Retrospective validation of a novel digital methodology for evaluating intraoral maxillary scans in newborns with craniofacial disorders

The study aimed to validate a computer-based approach for the geometric analysis of intraoral maxillary scans in newborns and infants with craniofacial disorders (CDs), focusing on validity and interrater reliability across diagnostic groups. Depending on CD type, 11–15 anatomical landmarks were defined and placed on intraoral scans in OnyxCeph3™ by two independent raters. Transversal and sagittal maxillary dimensions were derived from these landmarks. A total of 550 scans from 220 patients were analysed, including 91 infants with Robin sequence, 48 with Down syndrome, 53 with unilateral cleft lip and/or palate, and 28 with bilateral cleft lip and/or palate. The scans, collected during routine palatal plate treatment, were processed for landmark identification and exported to MATLAB for geometric analysis and to SPSS for statistical evaluation. Interrater reliability was assessed using intraclass correlation coefficients (ICC), and rater disagreement quantified as absolute differences. Additionally, the effect of scan completeness, age and CD type on rater disagreement was evaluated. ICC values ranged from good to excellent (0.74–0.99). Gingival and cleft-specific landmarks showed the highest relative rater disagreement. Median absolute disagreement values ranged from 0.23 to 1.22 mm. Age showed weak significant correlations with disagreement for some landmarks (r < 0.10). Scan completeness did not significantly affect disagreement (p > 0.05). CD type had significant effects on disagreement for certain landmarks (p < 0.05). The proposed digital workflow offers a standardized, replicable framework for 3D maxillary measurement, facilitating multicentric research and improving early diagnostic and therapeutic decision-making in craniofacial care. • Establishes a standardized digital method for maxillary measurements in newborns with craniofacial disorders. • Addresses interrater variability and improves measurement reliability across different conditions. • Demonstrates the impact of scan completeness and landmark standardization on accuracy. • Provides a framework for reproducible 3D analysis, ensuring comparability across studies. • Lays the foundation for future AI-driven advancements in craniofacial imaging and diagnosis.