Anshits, Alexander G

The rapid expansion of the electric vehicle (EV) market has raised significant safety concerns, particularly regarding fires caused by the thermal runaway of lithium-ion batteries. To address this issue, this study investigates the real-time fire detection performance of segmentation-based object detection models for EVs. The evaluated models include YOLOv5-Seg, YOLOv8-Seg, YOLOv11-Seg, Mask R-CNN, and Cascade Mask R-CNN. Performance is analyzed using metrics such as precision, recall, F1-score, mAP50, and FPS. The experimental results reveal that the YOLO-based models outperform Mask R-CNN and Cascade Mask R-CNN across all evaluation metrics. In particular, YOLOv11-Seg demonstrates superior accuracy in delineating fire and smoke boundaries, achieving minimal false positives and high reliability under diverse fire scenarios. Additionally, its real-time processing speed of 136.99 FPS validates its capability for rapid detection and response, even in complex fire environments. Conversely, Mask R-CNN and Cascade Mask R-CNN exhibit suboptimal performance in terms of precision, recall, and FPS, limiting their applicability to real-time fire detection systems. This study establishes YOLO-based segmentation models, particularly the advanced YOLOv11-Seg, as highly effective EV fire detection and response systems.

Background and Objectives: Weight-bearing activities exacerbate pain and fatigue in functional flat foot, with uphill walking presenting additional challenges due to increased external loads. The current study investigates whether 3D-printed customized arch-support insoles can enhance gait variables and ankle alignment during uphill walking. Materials and Methods: Twenty healthy young adults, divided into two groups (normal foot condition (control, n = 10), functional flat foot (FF, n = 10)), walked on a treadmill at a 10% incline under two conditions: wearing shoes alone (shoe) or wearing shoes with 3D-printed customized arch-support insoles (SI). Gait pattern, center of force (COF), and ankle joint angles were analyzed by OptoGait, Tekscan, and Kinovea, respectively. Results: The foot flat phase of the gait pattern was prolonged in individuals with FF compared to the control under both shoe and SI conditions, whereas the propulsive phase was shortened with the SI. Medial deviation of the COF during the propulsive phase, observed in individuals with FF under the shoe condition, was corrected to a more lateral alignment with the SI, resembling the COF alignment of the control. Additionally, individuals with FF under the shoe condition exhibited increased ankle pronation compared to the control, whereas the SI moderated pronation, achieving alignment closer to that of the control. Conclusions: These findings indicate that the 3D-printed customized arch-support insoles can improve gait mechanics and ankle alignment in individuals with FF, particularly under challenging conditions such as uphill walking.

It is essential for achieving green and sustainable economic development by using tax incentives to promote green investment. Using the data from the seventh, eighth, ninth, and tenth Chinese Private Enterprise Surveys (CPESs) conducted by the Private Enterprise Research Group and using China’s corporate income tax reform in 2008 as a quasi-natural experiment, this paper empirically analyses the effect of tax incentives on corporate green investment based on the difference-in-difference models. The research results show that tax incentives can significantly increase corporate green investment. The mechanism test shows that easing financing constraints is an important channel for tax incentives to promote corporate green investment. In addition, the role of tax incentives in promoting green investment varies depending on the type and location of the enterprise. Relatively speaking, tax incentives have a stronger effect in promoting green investment for corporates with low sales revenue, located in the eastern region, heavy pollution, and high innovation capability. By doing placebo tests and changing measurement methods of indicators for robustness tests, the conclusions of this paper are still valid. Therefore, the government should increase tax incentives to better promote corporate green investment.

Patient safety incidents are recognized as significant contributors to patient mortality, thus demanding immediate attention and strategic interventions in healthcare systems. The room-of-error education program serves as a solution, as it provides a case-based learning platform allowing nursing students to identify and resolve medical errors within a controlled environment systematically. This study aimed to identify the context, mechanisms, and outcomes of room-of-error training programs. This study adopted a systematic review methodology aligning with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines. Comprehensive searches were conducted across key databases, including OvidMEDLINE, Embase, Cochrane, and CINAHL, by utilizing specific terms related to healthcare providers, nursing students, room-of-error education, medical errors, simulation training, and virtual intervention. Included studies focused on healthcare providers or students, error recognition, RFE-related training, and randomized or quasi-experimental trials, while exclusion criteria were non-English/Korean studies, non-original articles, abstracts, and qualitative studies. Risk of bias in the selected studies was assessed using the Risk Of Bias In Non-randomized Studies version 2.0 tool. The search strategy yielded 2,447 articles, with eight studies meeting the inclusion criteria. Predominantly quasi-experimental in design, these eight studies primarily focused on nurses as the target population. Simulations were found to be widely integrated into room-of-error programs, emphasizing skill performance and critical thinking. Half of the studies provided preparation time, 37.5% included feedback, and 62.5% covered medication errors, with 87.5% using offline delivery, 62.5% offering individual education, and program durations ranging from 4 to 35 min, with 25% having no time limit for error inspection. Diverse content, including topics such as medication errors and infection control, was found to be delivered through offline or virtual formats and group-based or individual education. The findings provide valuable insights into the characteristics and outcomes of room-of-error training programs for healthcare professionals and students. This study emphasizes the significance of practical, case-based approaches in nursing education to augment knowledge, confidence, and competencies, thereby enhancing patient safety in clinical practice.

In time-series studies involving bearing sensor data, Gaussian noise and white noise techniques are commonly employed to evaluate model robustness. However, these conventional noise techniques are limited in their applicability to real-world industrial environments. This paper proposes three novel noise techniques—electrical interference noise, harmonic noise, and random shock noise—that more accurately reflect the complex noise encountered in industrial settings. Additionally, a new deep learning model, MultiPatchTST, is introduced, demonstrating robust performance under various noise conditions. Experimental results reveal that Gaussian noise has minimal impact on model performance, whereas the proposed noise techniques significantly affect performance, providing a more realistic evaluation of noise robustness. The proposed MultiPatchTST model achieves superior performance across all metrics in the presence of all four noise types, confirming its robustness and reliability.

In the 21st century, the increasing complexity and uncertainty of the global economy have heightened the need for accurate economic forecasting. Per capita GDP, a critical indicator of living standards, economic growth, and productivity, plays a key role in government policy-making, corporate strategy, and investor decisions. However, predicting per capita GDP poses significant challenges due to its sensitivity to various economic and social factors. Traditional methods such as statistical analysis, regression, and time-series models have shown limitations in capturing nonlinear interactions and volatility of economic data. To address these limitations, this study develops a per capita GDP forecasting model based on deep learning, incorporating key macroeconomic variables—the Consumer Price Index (CPI) and unemployment rate (UR)—to enhance predictive accuracy. This study employs five deep-learning regression models (RNN, LSTM, GRU, TCN, and Transformer) applied to real and placebo datasets, each incorporating combinations of CPI and UR. The results demonstrate that deep learning models can effectively capture complex, nonlinear relationships in economic data, significantly improving predictive accuracy compared to traditional models. Among the models, the Transformer consistently achieves the highest R-squared and lowest error values across various metrics (MSE, RMSE, and MSLE), indicating its superior ability to model intricate economic patterns. In addition, including CPI and UR as additional predictors enhances model robustness, with the TCN and Transformer models showing particularly strong performance in capturing short-term economic fluctuations. The findings suggest that the deep learning models, especially the Transformer, offer valuable tools for policymakers and business leaders, providing reliable GDP forecasts that support economic decision-making, resource allocation, and strategic planning. Academically, this study advances the understanding of deep learning applications in economic forecasting, particularly in integrating significant macroeconomic variables for enhanced predictive performance. The developed model is a foundation for informed economic policy and strategic decisions, offering a robust and actionable framework for managing economic uncertainties. This research contributes to theoretical and applied economics, providing insights that bridge academic innovation with practical utility in economic forecasting.

AbstractMelanoma is a serious type of skin cancer that originates from melanocytes. Rodent melanoma models have provided valuable insights into melanoma pathology; however, they often lack applicability to humans owing to genetic, anatomical, physiological, and metabolic differences. Herein, we developed a transgenic porcine melanoma model that closely resembles humans via somatic cell nuclear transfer (SCNT). Our model features the conditional oncogenes cassettes, TP53R167H and human BRAFV600E, controlled by melanocyte-specific CreER recombinase. After SCNT, transgenic embryos developed normally, with the capacity to develop porcine embryonic stem cells. Seven transgenic piglets with oncogene cassettes were born through embryo transfer. We demonstrated that Cre recombination-mediated oncogene activation remarkably triggered the mitogen-activated protein kinase pathway in vitro. Notably, intradermal injection of 4-hydroxytamoxifen activated oncogene cassettes in vivo, resulting in melanocytic lesions resembling hyperpigmented nevi with increased proliferative properties similar to early human melanomas. This melanoma-inducing system, heritably transmitted to offspring, supports large-scale studies. The novel porcine model provides a valuable tool for elucidating melanoma development and metastasis mechanism, advancing translational medicine, and facilitating preclinical evaluation of new anticancer drugs.

Yoo T.H., Woo H.J., Kim S.

Oh Y.J., Xuan X., Jung M., Kim S., Park Y., Woo H.J., Cho J., Kim S.

This study examines the relationship between cosmetic packaging color and consumer attention by analyzing changes in pupil size using eye-tracking technology. A controlled experiment with 25 participants (mean age: 24.7 ± 3 years, 14 males and 11 females) was conducted to investigate the impact of eight packaging colors (black, white, blue, yellow, orange, turquoise, pink, and sky blue) on pupil dilation during gaze fixation and movement. Pupil size data were analyzed using SAS 9.4, with T-tests used to determine significant differences across colors. The results revealed that pink packaging elicited significantly larger pupil sizes during fixation, indicating heightened attention, while black, white, blue, and orange led to smaller pupil sizes when fixated, suggesting greater focus on the surrounding environment rather than the packaging. In contrast, yellow and turquoise exhibited no significant differences in pupil size during fixation and movement. Additionally, the study highlights that gaze fixation is a more meaningful indicator of attention than gaze movement, as fixation reflects focused interest in specific stimuli. The findings suggest that pink packaging is most effective in attracting consumer attention, while black, white, blue, and orange are better suited for enhancing focus on the surrounding environment. These insights emphasize the growing importance of packaging design in influencing consumer behavior, particularly through color selection. This study contributes to marketing practices by providing empirical evidence for the visual impact of packaging colors, offering valuable guidance for cosmetic industry practitioners. Future research should expand sample sizes and explore additional packaging attributes, such as shape and material, to derive more comprehensive insights.

The identification of maritime objects is crucial for ensuring navigational safety, enabling effective environmental monitoring, and facilitating efficient maritime search and rescue operations. Given its ability to provide detailed spectral information, hyperspectral imaging has emerged as a powerful tool for analyzing the physical and chemical properties of target objects. This study proposes a novel maritime object identification framework that integrates hyperspectral imaging with machine learning models. Hyperspectral data from six ports in South Korea were collected using airborne sensors and subsequently processed into spectral statistics and RGB images. The processed data were then analyzed using classifier and convolutional neural network (CNN) models. The results obtained in this study show that CNN models achieved an average test accuracy of 90%, outperforming classifier models, which achieved 83%. Among the CNN models, EfficientNet B0 and Inception V3 demonstrated the best performance, with Inception V3 achieving a category-specific accuracy of 97% when weights were excluded. This study presents a robust and efficient framework for marine surveillance utilizing hyperspectral imaging and machine learning, offering significant potential for advancing marine detection and monitoring technologies.