A Neuroadaptive AR-Based Behavioral Training Framework for Therapist-Guided Psychological Support

This work presents a conceptual, inference-based framework for therapist-guided adaptive behavioral training using Augmented Reality (AR), artificial intelligence, and lightweight wearable sensing. The platform is designed as a non-diagnostic, non-therapeutic training support system that adapts to inferred engagement, regulation, and cognitive load without performing medical assessment or clinical decision-making. Ethical boundaries, therapist-in-the-loop governance, and privacy-by-design principles are embedded directly into system architecture. This release represents a Conceptual Open Science version intended to support transparency, reproducibility of design logic, and future research under appropriate ethical and clinical oversight.