and complexity, while facing challenges such as reduced computational power and reaction speed, increased storage requirements, and hardware dependency. The authors propose to integrate learned neural network pruning algorithms to simplify complex models, minimise hardware resource requirements without compromising accuracy, and improve operational capabilities with improved discriminants.

Predicting students’ academic progress and related attributes in first-year medical students: an analysis with artificial neural networks and Naïve Bayes

BackgroundDropout and poor academic performance are persistent problems in medical schools in emerging economies. Identifying at-risk students early and knowing the factors that contribute to their success would be useful for designing educational interventions. Educational Data Mining (EDM) methods can identify students at risk of poor academic progress and dropping out. The main goal of this study was to use machine learning models, Artificial Neural Networks (ANN) and Naïve Bayes (NB), to identify first year medical students that succeed academically, using sociodemographic data and academic history.MethodsData from seven cohorts (2011 to 2017) of admitted medical students to the National Autonomous University of Mexico (UNAM) Faculty of Medicine in Mexico City were analysed. Data from 7,976 students (2011 to 2017 cohorts) of the program were included. Information from admission diagnostic exam results, academic history, sociodemographic characteristics and family environment was used. The main dataset included 48 variables. The study followed the general knowledge discovery process: pre-processing, data analysis, and validation. Artificial Neural Networks (ANN) and Naïve Bayes (NB) models were used for data mining analysis.ResultsANNs models had slightly better performance in accuracy, sensitivity, and specificity. Both models had better sensitivity when classifying regular students and better specificity when classifying irregular students. Of the 25 variables with highest predictive value in the Naïve Bayes model, percentage of correct answers in the diagnostic exam was the best variable.ConclusionsBoth ANN and Naïve Bayes methods can be useful for predicting medical students’ academic achievement in an undergraduate program, based on information of their prior knowledge and socio-demographic factors. Although ANN offered slightly superior results, Naïve Bayes made it possible to obtain an in-depth analysis of how the different variables influenced the model. The use of educational data mining techniques and machine learning classification techniques have potential in medical education.

The deep learning applications in IoT-based bio- and medical informatics: a systematic literature review

AbstractNowadays, machine learning (ML) has attained a high level of achievement in many contexts. Considering the significance of ML in medical and bioinformatics owing to its accuracy, many investigators discussed multiple solutions for developing the function of medical and bioinformatics challenges using deep learning (DL) techniques. The importance of DL in Internet of Things (IoT)-based bio- and medical informatics lies in its ability to analyze and interpret large amounts of complex and diverse data in real time, providing insights that can improve healthcare outcomes and increase efficiency in the healthcare industry. Several applications of DL in IoT-based bio- and medical informatics include diagnosis, treatment recommendation, clinical decision support, image analysis, wearable monitoring, and drug discovery. The review aims to comprehensively evaluate and synthesize the existing body of the literature on applying deep learning in the intersection of the IoT with bio- and medical informatics. In this paper, we categorized the most cutting-edge DL solutions for medical and bioinformatics issues into five categories based on the DL technique utilized: convolutional neural network, recurrent neural network, generative adversarial network, multilayer perception, and hybrid methods. A systematic literature review was applied to study each one in terms of effective properties, like the main idea, benefits, drawbacks, methods, simulation environment, and datasets. After that, cutting-edge research on DL approaches and applications for bioinformatics concerns was emphasized. In addition, several challenges that contributed to DL implementation for medical and bioinformatics have been addressed, which are predicted to motivate more studies to develop medical and bioinformatics research progressively. According to the findings, most articles are evaluated using features like accuracy, sensitivity, specificity, F-score, latency, adaptability, and scalability.

Enhancing Orchid Inventory Management with K-Means Clustering: A Case Study in Sales Optimization

In the ornamental industry, particularly in orchid sales, manual inventory management has presented significant challenges. These challenges include excessive stock accumulation, financial pressure due to overstocks, and limited land resources to store overstocks. Utilizing information technology, data mining, specifically the K-Means Clustering algorithm, emerges as a solution. This research aims to optimize orchid inventory management by grouping products into three clusters: "highly sellable," "sellable," and "less sellable." Historical sales data was used as the main dataset, considering parameters such as sales volume, demand patterns, and seasonality. The K-Means Clustering algorithm categorized the orchid products, with 97 products in the "highly sellable" cluster, 68 products in the "sellable" cluster, and 22 products in the "less sellable" cluster. This analysis offers insights. Orchids that are "highly marketable" require higher stock levels to meet demand. "Sellable" varieties require standard stock levels, while "less sellable" types should maintain lower stock levels to avoid overstocking. In conclusion, using k-means clustering for orchid inventory management can optimize sales strategies. Orchids in high demand receive adequate stock, ensuring customer satisfaction. Effective stock allocation improves financial stability by overcoming the problem of overstock in the ornamental industry.

Improvement of emotion recognition methods based on neural networks

This article analyzes the use of microexpressions – subtle facial movements that are difficult for the human eye to notice, and even more difficult to immediately analyze, even specialists in the field do not always succeed in this perfectly, because their speed is only 1/5 to 1/3 of a second, for assessment of psychological state using artificial intelligence methods. The research is aimed at improving the analysis of micro-mimicry for accurate identification of emotions and psychological state. An overview of implemented technological solutions based on CNN was conducted, and a method for their improvement was found. An experimental test conducted on video recordings of people experiencing various emotions showed the high accuracy of the developed method in recognizing emotions and psychological state. Despite the challenges of the scarcity of microexpression datasets and the subtlety of facial movements, the paper presents a CapsuleNet model for microexpression recognition, builds a system architecture, and conducts testing. By combining three main data sets (SMIC, CASME II and SAMM) into a unified cross-database, the method developed in the work tests the possibility of generalization of the model by different subject characteristics. The performance of CapsuleNet, evaluated by cross-baseline benchmarking and Leave-One-Object-Out validation, significantly outperforms the baseline (LBP-TOP) and other improved of an CNN models. The paper shows that the performance of the developed model, determined by unweighted average recall and F1 scores, outperforms both the LBP-TOP baseline and other state-of-the-art CNN models. In a comprehensive microexpression recognition system. First, we process the data to identify the peak frames in the sequences and isolate the face region in these frames. These processed face images are then moved to CapsuleNet for the classification. The results of the work is to develop and complement methods of emotional artificial intelligence, offering new insights into micromimic assessment of psychological states that affect mental health, human-computer interaction, and social robotics. This technology has potential for development and expansion. This is an additional opportunity for companies that work with people and it is important for them to monitor their productivity, as it is directly related to the psychological state.

Unified Model for Children's Brain Image Segmentation with Co-Registration Framework Guided by Longitudinal MRI.

Accurate segmentation of brain structures is crucial for analyzing longitudinal changes in children's brains. However, existing methods are mostly based on models established at a single time-point due to difficulty in obtaining annotated data and dynamic variation of tissue intensity. The main problem with such approaches is that, when conducting longitudinal analysis, images from different time points are segmented by different models, leading to significant variation in estimating development trends. In this paper, we propose a novel unified model with co-registration framework to segment children's brain images covering neonates to preschoolers, which is formulated as two stages. First, to overcome the shortage of annotated data, we propose building gold-standard segmentation with co-registration framework guided by longitudinal data. Second, we construct a unified segmentation model tailored to brain images at 0-6 years old through the introduction of a convolutional network (named SE-VB-Net), which combines our previously proposed VB-Net with Squeeze-and-Excitation (SE) block. Moreover, different from existing methods that only require both T1- and T2-weighted MR images as inputs, our designed model also allows a single T1-weighted MR image as input. The proposed method is evaluated on the main dataset (320 longitudinal subjects with average 2 time-points) and two external datasets (10 cases with 6-month-old and 40 cases with 20-45 weeks, respectively). Results demonstrate that our proposed method achieves a high performance (>92%), even over a single time-point. This means that it is suitable for brain image analysis with large appearance variation, and largely broadens the application scenarios.