PURPOSE: Metopic craniosynostosis (MCS) is a congenital condition characterized by premature fusion of the metopic suture, leading to trigonocephaly and potential neurodevelopmental concerns. Strides in diagnosis and treatment have been made in the last two decades, with new objective metrics and imaging tools to improve accuracy and consistency in landmarking and evaluation of craniofacial abnormalities such as MCS. Traditional imaging such as serial computed tomography (CT) is discouraged in pediatric patients, leading to increased reliance on 3D photogrammetry. However, both modalities have limitations in capturing the full spectrum of craniofacial dysmorphology. This review comprehensively evaluates the use of current craniometric parameters and emerging machine learning (ML) models in assessing MCS morphology. METHODS: A systematic review was conducted via keyword search of PubMed and Google Scholar in accordance with PRISMA guidelines. We aim to highlight recent advances and evaluate hurdles left in craniofacial analysis and the translation into improved surgical practices. We included all English-language studies that reported on imaging-based craniometric parameters or machine learning applications for assessing MCS. The primary outcomes extracted included methods of severity assessment, their role in surgical decision-making, and the evaluation of postoperative results. Given the heterogeneity in study design and outcome reporting, the findings were synthesized descriptively. RESULTS: A total of 58 studies, including 9,068 patients, met the inclusion criteria. Among these patients, 2,425 (26.7%) were diagnosed with metopic craniosynostosis (MCS). The studies utilized various imaging modalities, with CT imaging being the most common (78.43%), followed by 3D photogrammetry (15.69%), and 2D photogrammetry (7.84%). Over 100 unique craniometric parameters were described across the studies. The most commonly reported parameters were CT-based, including the interfrontal angle (IFA) and the endocranial bifrontal angle (EBA). Few studies provided longitudinal follow-up of morphologic outcomes. Eighteen studies (31%) investigated the use of ML models in MCS analysis. ML models introduced indices such as the Metopic Severity Score, Cranial Morphology Deviation score, and Head Shape Anomaly index, which demonstrated high diagnostic accuracy and potential for severity assessment and outcome prediction. CONCLUSION: Traditional craniometric parameters based on CT imaging remain widely used in metopic craniosynostosis, but there is a growing shift toward ML models trained with advanced imaging to provide radiation-free, more objective, and reproducible assessments of dysmorphology. Future research should prioritize multicenter data sharing, standardization of morphologic variables, and incorporation of longitudinal postoperative imaging to better inform surgical decision-making in MCS treatment. This morphometric data should be combined with genotypic information and neuropsychological outcomes.

Abstract ID: CS54