Improving Age-at-Death Estimates from Human Skeletal Remains Through Spatial Analysis of Intracortical Remodeling Using Geographic Information Systems Software

This project will focus on the microscopic investigation of bone remodeling at the femoral midshaft to better estimate age from adult human skeletal remains. Historically, macroscopic indicators of age have been unable to accurately age older adults, often truncating them into a category of 50+ years which is of limited utility in decedent identification, especially considering increased life expectancy in the 21st century. Current standard methodology for quantification of remodeling examines only small regions of interest, and thus do not account for known regional variation in remodeling. As such, current histological methods are only marginally better than macroscopic methods.

The goal of this research is to provide forensic practitioners with an accurate age estimation method based on a large, modern adult sample that captures the relevant spatial variation in femoral remodeling. This project will build upon recent pilot research examining the totality of remodeling events at the femoral midshaft, spatially analyzing them using geographic information systems (GIS) software. The project will take place at Texas State University and will utilize femoral midshaft samples from modern donated adult cadavers or forensic cases from different populations in the U.S., Africa, Central America, Mexico, and Thailand. Doctoral and Graduate research assistants (DRA and GRA) will process femoral midshafts to remove a transverse cross-section to create a slide for histological examination. Microscopic images will be imported into GIS where remodeling events will be quantified by DRA/GRAs, allowing for spatial analysis of remodeling to more fully appreciate spatial heterogeneity in femoral remodeling. Spatial data will be analyzed using computing software to determine which region(s) of the femoral midshaft should be included in age estimation models. Remodeling event densities will be analyzed via random forest regression, recently recognized as more flexible alternative to standard linear regression, resulting in more accurate point age estimates, and importantly, 95% prediction intervals. Dissemination will occur throughout the project to create awareness of the research among target audiences including researchers, academic and applied forensic anthropologists, and practitioners within medical examiner and coroner’s offices. Ultimately, the goal of this project is to produce a method for adult age estimation that is accurate, unbiased, accessible, and easy to use. To that end, the findings will be incorporated into a graphical user interface web-application which will provide a platform where practitioners can input data collected from their own cases to generate statistically robust age estimates for use in forensic casework.