Neuro-Adaptive AI Systems for Predicting Root Resorption Susceptibility During Orthodontic Tooth Movement

Abstract

External root resorption remains one of the most significant iatrogenic risks associated with orthodontic tooth movement, often progressing asymptomatically until irreversible damage occurs. This study proposes a Neuro-Adaptive Artificial Intelligence (AI) system for predicting individual susceptibility to root resorption by dynamically integrating multimodal clinical, imaging, and biomechanical data. The proposed framework leverages brain-inspired adaptive learning models capable of continuously updating risk predictions in response to patient-specific biological responses and treatment progression. High-resolution CBCT-derived root morphology, orthodontic force vectors, treatment duration, patient demographics, and genetic and inflammatory biomarkers are incorporated to enable personalized risk profiling. By transitioning from static risk assessment to real-time, neuro-adaptive prediction, the system aims to support clinicians in optimizing force application, minimizing adverse outcomes, and enhancing treatment safety. The proposed approach aligns with emerging trends in precision orthodontics and AI-driven clinical decision support systems, offering a scalable pathway for early detection and prevention of root resorption in contemporary orthodontic practice.