Beyond the Therapeutic Cliff: A Clinical Engineering Framework for AI-Driven Home-Based Neurorehabilitation in Rural Veterans

Abstract

The therapeutic cliff, defined as the steep drop in rehabilitation intensity once a patient leaves inpatient care, is a documented driver of poor functional recovery after stroke and traumatic brain injury. For rural veterans, distance to specialised care, broadband disparity, and fragmented access amplify the drop. We propose a clinical engineering framework for AI-driven home-based neurorehabilitation grounded in SEIPS 2.0 sociotechnical principles. The framework integrates a wearable inertial measurement layer, an artificial intelligence exercise-quality and dose-tracking layer, a fall-prediction subsystem, a caregiver interface layer, and a Veterans Health Administration Care Coordination Home Telehealth integration layer. We synthesise eighteen peer-reviewed sources, map architecture components to validated evidence, and specify deployment constraints for rural cohorts including a bandwidth-tiered profile and a polytrauma-compatible cognitive guardrail. Validation pathways and policy implications for telerehabilitation reimbursement parity are outlined.