A CLOSED-LOOP AGE-FRIENDLY MEDICAL PLATFORM: FROM AI CONSULTATION TO POST-DIAGNOSIS CARE
TL;DR
The results indicate that an integrated consultation-escort-follow-up workflow can reduce digital barriers and support safer post-diagnosis management in community healthcare settings.
Abstract
This study designs and evaluates a WeChat mini-program that provides closed-loop, age-friendly medical support for older adults. The platform integrates AI consultation, hospital navigation, escort-service tracking, prescription recognition, medication reminders, and family health archives through a four-layer architecture consisting of a front end, FastAPI service layer, AI engine, and local data layer. Functional tests covered consultation, ordering, tracking, OCR prescription processing, security, and weak-network scenarios. AI consultation returned compliant reference suggestions rather than definitive diagnoses, and core pages loaded within one second. In usability testing, 20 older participants completed seven representative tasks with an overall completion rate of 88.6%, a mean task time of 50.3 seconds, and a SUS score of 77.8. Clear prescription images reached 94% field-level OCR accuracy, while the LLM verification layer flagged dosage and instruction risks. The results indicate that an integrated consultation-escort-follow-up workflow can reduce digital barriers and support safer post-diagnosis management in community healthcare settings.