Fuzzy Delphi-based Assessment of Privacy-Aware Healthcare Monitoring Architectures in Remote and Resource-Constrained Areas

Abstract

Intelligent healthcare monitoring in geographically isolated communities requires solutions that can operate under limited digital infrastructure while maintaining privacy, regulatory compatibility, and public confidence. Selecting an appropriate system architecture is therefore particularly important in areas where connectivity and computational resources are not always reliable. This study develops a fuzzy Delphi-based framework to assess and rank five privacy-aware healthcare monitoring architectures for such environments. Expert judgments were collected for the candidate architectures and processed through the fuzzy Delphi procedure to derive consensus-based fuzzy scores and establish their final priorities. The framework was applied to the Calabria region in Southern Italy, where dispersed settlements and uneven digital connectivity provide a relevant setting for examining healthcare monitoring solutions designed for remote communities. The findings identified "edge-based local processing" (A2) as the most suitable architecture, with a final weight of 0.4, followed by "differential privacy–enhanced platforms" (A5) with a weight of 0.25. "Federated learning architectures" (A3) is ranked third (0.18), while "Blockchain-enabled monitoring" (A4) (0.09) and "centralized AI monitoring" (A1) (0.05) occupied the fourth and fifth positions, respectively. The ranking indicates a clear preference for architectures that reduce dependence on continuous connectivity and limit the external transmission of sensitive healthcare data. Comparison with results reported in previous assessments showed a high level of agreement in the relative priorities of the alternatives, providing additional support for the consistency of the proposed fuzzy Delphi-based evaluation.