Investigation of subadult dental age-at-death estimation using transition analysis and machine learning methods

This project will focus on the investigation of dental root and crown development to estimate age in subadults using transition analysis. Further, we will explore dental development within machine learning methods to evaluate performance. The goal of this research is to provide forensic practitioners with an accurate age estimation method based on a large, demographically diverse, modern subadult sample that captures the variation in dental development. Current standard methodology based on tooth mineralization often underestimates age by one to more than two years as age increases, an issue we aim to minimize.

This research will use dental development data collected on radiographs from modern living subadults from different populations in the United States (US), the United Kingdom, and South Africa from birth to late teens. These samples include underrepresented populations, such as US Hispanics, that are lacking in the original publication samples because of a shift in population demographics since the methods were originally developed as compared to the present-day demographic profile of the US. The project will be carried out through collaboration of the following institutions: Texas State University (TXST), Michigan State University (MSU), University of Texas Health Science Center San Antonio (UTHSCSA), and Triservice Orthodontic Residency Program, 59th Dental Group (TORP). Dental radiographs from UTHSCSA and TORP will be anonymized and sent electronically to TXST or case information will be directly accessed from the Maxwell Museum of Anthropology Orthodontics Case File System. Once incorporated into the Dental Development Database, graduate research assistants will score dental development and enter data into a large repository, which already contains individuals currently retained by the PI. To increase the reference material, additional datasets will be pursued.

The data will be analyzed using statistical computing software to model the best age-at-transition values for each tooth and population group. Exploratory analyses will focus on determination of a standard set of teeth to be scored for the most accurate age estimation, ancestry specific models, and potentially sex estimation from the timing of tooth development. Finally, machine learning methods will be used to further refine age estimates by examining variable interaction on dental development.

Ultimately, the goal of this project is to produce a method that is accurate, user-friendly, and accessible. A Shiny web-application will provide a platform for analysis where users can select the methodology and known demographic variables to generate statistically robust age estimates for use in casework.

This project contains a research and/or development component, as defined in applicable law, and complies with Part 200 Uniform Requirements - 2 CFR 200.210(a)(14).

CA/NCF