Boston University
A critical assessment of population affinity estimation using dental crown dimensions from computed tomography scans
Abstract
dc:description.abstractThis study examined the tooth crown dimensions as a method of ancestry or population affinity estimation as well as the use of computed tomography (CT) scans to obtain these dimensions. Tooth crown dimensions, specifically the mesiodistal (MD) and buccolingual (BL) dimensions, represent one method of estimating population affinity. Population affinity estimations are frequently included in the biological profile created by forensic anthropologists; however, these estimations have a racist history and can be easily misinterpreted. Pilloud et al. (2014) developed an odontometric method of ancestry estimation that utilizes tooth crown dimensions from half of the dental arch and discriminant function equations to categorize an individual as African, Asian, or European. To assess their method on the variation observed in a modern United States population, the CT scans of recently deceased demographically diverse individuals from the New Mexico Decedent Identification Database (NMDID) were obtained, rendered, measured, and analyzed. The CT scans of these individuals more accurately reflected modern variation within the United States that is not seen in skeletal collections that lack ancestral diversity. The study sample was comprised of individuals with a known sex, either assigned female at birth (AFAB) and assigned male at birth (AMAB), and an African American, Asian American, European American, Latin American, or Native American population affinity. The CT scans of 31 AFAB and 58 AMAB individuals were examined using a Digital Imaging and Communications in Medicine (DICOM) viewer, and their dentition was isolated, and then exported to Meshmixer™. In Meshmixer™, the MD and BL crown dimensions of every tooth except the third molars on one side of the dental arch were measured from a 3D digital model. A chi-square goodness of fit test was performed to analyze the ability of Pilloud et al.’s (2014) method, and the values indicate that this method does not fit a modern and diverse U.S. population. Linear discriminant function analyses (DFA) were utilized to produce a model to estimate the affinity of an individual between the five population groups in this study. The cross-validated model which uses average MD and BL crown dimensions produced classification accuracies of 21.3%. Intraobserver error was assessed with the error analysis outlined in Pilloud et al. (2014) as well as intraclass correlation coefficients (ICC). While the intraobserver error was slightly higher than the error in Pilloud et al. (2014), ICC indicated that average MD and BL crown dimensions are reliable when examined at different times by the same observer. The results indicate that odontometric methods such as the one developed by Pilloud et al. (2014) and the model from this study may not be applicable to a U.S. population due to tooth loss, disparities in dental health care, increased admixture between population groups, and production of starburst effects on CT scans.
Degree
thesis:*- Name thesis:degree_name
- Master of Science
- Level thesis:degree_level
- masters
- Discipline thesis:degree_discipline
- Anatomy & Neurobiology
- Grantor
- Boston University
- Year dc:date.issued
- 2025
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Gibson, Maysel
- Advisors dc:contributor.advisor
-
- Tallman, Sean D.
- Moore, Tara L.
Subjects
dc:subject × 6Rights
dc:rights- Statement dc:rights
-
- Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/2144/53160