Artificial Intelligence to Identify Human Remains
Artificial intelligence has been trained to decipher wine reviews, classify tree pollen and even spot city code violations. Now, researchers at the University of North Texas Health Science Center in Fort Worth are applying the technology to a far more solemn purpose: identifying human remains.
This week, the university announced a $2 million grant from the National Institute of Justice to develop an AI-driven software to standardize and streamline how law enforcement and forensic anthropologists build biological profiles of the dead. The project — called MOSAIC, or Unifying Methods of Sex, Stature, Affinity & Age for Identification through Computational Standardization — aims to help solve thousands of cold cases. Other researchers involved in the effort hail from Texas State University and universities in Michigan, New Mexico and Kansas.
Kate Lesciotto, an assistant professor of anatomy at UNT Health’s College of Biomedical and Translational Sciences, is one of the lead investigators on the project.
UNT Health Fort Worth
How AI is Used in Forensic Science
AI is becoming an important tool in forensic science. Studies show it can estimate a person’s sex from a skull, and it has recently been used to develop a system that matches chest X-rays taken before and after death to identify the unknown. Scientists are also exploring how laser scans of crime scenes, paired with AI analysis, could reveal hidden evidence.
In forensic anthropology, a biological profile is essentially a snapshot of a person’s identity reconstructed from their bones. Four core traits define that profile: age, sex, ancestry and height. To estimate them, experts look for clues in skeletal details such as patterns of growth and wear, dental features and the thickness or strength of bones. Together, those markers help narrow the search.
The MOSAIC Project
Kate Lesciotto, an anatomy professor at UNT Health and one of the project’s lead investigators, said in a news release Monday that tools for creating biological profiles currently don’t account for how age, sex, ancestry and height interact. Despite long knowing these traits influence one another, researchers still estimate them one at a time using the methods at their disposal, Lesciotto added.
To build a stronger forensic tool that considers those interactions, Lesciotto and her collaborators will gather skeletal measurements and morphological data from collections in the U.S. and Mexico. Employing AI and machine learning, they plan to analyze hundreds of data points to improve the accuracy of biological profiles.
How MOSAIC Works
When finished, MOSAIC will be freely available to law enforcement and forensic anthropologists. By matching an investigator’s data against a reference database, the software would generate comprehensive biological profiles that can help guide identifications.
The ultimate aim, Lesciotto said, is to shrink the number of unidentified people listed in NamUs, a national database created in 2007 to track those missing, unidentified or unclaimed. NamUs was born out of a summit convened by the National Institute of Justice to address the challenges of such cases. In Texas alone, more than 2,000 unidentified people remain in the system.
Making identifications, Lesciotto said, could “help law enforcement and families find answers to their questions.”
The researchers plan to use AI and machine learning to create biological profiles — snapshots of a person’s identity reconstructed from their bones — using collections of humans remains in the U.S. and Mexico.
UNT Health Fort Worth
Conclusion
The use of artificial intelligence to identify human remains is a groundbreaking development in the field of forensic science. With the MOSAIC project, researchers aim to create a more accurate and efficient method of building biological profiles, which can help solve thousands of cold cases and bring closure to families of missing persons. As AI technology continues to advance, it is likely that we will see even more innovative applications in forensic science, leading to a safer and more just society.
Frequently Asked Questions
Q: What is the MOSAIC project? A: The MOSAIC project is an AI-driven software that aims to standardize and streamline the process of building biological profiles of human remains.
Q: How does MOSAIC work? A: MOSAIC uses AI and machine learning to analyze skeletal measurements and morphological data from collections in the U.S. and Mexico, and generates comprehensive biological profiles that can help guide identifications.
