U.S. flag

An official website of the United States government, Department of Justice.

Reconstructing Fragmentary Skeletal Remains

Date Published
February 1, 2017

When mass graves are unearthed, the bones of the dead are often commingled, presenting forensic anthropologists with the daunting problem of sorting and matching hundreds, sometimes thousands, of bone fragments in an effort to determine the number of deceased, as well as their age and sex.

"Our goal is to make [skeletal reconstruction] easier for law enforcement and paleoanthropology"

—Mohamed Mahfouz, Principal Investigator

The current method of piecing together fragmented bones from multiple individuals relies on a geographical information system designed to store, analyze, and manipulate spatial geographical data. The system isn’t specifically designed to analyze bone fragments and isn’t very efficient at doing so, said Mohamed Mahfouz, a biomedical and systems engineer with the University of Tennessee.

Mahfouz, supported by an NIJ award, developed software to aid in the reconstruction of fragmented bones, with a particular focus on mixed fragments from multiple individuals. The first iteration of the software is ready for testing by the forensic and anthropologic communities, he said.

The software enables forensic anthropologists “to quantify and reconstruct fragmentary human skeletal remains from three-dimensional surface files generated by computed-tomography or laser scans,” according to the researchers.  The software serves as an osteological case or scene management tool, with all scanned skeletal remains being reviewable within the software application. The software allows the user to visualize fragmentary elements, accept or reject specific fragments, and finally merge the fragments into a fully reconstructed bone.

“Our goal is to make [skeletal reconstruction] easier for law enforcement and paleoanthropology,” Mahfouz said. “It is intended to help the anthropologist to sort out and combine bone fragments and determine how a bone should look.”

Mahfouz and his team designed the software to work on fragments of four skeletal elements: the femur, humerus, pelvis, and skull. Fragments from a mass grave, for example, undergo CT or laser scanning, and then features are taken from each bone fragment by measuring surface roughness. Those features are matched to corresponding features on a “template” bone. The templates are derived from a database of 800 skeletal remains from the William M. Bass Donated Collection and Goldmine Collection. The software was tested on a subset of 2,061 scanned bone fragments from the Morton Shell Mound, an ancient ossuary in Louisiana that has yielded about 24,900 human bone fragments.

A newly scanned fragment can be visualized and then compared to template bones until a match is found. For the moment, the software is a tool that compares fragments to the templates and suggests which bone the fragment is from (left or right femur, for example). The ultimate goal, Mahfouz said, is improving the software so that it can automatically scan, score, and reconstruct fragments from commingled bones. The data involved in such a comprehensive analysis is enormous and the process so time consuming as not to be practical at the moment, he said.

Mahfouz said the software “will significantly impact forensic anthropologists’ and crime scene investigators’ ability to reconstruct mass disasters, commingled mass graves, and highly fragmentary individual burials or surface scatters.”

According to Mahfouz, the current version of the software is available “so people can use it, play with it, and suggest improvements,” at:  www.mmahfouz.org

About this Article

The research described in this article was funded by NIJ cooperative agreement number 2011-DN-BX-K537 awarded to University of Tennessee.

This article is based on the grantee report Computerized Reconstruction of Fragmentary Skeletal Remains for Purposes of Extracting Osteometric Measurements and Estimating MNI (pdf, 33 pages) by Mohamed R. Mahfouz, Natalie R. Langley, Nicholas Hermann, and Emam ElHak Abdel Fatah (Researchers from the University of Tennessee, Lincoln Memorial University, and Mississippi State University).

Date Published: February 1, 2017