Pasar al contenido principal

Computational Forensics

RYC2020-029454-I

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de Promoción del Talento y su Empleabilidad en I+D+i
Subprograma Subprograma Estatal de Incorporación
Convocatoria Ramón y Cajal
Año convocatoria 2020
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDADE DA CORUÑA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 4
Encontrada(s) 1 página(s)

RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Evidence evaluation in craniofacial superimposition using likelihood ratios

  • Martínez-Moreno, Práxedes
  • Valsecchi, Andrea
  • Mesejo, Pablo
  • Ibáñez, Oscar
  • Damas, Sergio
Financiado para publicación en acceso aberto: Universidad de Granada / CBUA, [Abstract]: Craniofacial Superimposition is a forensic identification technique that supports decision-making when skeletal remains are involved. It is based on the analysis of the overlapping of a post-mortem skull with ante-mortem facial photographs. Despite its importance and wide applicability, the process remains complex and challenging. To address this, computerized methods have been proposed, but subjectivity and qualitative reporting persist in decision-making. This study introduces an evidence evaluation system proposal based on Likelihood Ratios, previously used in other forensic fields, such as DNA, voice, fingerprint, and facial comparison. We present a novel application of this framework to Craniofacial Superimposition. Our work comprises three experiments in which our LR system is trained and tested under distinct conditions concerning facial images: the first utilizes frontal facial photographs; the second employs lateral facial photographs; and the last one integrates both frontal and lateral facial photographs. In the three experiments, the proposed LR system stands out in terms of calibration and discriminating power, providing practitioners with a quantitative tool for evidence evaluation and integration. However, the lack of massive actual data obliged us to focus our study on synthetic data only. Therefore, it should be considered a proof of concept. Nevertheless, the resulting likelihood-ratio system offers objective decision support in Craniofacial Superimposition. Further studies are required to validate in a real scenario the conclusions achieved., This research has been developed within the R&D project CONFIA (grant PID2021-122916NB-I00), funded by MICIU/AEI/10.13039/501100011033 and by ERDF/EU, and by grant FORAGE (B-TIC-456-UGR20) funded by Consejería de Universidad, Investigación e Innovación, both funded by “ERDF A way of making Europe”. Miss Martínez-Moreno is supported by grant PRE2022-102029 funded by MICIU/AEI/10.13039/501100011033 and the FSE+. Dr. Valsecchi's work is supported by Red.es under grant Skeleton-ID2.0 (2021/C005/00141299). Dr. Ibáñez's work is funded by the Spanish Ministry of Science, Innovation and Universities under grant RYC2020-029454-I and by Xunta de Galicia, Spain by grant ED431F 2022/21. The authors thank Pierre Guyomarc’h for providing us with the data used in this research. Funding for open access charge: Universidad de Granada / CBUA., Xunta de Galicia; ED431F 2022/21, Junta de Andalucía; B-TIC-456-UGR20
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Trustworthy AI-based Legal Age Estimation Using Orthopantomographs

  • Venema, Javier
  • Luca, Stefano de
  • Mesejo, Pablo
  • Ibáñez, Oscar
Financiado para publicación en acceso aberto: Universidade da Coruña/CISUG We are not allowed to share any of the images of the dataset. The network and its weights will not be shared for the time being, because there is a commercial interest in bringing this research to product by the company Panacea Cooperative Research S.Coop. However, the trained model can be used for free at https://panacea-coop.com/application-for-age-estimation-using-orthopantomographies/?lang=en so that other researchers can validate our results. Note that the published LAE tool also generates a linguistic summarization report and activation maps to facilitate explanation and interpretation of the results according to “Human agency and oversight”, “Technical robustness and safety” and “Transparency” principles of EGTAI., [Abstract]: In forensic anthropology, legal age estimation means estimating the age of a subject in order to determine if it is above or below a specific legal threshold (typically 18 years old). It is of great importance in many different scenarios such as migration of undocumented minors, child brides, exploitation or trafficking. The main goal of this work is to develop a legal age estimation method, based on artificial intelligence, that can be applied in day-to-day forensic practice. For this purpose, we use a sample of 10,739 orthopantomographs of individuals ranging from 14–26 years, across twelve countries and four continents. Our best model fine-tuned ResNeXt50 (adapted to regression) and obtained a mean absolute error of 1.12 years in testing. When thresholding the regression estimates at 18 years old, it achieved a classification accuracy of 88.38 %. In order to show the validity of the model in real forensic scenarios, we evaluated it with four samples from population origins not seen during training. Over these samples, we obtained mean absolute errors between 1.21 and 1.47 years, and accuracies between 83 and 92 % when setting the 18-year threshold. Moreover, we obtained prediction intervals of different coverage levels to address the ethical problem of overestimating the age of a minor. This allows the model to estimate the minimum of an interval with a certain coverage level, where the higher the coverage the fewer minors are estimated as adults. As an example, if estimating the minimum of a prediction interval of 95 % coverage, 98.2 % of minors in the test set are classified as such. Our robustness to geographical origin, state-of-the-art accuracy and the estimation of prediction intervals, which are the main contributions of this work, make the proposed method, to the best of our knowledge, the only one that has been proven to be applicable in real forensic scenarios. As a result, our method properly deals with the following principles of Ethics Guidelines for Trustworthy Artificial Intelligence: Human agency and oversight; Technical robustness and safety; Transparency; Diversity, non-discrimination and fairness; and Societal and environmental well-being. Furthermore, it provides both the most probable age and a prediction interval, which in turn allows the estimation of a minimum age with fixed statistical support. As a result, the developed method can be used in conjunction with other methods according to the European protocols of legal age estimation., This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No — UMAFAE — H2020-MSCA-IF-2020. Javier Venema’s work is funded by Project number 09/942572.9/23 within the 2023 call for aid from the Community of Madrid to finance projects that have obtained a Seal of Excellence within the European Innovation Council’s Accelerator Program. Dr. Mesejo’s work is supported by grant CONFIA (PID2021-122916NB-I00) funded by MCIN/AEI/ 10.13039/501100011033, and by grant FORAGE (B-TIC-456-UGR20) funded by Consejería de Universidad, Investigación e Innovación, both funded by “ERDF A way of making Europe”. This publication is part of the R&D&I project PID2024-156434NB-I00 (CONFIA2), funded by MICIU/AEI/10.13039/501100011033 and ERDF/EU. Dr. Ibáñez’s work is funded by the Spanish Ministry of Science, Innovation and Universities under grant RYC2020-029454-I and by Xunta de Galicia through grant ED431F 2022/21. We wish to acknowledge the support received from the Centro de Investigación de Galicia “CITIC”, funded by Xunta de Galicia and the European Union (European Regional Development Fund- Galicia 2014–2020 Program), through grant ED431G 2019/01. Funding for open access charge: Universidade da Coruña, CISUG., Xunta de Galicia; ED431F 2022/21, Xunta de Galicia; ED431G 2019/01, Junta de Andalucía; B-TIC-456-UGR20
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

A Survey on Forensic Age Estimation through Dna Methylation: Recent Progress and Future Directions Towards Legal Age Estimation

  • González-Romero, Rodrigo
  • Suárez-Ulloa, Victoria
  • Ibáñez, Oscar
Financiado para publicación en acceso aberto: Universidade da Coruña/CISUG, [Abstract]: DNA methylation (DNAm) has emerged as one of the most promising molecular biomarkers for age estimation (AE), offering a powerful complement to traditional anthropological and radiological methods. Its potential spans forensic, biomedical, and humanitarian domains, particularly in the identification of deceased individuals and the estimation of legal age in living persons. Despite its rapid development, the forensic implementation of DNAm-based AE models remains challenged by issues of validation, dataset diversity, and tissue-specific variability. This survey provides an up-to-date overview of DNAm-based AE models developed between 2019 and 2025. We catalogued and analyzed published models according to their data origin, tissue type, methylation quantification method, and mathematical approach. We also distinguish between models designed to estimate chronological age in living individuals and those focused on age-at-death estimation in deceased persons. Our analysis highlights trends toward tissue-specific model optimization, targeted approaches with a small number of markers, and the integration of machine learning to improve predictive performance. Special attention is given to the development of pediatric and young-adult epigenetic clocks tailored to the estimation of legal age, an area with growing societal and humanitarian relevance, particularly in migration and child protection contexts. By identifying the main methodological advances, limitations, and challenges, this review outlines a roadmap for the next generation of DNAm-based AE models, emphasizing the need for standardized validation protocols, increased population diversity, and interdisciplinary approaches that integrate molecular, radiological, and computational tools., Dr. Gonzalez-Romero’s work is supported by MICIU/AEI/10.13039/501100011033 under grant PTQ2023-013171. Dr. Suarez-Ulloa’s work is funded by MCIN/AEI/10.13039/501100011033 under grant PTQ2022-012683. Dr. Ibáñez’s work is funded by the Spanish Ministry of Science, Innovation and Universities under grant RYC2020-029454-I and by Xunta de Galicia by grant ED431F 2022/21. We wish to acknowledge the support received from the CITIC, belonging to the Galician University System Network of Research Centers, and funded by Xunta de Galicia and the European Union (European Regional Development Fund- Galicia 2021–2027 Program) by grant ED431G 2023/01. Also, this publication is part of the R&D&I project PID2024-156434NB-I00 (CONFIA2), funded by MICIU/AEI/10.13039/501100011033 and ERDF/EU. Funding for open access charge: Universidade da Coruña/CISUG, Xunta de Galicia; ED431F 2022/21, Xunta de Galicia; ED431G 2023/01
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

International Validation Study of AI-guided Craniofacial Superimposition in a Contemporary Population Sample

  • Guerra, Rosario
  • Martos, Rubén
  • Ibáñez, Oscar
  • Valsecchi, Andrea
  • Bermejo, Enrique
  • Luca, Stefano de
  • Guativonza, María Alejandra
  • R-García, Guillermo
  • Martínez-García, Verónica
  • Casallas, Daniel
©2025 Elsevier B.V. All rights reserved. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/. This version of the article has been accepted for publication in Forensic Science International. The Version of Record is available online at https://doi.org/10.1016/j.forsciint.2025.112628 All the data employed in this study could be shared under certain conditions. To access primary data (facial photographs and 3D skull models) ask Dr. Erin Kimmerle. To access secondary data (landmarks, vectors and resulting SFOs), contact the corresponding author., [Abstract]: Reliance on primary identifiers (DNA, fingerprints) alone, can delay or hamper forensic human identification. At the same time, the use of techniques with unknown accuracy can compromise identification efforts. Craniofacial Superimposition has faced scrutiny due to limitations in reliability studies, such as small samples, non-representative case selections and unrealistic experimental setups. These studies often lacked comprehensive cross-comparisons and disregarded recommended practices. Building upon insights from a previous validation study, which highlighted limitations caused by material quality and quantity, a blind validation study focusing on a contemporary sample with superior photographic quality has been conducted. Ten practitioners with different levels of expertise addressed the identification scenario under blind conditions employing a novel AI-guided Craniofacial Superimposition technology, which automates tasks within the application of the technique. Two conclusions are reached: (i) the combination of human expertise and AI assistance can lead to reliable identification outcomes. In particular, it is noteworthy that experienced participants achieved a 100 % correct decision rate, suggesting that prior experience and expertise in the technique contributes to improved decision-making ability, even when aided by AI tools; (ii) the utilization of an automatic AI-based ranking tool significantly reduced the workload for participants by providing a prioritized list of potential matches, narrowing down the number of comparisons from 1175 to 141. Additional constraints and sources of error have been identified. The study was conducted with Skeleton‑ID v4.2, a commercial AI tool (Panacea Cooperative Research, Spain); authors’ competing interests are disclosed in full., This research has been developed within the R&D project CONFIA (grant PID2021-122916NB-I00), funded by MICIU/AEI/10.13039/5011000 11033/ and by EDRF/EU. This work is part of the doctoral thesis of Ms. Guerra in the program of PhD in Biomedicine (B11.56.1) at the University of Granada. We would like to thank the Institute of Forensic Anthropology Applied Sciences (IFAAS) and the Digital Heritage and Humanities Collections (DHHC) teams at the University of South Florida for data collection. Rosario Guerra is supported by the Spanish Ministry of Science, Innovation and Universities, and the State Research Agency (AEI), under grant Doctorados Industriales (DIN2020-011619), including European NextGeneration EU funds. Dr. Valsecchi’s work is supported by “EIC Accelerator - Seal of Excellence” project, funded by Consejería de Educación y Universidades de la Comunidad de Madrid. Dr. Ibañez’s work is funded by the Spanish Ministry of Science, Innovation and Universities under grant RYC2020-029454-I and by Xunta de Galicia, Spain by grant ED431F 2022/21. Dr. Bermejo’s work has been funded by MCIN/AEI/10.13039/501100011033 under grant PTQ2022-012457. This research has been developed within the R&D project CONFIA (grant PID2021-122916NBI00), funded by MICIU/AEI/10.13039/5011000 11033/ and by EDRF/EU., Xunta de Galicia; ED431F 2022/21