AIdentifyAGE Ontology for Decision Support in Forensic Dental Age Assessment

Age assessment is crucial in forensic and judicial decision-making, particularly in cases involving undocumented individuals and unaccompanied minors, where legal thresholds determine access to protection, healthcare, and judicial procedures. Dental age assessment is widely recognized as one of the most reliable biological approaches for adolescents and young adults, but current practices are challenged by methodological heterogeneity, fragmented data representation, and limited interoperability between clinical, forensic, and legal information systems. These limitations hinder transparency and reproducibility, amplified by the increasing adoption of AI- based methods. The AIdentifyAGE ontology is domain-specific and provides a standardized, semantically coherent framework, encompassing both manual and AI-assisted forensic dental age assessment workflows, and enabling traceable linkage between observations, methods, reference data, and reported outcomes. It models the complete medico-legal workflow, integrating judicial context, individual-level information, forensic examination data, dental developmental assessment methods, radiographic imaging, statistical reference studies, and AI-based estimation methods. It is being developed together with domain experts, and it builds on upper and established biomedical, dental, and machine learning ontologies, ensuring interoperability, extensibility, and compliance with FAIR principles. The AIdentifyAGE ontology is a fundamental step to enhance consistency, transparency, and explainability, establishing a robust foundation for ontology-driven decision support systems in medico-legal and judicial contexts.

💡 Research Summary

The paper presents the development and validation of the AIdentifyAGE ontology, a domain‑specific semantic framework designed to standardize and integrate all aspects of forensic dental age assessment (DAA). Recognizing that age determination is pivotal for undocumented migrants and unaccompanied minors—populations whose legal rights to protection, health care, and judicial processes hinge on precise age thresholds—the authors argue that dental development remains the most reliable biological marker for adolescents and young adults. However, current DAA practice suffers from methodological heterogeneity, fragmented data representations, and limited interoperability among clinical, forensic, and legal information systems. The situation is further complicated by the rapid adoption of artificial‑intelligence (AI) models that analyze orthopantomograms (OPG) but generate additional metadata (training data provenance, hyper‑parameters, inference logs) that are rarely captured in a unified manner.

To address these challenges, the authors built the AIdentifyAGE ontology on top of the Ontology for Biomedical Investigations (OBI), reusing and aligning with established biomedical, dental, radiology, and machine‑learning ontologies. Every term receives an Internationalized Resource Identifier (IRI) and rich annotations (formal definitions, usage notes, domain‑specific descriptions) in compliance with FAIR principles. The ontology is organized into three interrelated sub‑domains:

1. Judicial/Forensic Domain – models case identifiers, requesting authorities, examination dates, expert roles, and the final judicial report that contains the age‑assessment conclusion.
2. Manual DAA Domain – captures tooth‑development staging systems (e.g., Demirjian, Willems, Li‑versidge), links each stage to population‑specific reference studies, and records statistical outputs such as mean estimated age, standard deviation, confidence intervals, and age‑threshold classifications.
3. AI‑Based DAA Domain – represents the full machine‑learning workflow: data collections (OPG image sets), model characteristics (algorithm type, hyper‑parameters, training dataset provenance), inference runs, and model outputs (regression values or classification labels).

The ontology creation followed a three‑phase methodology: (i) knowledge‑base creation, where the scope was defined and OBI’s class hierarchy was extended with dental‑specific concepts; (ii) definition chain, involving semantic annotation, detection of term redundancy, and linking to external ontologies; and (iii) validation, which combined logical consistency checks using the HermiT reasoner with functional adequacy testing via eleven competency questions (CQs). Each CQ reflects a realistic forensic‑legal information need (e.g., “What is the estimated age and confidence interval for case X?”) and was successfully answered through SPARQL queries, demonstrating that the ontology can retrieve examination context, methodological parameters, statistical results, and AI model provenance in a traceable manner.

Interoperability was further verified by ensuring that reused classes retain their original semantics and can be queried across integrated knowledge bases. The ontology is openly published on BioPortal in OWL format and accompanied by a GitHub repository containing the full class/property definitions, SPARQL examples, and a traceability matrix linking CQs to ontology elements.

Importantly, the design anticipates future extensions: new AI models, additional tooth‑development indices, or novel statistical reference datasets can be incorporated by adding subclasses or properties without disrupting existing structures. This extensibility, combined with FAIR compliance, positions AIdentifyAGE as a robust backbone for ontology‑driven decision‑support systems (DSS) that can assist forensic odontologists, legal authorities, and AI developers alike.

In summary, the AIdentifyAGE ontology unifies heterogeneous forensic, clinical, and AI data streams into a coherent, machine‑readable model. By doing so, it enhances transparency, reproducibility, and explainability of dental age assessments, thereby strengthening the evidentiary value of age estimates in medico‑legal contexts and facilitating international collaboration and standardization in forensic age determination.