Daniele Gibelli1, Danilo De Angelis2, Francesco Riboli3, Claudia Dolci4, Cristina Cattaneo2, Chiarella Sforza4. 1. LAFAS, Laboratorio di Anatomia Funzionale dell'Apparato Stomatognatico, Dipartimento di Scienze Biomediche per la Salute, Università degli Studi di Milano, V. Mangiagalli 31, Milan, Italy. daniele.gibelli@unimi.it. 2. LABANOF, Laboratorio di Antropologia e Odontologia Forense, Dipartimento di Scienze Biomediche per la Salute, Università degli Studi di Milano, Milan, Italy. 3. , Milan, Italy. 4. LAFAS, Laboratorio di Anatomia Funzionale dell'Apparato Stomatognatico, Dipartimento di Scienze Biomediche per la Salute, Università degli Studi di Milano, V. Mangiagalli 31, Milan, Italy.
Abstract
PURPOSE: The introduction of modern 3D image acquisition systems has enabled researchers to develop novel procedures for personal identification. The present study aimed to assess differences between dental scans belonging to the same or different subjects, through an innovative 3D-3D superimposition and registration method. METHODS: Twelve subjects (6 males and 6 females) with pre- and post-orthodontic treatment dental casts were recruited. A 3D scan from each cast was obtained through a laser scanner and the 3D model of the upper first and second molar on the post-treatment cast was superimposed on the pre-treatment scan, for a total of 12 matches and 100 mismatches. Point-to-point RMS (root mean square) distance was then calculated. Student's t test verified possible statistically significant differences according to group (matches/mismatches; p < 0.05). RESULTS: In case of matches, on average the point-to-point distance RMS was 0.29 mm (SD: 0.08 mm), while it was 0.94 mm (SD: 0.30 mm) for mismatches, with a statistically significant difference (p < 0.001). CONCLUSIONS: Results show that the novel procedure was able to distinguish matches from mismatches through an RMS threshold (0.50 mm): a possible method for personal identification is described, which needs to be verified through the application to a larger sample of casts.
PURPOSE: The introduction of modern 3D image acquisition systems has enabled researchers to develop novel procedures for personal identification. The present study aimed to assess differences between dental scans belonging to the same or different subjects, through an innovative 3D-3D superimposition and registration method. METHODS: Twelve subjects (6 males and 6 females) with pre- and post-orthodontic treatment dental casts were recruited. A 3D scan from each cast was obtained through a laser scanner and the 3D model of the upper first and second molar on the post-treatment cast was superimposed on the pre-treatment scan, for a total of 12 matches and 100 mismatches. Point-to-point RMS (root mean square) distance was then calculated. Student's t test verified possible statistically significant differences according to group (matches/mismatches; p < 0.05). RESULTS: In case of matches, on average the point-to-point distance RMS was 0.29 mm (SD: 0.08 mm), while it was 0.94 mm (SD: 0.30 mm) for mismatches, with a statistically significant difference (p < 0.001). CONCLUSIONS: Results show that the novel procedure was able to distinguish matches from mismatches through an RMS threshold (0.50 mm): a possible method for personal identification is described, which needs to be verified through the application to a larger sample of casts.
Keywords:
3D superimposition; Anatomy; Root mean square (RMS); Upper molars
Authors: Ademir Franco; Guy Willems; Paulo Henrique Couto Souza; Geertruida E Bekkering; Patrick Thevissen Journal: Int J Legal Med Date: 2014-11-15 Impact factor: 2.686
Authors: Ademir Franco; Patrick Thevissen; Walter Coudyzer; Wim Develter; Wim Van de Voorde; Raymond Oyen; Dirk Vandermeulen; Reinhilde Jacobs; Guy Willems Journal: J Forensic Leg Med Date: 2012-10-10 Impact factor: 1.614