Abstract: Background: Foreign body ingestion is a common pediatric emergency, especially among children aged 1 to 10 years. Prompt diagnosis and management are crucial to prevent complications. This study aimed to assess the prevalence and patterns of foreign body ingestion in the upper aerodigestive tract in pediatric patients presenting to the Department of Gastroenterology, Madhesh Institute of Health Sciences (MIHS), Janakpurdham, Nepal. Methods: A hospital-based cross-sectional study was conducted in the Department of Gastroenterology, MIHS, Janakpurdham, over a period of 3 years (January 2021 to December 2023). A total of 50 pediatric patients, aged 1 to 10 years, who presented with a history or suspicion of foreign body ingestion were included. Data on demographics, type and location of foreign body, symptoms, radiological findings, endoscopic procedures, and outcomes were collected. Statistical analysis was performed using SPSS version 25. P-value < 0.05 was considered statistically significant. Results: The mean age of the patients was 4.7 ± 2.3 years, with a male-to-female ratio of 1.27:1. The most common age group affected was 2-5 years (56%). Coins were the most commonly ingested foreign bodies (40%), followed by toy parts (20%), batteries (14%), fish bones (10%), and others (16%). The esophagus was the most frequent site of foreign body lodgment (52%), followed by the oropharynx (24%) and stomach (18%). Endoscopic removal was successful in 92% of cases. Complications such as mucosal erosion were observed in 8% of cases. There was a statistically significant association between the type of foreign body and site of lodgment (P = 0.023). Conclusion: Foreign body ingestion in the upper aerodigestive tract is most prevalent among children aged 2 to 5 years, with coins being the most commonly ingested objects. Early diagnosis and endoscopic management are effective in preventing serious complications. Public awareness and parental supervision are vital in reducing the incidence.

Abstract: Background: Foreign Direct Investment (FDI) is a critical driver of economic development for developing nations, yet its effectiveness is contingent on host country conditions. Burundi, a landlocked East African nation, has struggled to attract and leverage FDI effectively due to political instability, infrastructural deficits, and low human capital. This study investigates the impact of FDI on Burundi's economic growth from 1987 to 2023, emphasizing the mediating roles of human capital and trade openness. Methods: This empirical study employs time-series econometric techniques using annual data from 1987 to 2023. The Johansen Co-integration Test and Vector Error Correction Model (VECM) are used to analyze both long-run and short-run relationships between GDP growth, FDI inflows, human capital (proxied by school enrollment rates), and trade openness. Results: The analysis reveals a positive and statistically significant long-run relationship between FDI and economic growth, with a coefficient of 0.25. Human capital significantly enhances FDI's productivity (coefficient: 0.15), though low enrollment rates limit this effect. Trade openness also positively correlates with growth (coefficient: 0.20), but logistical barriers hinder its potential. Short-run dynamics show a slow adjustment speed to equilibrium (Error Correction Term: -0.35), indicating structural rigidities. Conclusions: FDI is a key driver of growth in Burundi, but its potential is constrained by structural obstacles. To maximize benefits, policymakers must prioritize education reforms, particularly vocational training, invest in infrastructure to reduce transaction costs, implement trade facilitation measures, and strengthen governance to create a stable investment climate. These steps are crucial for Burundi to harness FDI for sustainable economic development.

Abstract: This paper is research based on the impact of corporate governanceontheenvironmental and social (E&S) scores of publicly listed companiesintheEast African Community (EAC). The study runs the fixed-effectsandsystem GMM estimations using a panel data set of firms inKenya,Tanzania, Uganda, and Rwanda, 2015-2024 to investigate the effect of boardcomposition, ownership structure, and leadership structureontheenvironmental and social score provided by ESG rating providers. Thefindings show that independence of boards, gender board diversity, andthepresence of sustainability committees are positively andsignificantlycorrelated with better E&S scores whereas CEO duality andconcentratedownership are negatively correlated. Sub sample results indicatethat thereare stronger governance impacts in Kenya and Rwanda, where regulatoryreforms have implemented mandatory ESG rules and in environmentallysensitive sectors such as energy and utilities, where the visibilityofstakeholders is greater. Event-studies support the claimthat theimplementation of ESG disclosure rules in Kenya (2021) andRwanda(2024)induced identifiable shifts in sustainability performance, especiallyinthecase of well-managed companies. The consistency of the results ischeckedby robustness checks on several ESG data providers. The findings addtothebody of corporate governance and sustainability literature byillustratingthat governance structures are key levers to pursue environmental andsocial responsibility in frontier markets. The policy recommendationsarethat the regulators of EAC need to intensify the codes of governance, toalign ESG reporting frameworks and to protect minority investorsinorderto make corporate operations in the region more sustainable.

Abstract: This study examines the impact of digital transformation (DT) on financing constraints (FC) in Chinese A-share listed companies on the Shanghai and Shenzhen Stock Exchanges from 2010 to 2023. It uses the SA index as a proxy for FC and the frequency of keyword indicators in annual reports, across five core technologies: artificial intelligence, blockchain, cloud computing, big data, and digital applications. Using fixed-effects regression with Driscoll- Kraay standard errors, the results show that DT has a strong negative relationship with FC, with a coefficient significant at the 1% level. Additionally, being a state-owned enterprise (SOE) and owning shares in other financial institutions each has a significant negative impact at the 1% level. CEO duality and profit volatility (PV) have positive impacts, significant at 5% and 1%, respectively. A clean audit opinion negatively affects FC at the 10% level. Inflation rate positively influences FC, while GDP growth rate negatively influences FC, each at the 1% significance level. Furthermore, the findings indicate that DT moderates the link between PV and FC. The findings are robust to the use of the Whited-Wu index as an alternative measure of FC and apply a one-period lag to check for the causality effect. Further analysis reveals that artificial intelligence, cloud computing, big data, and digital applications significantly reduce FC, whereas blockchain’s effect is insignificant. The study concludes that strategic investment in digital technology is an effective approach for enhancing access to external capital and mitigating the positive influence of PV on FC. Future studies could examine the different impacts across industries and explore the long-term impact of DT on financial stability and sustainable growth.

Abstract: Traditional image processing techniques for corrosion detection are often inadequate, struggling with complex backgrounds, occlusions, and high sensitivity to variable lighting and environmental conditions. Deep learning offers a robust alternative, demonstrating superior adaptability for object detection in such challenging, real-world scenarios. To address the critical need for automated inspection of coastal public infrastructure, this study proposes the CBG-YOLOv5s model for the detection and grading of metal surface corrosion. The model automatically identifies, localizes, and classifies corrosion into three severity levels: mild (LC), moderate (MC), and severe (HC). Our key innovations are threefold: (1) We integrate the Convolutional Block Attention Module (CBAM) into the C3 module to create C3CBAM, enhancing the model's ability to focus on salient corrosion features. (2) We augment the feature fusion neck with a BiFPN-CBAM structure and an additional small-object detection layer to improve multi-scale feature integration, particularly for early-stage, small corrosion spots. (3) We design a lightweight C3Ghost module to replace standard convolutions, significantly reducing computational complexity without sacrificing performance. The model was trained and evaluated on a custom dataset of 6,000 annotated images of coastal metal corrosion. Experimental results demonstrate that CBG-YOLOv5s achieves an 86.0% mAP@0.5 and a 92.8% precision, outperforming the baseline YOLOv5s by 2.7 percentage points in mAP and 3.2 percentage points in precision, while simultaneously reducing its parameter count by 20.5% (from 6.728M to 5.354M). This makes it highly suitable for real-time deployment on edge devices for preventative maintenance.