Dongseo University

This study develops and evaluates a comprehensive deep-learning framework for early detection of Autism Spectrum Disorder (ASD) through facial image analysis. Five state-of-the-art convolutional neural network (CNN) architectures, VGG16, VGG19, ResNet50, InceptionV3, and MobileNet, were systematically assessed using a balanced dataset of 5,000 images (2,500 ASD, 2,500 non-ASD). Transfer learning and data augmentation enhanced model generalization. VGG19 achieved the highest overall accuracy (77.89%) and F1-score (0.7962), ResNet50 attained the best precision (82.53%), and InceptionV3 produced the highest recall (99.67%), indicating strong screening potential. The findings confirm that deep CNNs can capture subtle facial morphological cues linked to ASD, supporting their feasibility as non-invasive diagnostic tools. This work provides a benchmark for future multimodal, explainable, and clinically validated AI systems for autism detection.

In urban areas, the pavement rate increases significantly; therefore, even under the same rainfall conditions, flooding increases. In addition, when heavy rainfall occurs and drainage system capacity is limited, the watershed tends to be flooded overall, including low-lying areas. Despite many efforts to reduce flooding, the damage continues to recur. As a structural measure to reduce flooding, relatively large reservoirs or pumping stations are installed in upper- and middle-stream areas or at river discharge points. However, due to the characteristics of the urban drainage system, inundation can occur across the watershed, which limits the effectiveness of these measures in reducing overall flooding. This study proposes a distributed installation procedure for small-scale reservoirs, considering economic feasibility and flood reduction, using a Pareto optimization approach that considers flooding characteristics. In this study, the multi-objective optimization technique applied to the Pareto optimization approach is Genetic Algorithms (GA). The objective functions are the minimum cost and maximum flood reduction rate, and the Storm Water Management Model (SWMM) was adopted for runoff simulation. As a result, the size and location of optimal reservoirs representing maximum flood reduction at minimum cost could be selected using the Pareto optimization approach.

Pulsed electric field technology possesses a high potential and a bright future in food processing to inactivate microorganisms and reduce enzymatic activity. Processed food shows a higher retention of health-related compounds and an extension of the shelf-life compared to conventional pasteurization methods. This technology is gradually moving from the laboratory and pilot-plant scale to the commercial scale. In the current review, we focus on the way existing knowledge on mathematical modeling and computational approaches is structured, connected, and interpreted across scales. We start with the electroporation models, progressing from those that are derived from simple physical and chemical considerations to those that are based on more complex probabilistic approaches. They attempt to predict how electric pulses create pores in cell membranes and form the basis of kinetic inactivation models. Subsequently, we examine the most common kinetic models of microorganism inactivation, from first-order models to models based on random and probabilistic considerations. We then review the works carried out on the numerical simulations of the electric field in a continuous PEF chamber and the works related to coupled simulations of the electric, fluid flow, temperature, and inactivation kinetic field. Finally, we conclude the manuscript with a section dedicated to the current applications of the PEF process to demonstrate its effectiveness.

1.1. , . . 222 () , .

Introduction: Glucose serves as the brain’s primary energy substrate and plays a critical role in maintaining cognitive function. Even slight fluctuations in glucose levels can influence attention, learning, and emotional regulation. Materials and Methods: The authors organised a screening of PubMed and Google Scholar reference databases on relevant articles from 1995 till 2025 to perform a systematic review. Results: Glucose metabolism and its regulation in the brain: The brain consumes a significant portion of the body’s glucose, especially in cortical and subcortical regions involved in cognitive processing. Glucose regulation is mediated by neuroglia, particularly astrocytes, which are essential for energy transfer and utilization. Neurobehavioral and neuroimaging markers: Neuroimaging studies using fMRI and PET reveal distinct patterns of activation in the brain ’ s reward system when comparing glucose and fructose consumption. Glucose more strongly activates the hypothalamus and regions related to satiety and appetite regulation, whereas fructose is associated with a reduced sense of fullness. Effects of sweet substances on cognitive function: Comparisons between glucose and non-caloric sweeteners (e.g., sucralose) highlight differences in the engagement of dopaminergic pathways and reward anticipation mechanisms. Glucose demonstrates a more stable effect on short-term memory and attention. Role of incretins and calorie awareness: Gut hormones such as GLP-1 are involved in satiety signaling and influence activity in limbic structures. Additionally, cognitive awareness of caloric content and food composition can modulate neural responses, underscoring the importance of cognitive factors in regulating eating behavior. Conclusion: Understanding the differences in perception and processing of glucose, fructose, and non-caloric sweeteners is crucial for developing dietary strategies and interventions for metabolic and neuropsychiatric disorders. Further research into neuroenergetics and individual variability in nutrient responses may support more precise and personalized approaches in medicine.

Purpose This study investigates the relationship between the ESG (environment, social and governance) modes of companies and their innovation activities from the perspective of intellectual capital. Design/methodology/approach A three-stage process was used to analyze a dataset of 1,267 companies from knowledge-intensive industries. In the first stage, principal component analysis was applied to identify the types of ESG-related activities. Next, using cluster analysis, we identified patterns or modes of ESG-related behavior. Third, the identified clusters were compared in terms of their innovation-related behavior. Findings Compared to companies that demonstrated low ESG engagement, enterprises that have adopted comprehensive ESG strategies were found to be more likely to foster innovation and thus achieve sustainable competitive advantages. Our results highlight the particular importance of relational capital in stakeholder collaboration and knowledge sharing. Research limitations/implications While measuring ESG and innovation at the corporate level presents inherent challenges, this study contributes positively by offering a structured categorization of ESG modes, enabling more precise and meaningful comparisons. Our framework reduces ambiguity in assessing ESG-innovation dynamics and provides a solid foundation for future research. Moreover, the study opens exciting avenues for exploration, such as investigating the causal relationship between ESG engagement and innovation outcomes. Future research can build on these insights to examine how specific ESG models drive distinct innovation characteristics, particularly in developing economies and among SMEs, thereby enriching the global understanding of sustainable corporate strategies. Practical implications The study’s findings provide actionable insights for companies, policymakers, and other stakeholders on strategically integrating ESG and innovation. Businesses can leverage these insights to better align their ESG and innovation strategies, recognizing the critical role of relational and structural capital. Fragmented ESG communication yields limited value compared to cohesive strategies that foster external collaboration. Policymakers, in turn, should link innovation incentives to broader ESG strategies, ensuring a holistic integration that encompasses all ESG dimensions to enhance long-term competitiveness and sustainability. Social implications Adopting a broad ESG approach, rather than focusing on isolated aspects, is crucial for fostering sustainable innovation and generating societal benefits. Companies should assess ESG holistically, considering environmental, social and governance dimensions to maximize synergies with innovation. Stakeholder engagement and external cooperation are essential for developing effective and inclusive ESG initiatives. Furthermore, greater emphasis should be placed on the social dimension of ESG, which remains underdeveloped in many organizations. Future research could explore how small and medium-sized enterprises navigate this relationship and examine regional differences to better understand the social impact of ESG and innovation. Originality/value The results offer meaningful insights for corporate strategy, including the need for a holistic approach to intellectual capital management in ESG-driven innovation. In order to promote long-term sustainability and value creation, policymakers should encourage the alignment of innovation incentives with broader ESG frameworks.

PurposeThe purpose of the study was to evaluate the validity and reliability of a short-form Revised Lifestyle Evaluation Scale for Metabolic Syndrome (RLES-MS)-Korean.MethodsA cross-sectional survey was administered to 211 people with diabetes who were taking medication for metabolic syndrome and attending an outpatient clinic at a university-affiliated hospital in Suwon, Korea. Structured questionnaires were used for the psychometric evaluation. The RLES-MS validity was examined using exploratory factor analysis, and reliability was assessed using Cronbach's alpha and McDonald's omega.ResultsThe interitem correlation analyses identified 2 low-correlation items, resulting in a 15-item scale. Factor analysis identified 4 dimensions of the RLES-MS: healthy eating and being active, monitoring and managing target goals, healthy coping, and awareness and taking medication. The 4-factor model explained 58.65% of the total variance. Item 8 (taking medication) scored the highest, and Item 2 (healthy eating) scored the lowest, followed by Item 13 (monitoring and managing the target values). Overall, Cronbach's alpha and McDonald's omega were .864 and .866, respectively. The RLES-MS showed significant associations with quality of life (r = .426, P < .001), self-efficacy (r = .530, P < .001), depressive symptoms (r = -.430, P < .001), and A1C (r = -.163, P = .018).ConclusionsThe RLES-MS showed good validity and moderate reliability, supporting its clinical use for evaluating lifestyle behaviors in Korean people with diabetes and metabolic syndrome.

As of May 7, 2024, the securities crowdfunding cycle has facilitated the funding of 356 securities, comprising 231 equity securities, 20 debt securities, and 105 sukuk securities. Sukuk experienced an increase in the share of sharia securities funding with a percentage of 30%, up from 22% in 2023. This increase in funding was not accompanied by a specific, legally binding screening standard for sharia securities issuers in the securities offering mechanism. Financial Services Authority or Otoritas Jasa Keuangan (OJK) Regulation Number 35 of 2017 cannot be legally applied as a screening standard for this funding. This research employs a library research methodology with a socio-legal approach to examine this regulatory void. Primary data consists of OJK Regulation No. 35 of 2017 and OJK Regulation No. 57 of 2020, supported by secondary data on offering mechanisms, funding growth, and relevant literature. The study found a lack of specific regulations governing screening standards (core business and financial ratio) for Sharia securities in the SCF cycle. Currently, guidelines are often informally drawn from OJK Regulation No. 35 of 2017, which is legally inappropriate as it is intended for the public-company setting of the Indonesia Stock Exchange. The findings highlight the need for a specific regulation governing screening standards for securities crowdfunding funding. Adopting an economic approach to legal science, this study proposes the development of a specific OJK Regulation (POJK) that provides clear screening criteria and permanent legal consequences for issuers or publishers who violate the Sharia securities criteria.

Digital disruption and AI-driven transformation are reshaping international management at a systemic, organizational, and leadership level. Problem:&amp;nbsp;Existing multinational corporations (MNCs) face growing uncertainty arising from algorithmic governance, global data regulation, geopolitical fragmentation, and virtual workforce restructuring. Gap:&amp;nbsp;Although prior research recognizes AI as a strategic driver, scholarly work insufficiently explains how AI reshapes international leadership competencies and strategic transformation pathways, particularly within emerging-market multinational contexts. Method:&amp;nbsp;This study adopts a conceptual&amp;ndash;comparative approach, integrating international management theories with digital transformation reports, and synthesizes cross-disciplinary literature from AI innovation, global strategy, and digital leadership research. Contribution:&amp;nbsp;The paper develops a three-stage conceptual framework: &amp;ldquo;AI Innovation &amp;rarr; Leadership Adaptation &amp;rarr; Strategic Outcomes,&amp;rdquo; clarifying how AI capabilities reshape leadership roles, competencies, and multinational strategy formation. It further provides policy implications, practice guidelines, and future research directions to support responsible and sustainable human&amp;ndash;AI collaborative global management.