Reliability of digital impressions for human identification based on palatal structures
Mikolicz Ákos
Dental Research Division
Dr. Varga Gábor
SE FOK Oktatási Centrum Árkövy terem
2026-06-05 14:00:00
Dental Research
Dr. Varga Gábor
Dr. Vág János
Dr. Vecsei Bálint
Dr. Sándor Balázs
Dr. Liliom Károly
Dr. Tajti Péter
Dr. Fráter Márk
Objectives: Previous studies suggest that the palate and palatal rugae can serve as individual markers in human identification. This thesis aimed to confirm the feasibility of a palate-based identification method, evaluate its applicability across different digitization techniques, and assess the influence of ethnic background.
Methods: The palate was scanned three times and two years later in 20 pairs of monozygotic twins (40 individuals) using intraoral scanners (IOS) to assess repeatability. Additionally, elastic impressions and plaster models were created and digitized with a laboratory scanner. The mean absolute distance (MAD) between scans was compared after best-fit alignment to evaluate forensic reproducibility. Comparisons across sessions assessed the effects of aging, orthodontic treatment, and digitization methods, while scans from different digitization techniques in the second session examined technical reproducibility. Sibling differences over time were analyzed to evaluate aging effects. In a second study, 23 individuals from 11 countries were scanned, and superimposition (MAD) and geometric dimensions (height, width, depth) were assessed.
Results: The anterior palatal area showed significantly better repeatability and forensic reproducibility than the entire palate (p < 0.001), with orthodontic treatment having no effect. Indirect digitization produced lower reproducibility than IOSs, with forensic MAD values of 75–77 µm and technical MAD of 37 µm. IOS scans demonstrated a repeatability of 22 µm, significantly better (p < 0.001). No significant differences emerged in sibling comparisons over time; the closest sibling MAD (239 µm) exceeded the highest forensic MAD (141 µm), indicating good discriminative capacity. The geometric evaluation achieved 91.2% sensitivity and 97.1% specificity. Latitude and longitude did not significantly affect geometric matches, and MAD ranges for superimposition (1.068–0.214 mm) and repeatability (0.011–0.093 mm) did not overlap. For sex determination, the method recognized females over males with 69.0% sensitivity and 62.5% specificity.
Conclusions: Reproducibility between different IOS devices remains acceptable over two years, though poor between IOS and indirect digitization. The anterior palate appears relatively stable in young adults. Geometric and superimposition methods demonstrated robust reliability, unaffected by population differences, whereas sex prediction showed only moderate accuracy.
