The aim of this paper is to identify the factors that influence employees’ attitudes and perceptions of corporate social responsibilities (CSR), as well as to discuss if there are some significant differences in the application of CSR in the analyzed Western Balkan countries, especially in terms of public and private sector affiliation. This is the first academic survey which examined the employees’ attitudes and perceptions of CSR integrally in five Western Balkan countries. Multivariate factor analysis was applied to data collected from 2410 employees in the analyzed countries. In order to test additionally the results obtained by factor analysis, the Structural Equation Model (SEM) was applied. The results of the research, obtained by multivariate factor analysis and SEM model, show that the largest percentage of employees believe that the CSR concept enables the generating of new values and success of the company, as well as welfare for the whole society. Additionally, companies are increasingly recognizing the strategic importance of CSR for the sustainability of their business, as well as the importance of development of sustainable practices in the environment. Based on the results of the research, the authors found out that there were significant differences in the understanding and acceptance of CSR in the analyzed countries, while there were significantly smaller differences between the private and public sectors. The results of the research may serve as a guide for managers when making decisions about implementing the strategy and future activities regarding CSR.

Regional competitiveness and growth performance in EU enlargement contexts depend critically on institutional foundations, yet the extent to which governance quality shapes macroeconomic policy transmission remains insufficiently examined in comparisons between the European Union and the Western Balkans. Using a panel Vector Error Correction Model (VECM) for 33 countries over 2002–2022, this paper compares cointegration patterns, adjustment dynamics, and shock transmission channels across both regions. The results show pronounced structural differences: in the EU panel, macroeconomic policies transmit more effectively into real outcomes, consistent with a stable institutional environment and stronger policy credibility. Long-run estimates further indicate that institutional quality is more strongly associated with sustained growth in the EU than in the Western Balkans, where weaker institutional anchoring limits reform effectiveness. These findings suggest that stronger institutions facilitate resilient regional competitiveness and more effective macroeconomic policy, implying that governance reforms should accompany, and in many cases precede, stabilization and liberalization efforts in EU pre-accession settings.

This paper studies the dynamics and trends in the intricate interplay of factors shaping the environmental landscape within the Balkan region and several European countries in a context of a highly competitive and global economic environment. The aim of this study is to analyse the dynamic relationships among expenditure on green technologies, expansion of green territories, waste management practices, energy efficiency, and activation of tourism. A comparative analysis of the green territories coefficient was conducted for Albania, Bosnia and Herzegovina, Bulgaria, Croatia, Central Europe and the Baltics, Georgia, the Republic of North Macedonia, Moldova, Montenegro, Romania, Serbia, and Ukraine. It was established that the leaders are Bulgaria and Albania, as the coefficient of green territories expansion for these countries, like for the majority analyzed, does not undergo significant growth. However, the difference between the coefficient of green territory expansion and the growth rates of the total emissions of pollutants increases, indicating the implementation of "green technologies" in production by these countries, resulting in a reduction in emissions of various types of pollutants. Through a comprehensive analysis of regional data and case studies, the authors explore the causal relations between terrestrial and marine protected areas, nitrous oxide emissions, methane emissions, total greenhouse gas emissions, CO2 emissions, forest area, GDP per capita. The authors' contribution to scientific knowledge lies in the formulation of the green territory expansion coefficient, defined as the difference between green territories and the incremental growth in emissions. The study produces a definition and discusses benefits such as enhanced understanding of environmental dynamics and trends and potential for fostering sustainable practices. However, drawbacks may include the complexity of integrating various factors and the challenges associated with implementation and coordination across diverse regions. Practical recommendations for policymakers and other stakeholders committed to fostering resilient and sustainable regional ecosystems in the context of global environmental changes and the growth of tourism are proposed.

We evaluated the performance and practical impact of several deep learning models, including ResNet and transformer-based architectures, for classifying colorectal adenomas with and without high-grade dysplasia in histopathology images, to support an artificial intelligence-assisted triage workflow that reduces pathologist workload while maintaining diagnostic quality and workflow efficiency. Models were trained and validated on 1,669 histopathology image tiles from colorectal polyp samples (368 with high-grade dysplasia, 767 low-grade adenomas, and 534 non-neoplastic controls) that were expert-reviewed and analysed at 100× magnification. Performance was assessed using 5-fold cross-validation, and Grad-CAM++ heat maps were generated to visualise model attention and support explainable pathologist review. ResNet101 achieved 99.17% accuracy in the best fold and 97.40% on average. The explanatory heat maps consistently highlighted diagnostically relevant epithelial regions, supporting rapid pathologist verification. In a simulated workflow of 1,000 colorectal polyps, artificial intelligence-assisted triage reduced pathologist review volume by 41-61% while maintaining 98.91% detection accuracy for high-grade dysplasia and 99.47% accuracy for adenomas without high-grade dysplasia. These reductions correspond to meaningful full-time equivalent savings in high-volume gastrointestinal pathology practice, where repetitive biopsy assessment contributes substantially to the daily workload. These findings indicate that deep learning models can accurately classify colorectal adenomas and detect highgrade dysplasia, and that artificial intelligence-assisted triage and sign-out, supported by interpretable heat maps, may provide a practical pathway towards more sustainable and efficient digital pathology workflows.