Educational and Clinical Platform for Continuous Diabetes Monitoring: A Conceptual Architecture for Decision Support, Bidirectional Alerts, Traceability, and FHIR Interoperability

Olmedo-Cisneros J.; Gallegos-Acosta A.E.; Barra Á.V.; Cardenas-Nuñez Y. · Publicaciones de la Facultad de Educacion y Humanidades del Campus de Melilla (2026) · DOI: 10.30827/publicaciones.v56i1.36841

Resumen

Introduction: Ambulatory diabetes monitoring often suffers from a fragmented data-decision cycle due to manual reminders, informal messaging, and delayed documentation. This work presents a conceptual, interoperable educational and clinical platform architecture designed to close the data-decision-alert-record cycle while enabling supervised clinical training. Method: Orchestrated by n8n with human-in-the-loop artificial intelligence, the event-oriented architecture processes inputs from fingerstick, CGM, or non-invasive sensors. It structures operations into technical domains: ingestion/QA validation (schema and physiological checks), pairing/normalization within short time windows, AI-supported correction and short-term trend prediction, rule-based n8n orchestration, bidirectional alerts (via WhatsApp, SMS, or email), and FHIR/EHR-compatible resource mapping (Patient, Observation, Task, Provenance). The platform features a clinical mode for patient-professional interaction and an educational mode where students manage simulated or shadow cases under auditable supervisory feedback. Results: The expected results are framed as conceptual and operational validation through case-based walkthrough scenarios (S1–S5) and instrumented artifacts. The architecture successfully produces flow-state logs, correlation keys, versioned datasets/rules/models, and structured FHIR records to provide an auditable trace of every automated or manual decision. Conclusions: The proposal provides a functionally integrated, feasible architectural design that reinforces safety, traceability, and interoperability in diabetes follow-up. Because validation is limited to scenario-based conceptual verification and simulation, no clinical effectiveness or autonomous medical decision-making is claimed. Future work includes developing a minimum viable prototype and conducting controlled pilots.

Instituciones: Universidad Católica de Santa María