African Journal of Business, Economics and Management

Background: The convergence of biomedical engineering (BME) with artificial intelligence (AI) and advanced computational methodologies is fundamentally reshaping modern healthcare. This interdisciplinary synergy enables unprecedented capabilities in medical diagnostics, therapeutic interventions, patient monitoring, and healthcare delivery systems. As AI technologies continue to evolve rapidly, understanding their integration with biomedical engineering becomes essential for researchers, clinicians, and policymakers. Objective: This comprehensive review systematically examines the transformative impact of artificial intelligence on biomedical engineering, exploring key application domains including medical imaging, diagnostics, functional genomics, healthcare informatics, and therapeutic systems. The study aims to synthesize current knowledge, identify emerging trends, and propose future research directions at the intersection of AI and BME. Methods: A systematic literature review was conducted across major scientific databases including PubMed, IEEE Xplore, Scopus, and Web of Science for publications between 2015 and 2025. The search strategy combined terms related to artificial intelligence, machine learning, deep learning, biomedical engineering, medical imaging, genomics, and healthcare informatics. Studies were included if they addressed AI applications in biomedical contexts with empirical validation or comprehensive theoretical frameworks. Thematic analysis was employed to synthesize findings across multiple domains. Results: The review reveals that AI technologies, particularly machine learning and deep learning, have achieved remarkable success across diverse biomedical applications. In medical imaging, AI algorithms demonstrate diagnostic accuracy comparable to or exceeding human experts in detecting pathologies from X-ray, CT, MRI, and ultrasound images. In functional genomics, machine learning enables analysis of high-throughput sequencing data, identification of genetic variants, and prediction of gene function and regulation. Healthcare informatics has been transformed through natural language processing for electronic health record analysis, predictive modeling for patient outcomes, and clinical decision support systems. Therapeutic applications include AI-assisted robotic surgery, personalized drug delivery systems, and brain-computer interfaces for neural rehabilitation. The integration of AI with wearable sensors and Internet of Medical Things (IoMT) enables continuous patient monitoring and proactive healthcare interventions. Conclusion: Artificial intelligence is revolutionizing biomedical engineering by enabling data-driven insights, personalized medicine, and intelligent healthcare systems. The synergy between AI algorithms and biomedical technologies enhances diagnostic accuracy, treatment precision, and patient outcomes. However, challenges remain regarding data privacy, algorithmic bias, regulatory frameworks, and ethical considerations. Future research should focus on explainable AI, federated learning for privacy-preserving analytics, multimodal data integration, and robust validation in clinical settings. The continued collaboration between engineers, data scientists, clinicians, and policymakers will be essential for realizing the full potential of AI-driven biomedical innovation.

Background: Environmental, Social, and Governance (ESG) practices have become increasingly important for businesses worldwide, yet research on ESG adoption in Small and Medium-sized Enterprises (SMEs), particularly in the restaurant sector, remains limited. Macau's restaurant industry, comprising over 2,400 establishments employing more than 36,000 workers, provides a valuable context for understanding how SMEs approach sustainability in a unique cultural and regulatory environment. Objective: This study explores the specific ESG practices adopted by SME restaurants in Macau, investigates the motivations driving these implementations, and identifies the barriers and enablers influencing ESG adoption in this sector. Methods: Employing a qualitative research design, semi-structured in-depth interviews were conducted with eight owners and managers of SME restaurants in Macau. Participants operated between one and six establishments, employing 6 to 70 individuals each. Thematic analysis was applied to interview transcripts to identify patterns, themes, and insights regarding environmental initiatives, social responsibility, governance practices, and factors affecting ESG adoption. Results: Findings reveal that government regulations serve as the primary driver of environmental initiatives, particularly in reducing plastic usage and implementing food waste management programs. However, adoption remains uneven due to financial constraints and operational limitations such as space restrictions. Social responsibility practices, including collaboration with charitable organizations and hiring individuals with disabilities, are undertaken cautiously and often suspended during economic downturns. Fair trade product adoption faces consumer resistance due to price sensitivity. Governance initiatives focus predominantly on operational efficiency through electronic workflow adoption and risk management practices, with stakeholder relationships characterized by informal, personalized approaches rather than formal governance structures. A significant knowledge gap exists regarding ESG frameworks and certifications, with participants unaware of formal sustainability standards despite engaging in some related practices. Conclusion: ESG adoption in Macau's SME restaurants is primarily driven by cost-saving opportunities and regulatory compliance rather than investor interest or consumer trust considerations. The sector requires tailored approaches including industry-specific ESG frameworks, financial incentives, hands-on training, and peer-to-peer learning networks to overcome adoption barriers. This study contributes to understanding how SME restaurants engage with sustainability in a context where formal ESG awareness remains limited but practical implementation is emerging through regulatory and operational drivers.

Background: Machine learning systems deployed in supply chain operations operate in inherently non-stationary environments where data distributions, label definitions, and operational constraints continuously evolve. Promotions, assortment changes, supplier transitions, external shocks, and sensor drift routinely destabilize model assumptions, leading to degradation in predictive accuracy, service levels, and decision latency. Traditional static models fail to maintain performance under such conditions, necessitating adaptive approaches. Objective: This systematic review examines the landscape of drift-aware, continual learning pipelines for supply chain operations and proposes a governed, self-healing MLOps architecture that enables reliable, auditable, and resilient deployments under real-world non-stationarity. Methods: A structured literature search was conducted across Scopus, Web of Science, IEEE Xplore, ACM Digital Library, and INFORMS PubSOnline for publications between 2015 and 2025. Studies were included if they addressed supply chain operations using operational tabular or time-series data with business or reliability metrics, implemented adaptive mechanisms, and reported empirical results. Thematic synthesis was performed across three analytical axes: drift landscape and adaptation mechanisms, operational effectiveness and risk, and self-healing MLOps design. Results: Fifteen applied studies spanning retail, e-commerce, logistics, manufacturing, pharmaceutical, and cold-chain contexts met inclusion criteria. Evidence demonstrates that drift-aware, self-healing approaches reduce operational costs and prediction errors, maintain service stability under changing conditions, enable fairness-aware dispatch, improve ETA and promise-date accuracy, and reduce false alerts in anomaly detection systems. However, explicit drift detectors, reliability key performance indicators (time-to-detect, time-to-recover), and governance artifacts were inconsistently reported across studies, limiting cross-study comparability and reproducibility. Conclusion: Self-healing machine learning systems that integrate continuous monitoring, drift detection, automated attribution, controlled adaptation, and gated rollout can materially improve supply chain decisions under non-stationarity. We propose a reference architecture codifying data validation, versioned storage, observability, trigger thresholds, and controlled rollout paths, coupled with a standardized benchmarking and reporting protocol. Future research should prioritize reliability metric standardization, temporal benchmark development, comparative evaluation of composite monitors, and extension to federated, privacy-preserving settings with energy accounting.

Background: Foreign Direct Investment (FDI) has been widely recognized as a catalyst for economic development in emerging economies. However, the effectiveness of FDI in stimulating growth is contingent upon the prevailing political and institutional environment. Myanmar's experience with FDI across four distinct governmental periods between 2000 and 2024 provides a unique natural experiment for examining this relationship. Objective: This study investigates the long-term impact of FDI on Myanmar's economic growth across four successive government administrations: the State Peace and Development Council (SPDC, 2000-2011), the Union Solidarity and Development Party (USDP, 2011-2016), the National League for Democracy (NLD, 2016-2021), and the State Administration Council (SAC, 2021-2024). The research examines how political stability, human capital, inflation, and government expenditure moderate the FDI-growth nexus. Methods: Employing a quantitative research design, the study utilizes annual time-series data from 2000 to 2024 sourced from the World Bank, International Monetary Fund, and Myanmar's Central Statistical Organization. Econometric techniques including Analysis of Variance (ANOVA), multiple regression analysis with dummy variables, the Chow test for structural breaks, and independent samples t-tests were applied to examine relationships and differences across government periods. Results: The findings reveal significant variations in economic performance across the four governmental periods (F-statistic = 12.45, p < 0.001). FDI exhibited a strong positive relationship with GDP growth (coefficient = 0.52, p < 0.001), while inflation demonstrated a negative impact (coefficient = -0.15, p < 0.001). The NLD period (2016-2021) experienced significantly higher growth compared to preceding administrations, while the SAC period (2021-2024) witnessed a dramatic decline (coefficient = -10.50, p < 0.001). Chow tests confirmed structural breaks at each government transition, with the most pronounced break occurring between the NLD and SAC periods (F-statistic = 15.40, p < 0.001). Conclusion: Political stability emerges as a critical determinant of FDI effectiveness in Myanmar. While FDI significantly contributes to economic growth under stable governance conditions, political instability undermines investor confidence and diminishes FDI's growth-enhancing potential. Strategic investments in human capital, inflation control, and productive government expenditure are essential for maximizing FDI's developmental impact. These findings offer valuable insights for policymakers in politically transitioning economies seeking to optimize FDI for sustainable development.

Background: The pharmaceutical industry is experiencing a paradigm shift in its commercial operations, driven by the integration of Artificial Intelligence (AI) technologies. This transformation is particularly evident in sales and marketing functions, where data-driven decision-making has become essential for maintaining competitive advantage in an increasingly complex healthcare landscape. Objective: This comprehensive review examines the role, applications, and impact of AI in pharmaceutical sales and marketing, synthesizing current literature and industry practices to provide a holistic understanding of this evolving field. Methods: A systematic literature review was conducted across multiple databases including PubMed, Scopus, Web of Science, and Google Scholar for publications between 2015 and 2025. Additionally, industry reports and white papers from leading consulting firms and pharmaceutical organizations were analyzed to capture contemporary commercial applications. Results: AI technologies including machine learning, natural language processing, and predictive analytics are being deployed across the pharmaceutical commercial value chain. Key applications include customer segmentation, sales forecasting, personalized marketing, chatbots for healthcare professional engagement, sentiment analysis, and compliance monitoring. Major pharmaceutical companies including Pfizer, Novartis, Roche, Sanofi, and AstraZeneca have reported measurable improvements in sales efficiency, marketing ROI, and customer engagement through AI implementation. Conclusion: AI is transforming pharmaceutical sales and marketing from traditional, experience-based approaches to intelligent, analytics-driven strategies. While significant benefits have been demonstrated, challenges including data quality concerns, implementation costs, regulatory compliance, and the need for human-AI collaboration require careful consideration. Future advancements in generative AI and real-world data integration are expected to further revolutionize pharmaceutical commercial operations.