Unidentified Human Remains in Mexico (2006–2019): A Computational, Epidemiological, and Forensic Dental Analysis of the National Identification Crisis
forensic odontology; human identification; unidentified human remains; machine learning; Mexico forensic crisis; dental biometrics; disaster victim identification; postmortem analysis; artificial intelligence; clandestine graves.
AuthorsAbstractBackground: Mexico faces an unprecedented humanitarian and forensic crisis. Between 2006 and 2019, at least 38,891 unidentified bodies accumulated in Servicios Médicos Forenses (SEMEFO) facilities across the country. Simultaneously, over 112,000 persons remain officially reported as missing, and an estimated 72,100 additional sets of unidentified human remains exist in state custody as of 2024. Forensic identification infrastructure has proven systemically insufficient: the United Nations Committee on Enforced Disappearances concluded that, under prevailing conditions, it would take 120 years to process existing remains without accounting for new cases. Objective: This study presents a dual-component investigation integrating (I) advanced statistical analysis, machine learning, and data mining applied to a national public dataset of unidentified bodies, and (II) a critical scientific synthesis of forensic dentistry as an evidence-based, scalable, and technologically advanced solution for mass human identification in contexts of decomposition, carbonization, skeletonization, and clandestine burials. Methods – Part I: The Quinto Elemento Lab – Crisis Forense dataset (N = 38,891 records; 11 variables; 2006–2019) was analysed using descriptive statistics, temporal trend modelling, Pearson and Spearman correlation coefficients, chi-square independence testing (χ² = 4,371.26; p < 0.0001), Kmeans clustering (k = 3), and a composite forensic documentation index (FDI; range 0–4). Missing data patterns were systematically characterised. Methods – Part II: A structured literature review was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore (2020–2026) using terms including forensic odontology, dental biometrics, AI human identification, disaster victim identification, and postmortem dental analysis. Eighty-one peer-reviewed articles were critically evaluated. Results – Part I: Unidentified bodies increased by 2,655.6% between 2006 and 2019 (CAGR = 29.1%; r = 0.943; p < 0.0001). The male-to-female ratio averaged 10.5:1. A total of 67.1% of bodies were disposed in common graves (fosa común). The Forensic Documentation Index (FDI) averaged 2.07/4.0, declining significantly over time (Spearman r = −0.626; p = 0.017). Cluster analysis identified three distinct state-level documentation profiles. Cause of death was categorised in only 45.4% of cases. A total of 17.7% of records lacked registration year. Results – Part II: Current literature confirms that dental hard tissues (enamel, dentin, cementum) constitute the most biologically resistant structures in the human body, withstanding temperatures exceeding 1,000°C, prolonged burial, immersion, and advanced decomposition. Deep learning models applied to cone-beam computed tomography (CBCT) achieve identification accuracy of 85–99.1% depending on tooth type and algorithm. Dental prosthetics, implants, and restorations create individually unique, traceable forensic signatures. No national dental forensic database exists in Mexico. Conclusions: This study demonstrates that Mexico's forensic crisis is statistically characterised by exponential growth, systemic documentation failure, and extreme institutional heterogeneity. Forensic dentistry — augmented by artificial intelligence and digital imaging — represents a scientifically robust, cost-effective, and scalable solution for the •••••••••••••••••••••••••••••••• ejprd.org- Published by Riset Publishing Services LLC.
EJPRD
Copyright © 2026 by Riset Publishing Services LLC