Wellness tourism is among the fastest-growing segments of the global health economy, yet its development in Central Asian heritage regions remains constrained by fragmented service delivery, limited digital infrastructure, and a shortage of evidence-based planning tools. In this Perspective, we argue that advancing wellness tourism in such regions requires coupling econometric diagnosis of revenue drivers with the design of a digital platform that operationalizes those drivers, and we illustrate this dual approach using Bukhara, Uzbekistan—a UNESCO World Heritage Site rich in thermal springs, therapeutic hot sands, and mineral-rich muds. Drawing on panel data from 12 wellness facilities observed over 2021–2024, a weighted least squares model identifies three revenue determinants: client base size, service breadth, and qualified staffing. Client base expansion and qualified staffing emerge as the strongest positive determinants, while service breadth shows a paradoxical negative effect, suggesting that resource dispersion outweighs portfolio benefits in this setting. Revenue projections indicate substantial sectoral growth by 2030, with nature-oriented sanatoriums leading in relative terms. Building on these patterns, we propose the “Wellness Bukhara Voucher System”—a digitally integrated platform connecting disparate facilities through standardized vouchers, QR-code authentication, automated analytics, and a public-private partnership financial model. The platform addresses the diversification paradox through “network specialization,” allowing each facility to deepen its core competencies while the system as a whole expands service breadth via cross-referrals. We discuss infrastructure, stakeholder, regulatory, and privacy conditions for viable deployment, and argue that this perspective offers a transferable model for heritage regions seeking to convert natural healing assets into digitally coordinated wellness economies.
Deep learning for stress oriented human activity recognition
IntroductionHuman Activity Recognition (HAR) using sensor-generated time-series data has gained significant attention for assessing mental and physical states to address various behavioral disorders. This study



