AI-driven surgical decision making in oral and maxillofacial surgery: from diagnostic ambiguity to personalized therapy via data structuring, predictive modeling, and XR-enhanced execution

Abstract Objectives This study proposes a dual-pathway framework for AI-supported decision-making in OMFS on the background of heterogeneous data sources, varying clinical experience, and inconsistent integration into medical classification systems. The approach targets (I) the classification of findings, symptoms, treatments in general hospitals to guide triage and referral, and (II) individualized risk assessment and surgical planning in tertiary centers, with (III) intraoperative execution supported by 3D and mixed/augmented/extended reality (MR/AR/XR) technologies. Methods Using structured data, exploratory data analysis (EDA), machine learning, and large language models (LLMs), the system generates context-aware risk predictions and personalized treatment recommendations. Clinical and imaging data (MDCT, CBCT, MRI) are harmonized and modeled. These are translated into assessment and operative workflows using additive manufacturing, MR/AR/XR. The framework improves diagnostic consistency, enables earlier triage, enhances surgical reproducibility and is designed for scalability, multilingual use, and interoperability across healthcare systems. Results The described workflow was proved to be technically feasible and clinically coherent. Structured integration of retrospective clinical data enabled stable semantic harmonization and accurate temporal modeling. ML-models trained on these datasets achieved reliable complication prediction in mandibular trauma. The operational translation 3D-planning into surgical practice using printed guides and MR-visualization demonstrated high accuracy in procedures and supported reproducible intraoperative execution. These results show that AI-derived diagnostics, predictive modelling and augmented surgical workflows can be integrated into an OMFS pathway. Conclusions This AI-based concept addresses diagnostic and procedural gaps in OMFS, supporting both peripheral care providers and surgical specialists and lay a foundation for an ethically grounded and globally applicable surgical decision support system.