Applying the Scretching Diagnostic Matrix (SDM) and Dynamic SDM (D-SDM) to AI Chatbot Psychiatry

This study applies the Scretching Diagnostic Matrix (SDM) and Dynamic Scretching Diagnostic Matrix (D-SDM) to AI-assisted psychiatry, examining chatbot effectiveness, symptom progression, emotional dependence, belief reinforcement, and clinical safety. The proposed framework integrates multidimensional psychiatric assessment, longitudinal state tracking, episode-transition modeling, evidence validity, and safety safeguards. Two fictional clinical cases demonstrate potential applications involving symptom improvement, chatbot reliance, and problematic belief reinforcement. Independently verified calculations illustrate reductions in depression and anxiety scores, while a synthetic 200-case classification model demonstrates diagnostic-performance metrics. The research establishes a testable mathematical and clinical decision-support framework for evaluating AI-mediated mental-health interactions, distinguishing theoretical demonstrations from validated clinical outcomes and identifying priorities for prospective empirical testing.