Multivariate methods for the analysis of complex and big data in forensic sciences. Application to age estimation in living persons

A new model of classification of third molars development and its correlation with chronological age in a Brazilian subpopulation

The present study correlated the mineralization of third molars to chronological age using a modified classification based on Demirjian's stages in a Brazilian subpopulation and compared with the original classification. A total of 1082 patients with age ranging from 6 to 26 years were included in the sample, with at least one third molar on panoramic radiographs. The third molars were classified according to the original Demirjian classification (8 stages) and a new model based on the Demirjian method, where the original stages were grouped into four stages: AB-enamel mineralization; CD-crown dentin mineralization; EFG-root formation; and H-complete development. Statistical analyses were performed by Kruskal-Wallis/Dunn tests (α = 0.05) and the multinomial logistic regression model. Data were analyzed according to percentiles for the probability of an individual being over 18 years old. The mean ages of the stages in both classifications did not present a significant difference between superior and inferior arches (p < 0.05). The differences in mean ages between all the stages of mineralization were statistically significant (p < 0.001) only for the 4-stage classification. Males attained root formation and complete formation earlier than females (p < 0.05) in the 4-stage classification. The modified classification system showed dependence between chronological age and mineralization stages of third molars, simplifying the age estimation process. At stage H, females present a 95.7% chance of being over 18, while for males, this probability is 89.6%. This modified classification system simplifies the dental age estimation process based on third molars and can be used as a reference for future studies.

Forensic Discrimination Potential of Blue, Black, Green, and Red Colored Fountain Pen Inks Commercially Used in Pakistan, by UV/Visible Spectroscopy, Thin Layer Chromatography, and Fourier Transform Infrared Spectroscopy

Examination and comparison of fountain pen inks are very important in forensic questioned documents examination in developing countries where the chances of fraud are greater in cases of cheques, marriage papers, entry of birth and death, etc. In this study, fountain pen inks of blue, black, green, and red colours that are commercially used in Pakistan have been discriminated by UV-Vis spectroscopy, TLC, and FTIR spectroscopy. We have calculated and compared the results in terms of discriminating power. UV/Visible Spectroscopy of fountain pen inks of different brands showed different composition despite their similar colours. TLC was effectively used to differentiate between the colored components of inks. FTIR results showed that each brand could be distinguished by studying the pattern of their absorption spectra that appeared due to the presence of different functional groups. On the basis of combined results of UV-VIS, TLC, and FTIR, the DP was found from 0.73-0.8 for blue, 0.80-1.0 for black, 0.5-1.0 for green, and 1.0 for red colored fountain pen inks. Overall, this study demonstrated the elevated worth of analysis of fountain pen inks commercially used in Pakistan as the study for fountain pen inks, while not very common, remains an interesting target study.

Semantic interoperability challenges to process large amount of data perspectives in forensic and legal medicine

This article is a position paper dealing with semantic interoperability challenges. It addresses the Variety and Veracity dimensions when integrating, sharing and reusing large amount of heterogeneous data for data analysis and decision making applications in the healthcare domain. Many issues are raised by the necessity to conform Big Data to interoperability standards. We discuss how semantics can contribute to the improvement of information sharing and address the problem of data mediation with domain ontologies. We then introduce the main steps for building domain ontologies as they could be implemented in the context of Forensic and Legal medicine. We conclude with a particular emphasis on the current limitations in standardisation and the importance of knowledge formalization.

A data model for clinical legal medicine practice and the development of a dedicated software for both practitioners and researchers

OBJECTIVE

To present a data model for clinical legal medicine and the software based on that data model for both practitioners and researchers. The main functionalities of the presented software are computer-assisted production of medical certificates and data capture, storage and retrieval.

METHODS

The data model and the software were jointly developed by the department of forensic medicine of the Jean Verdier Hospital (Bondy, France) and an bioinformatics laboratory (LIMICS, Paris universities 6-13) between November 2015 and May 2016. The data model was built based on four sources: i) a template used in our department for producing standardised medical certificates; ii) a random sample of medical certificates produced by the forensic department; iii) anterior consensus between four healthcare professionals (two forensic practitioners, a psychologist and a forensic psychiatrist) and iv) anatomical dictionaries. The trial version of the open source software was first designed for examination of physical assault survivors.

RESULTS

An UML-like data model dedicated to clinical legal practice was built. The data model describes the terminology for examinations of sexual assault survivors, physical assault survivors, individuals kept in police custody and undocumented migrants for age estimation. A trial version of a software relying on the data model was developed and tested by three physicians.

DISCUSSION

The software allows files archiving, standardised data collection, extraction and assistance for certificate generation. It can be used for research purpose, by data exchange and analysis. Despite some current limitations of use, it is a tool which can be shared and used by other departments of forensic medicine and other specialties, improving data management and exploitation. Full integration with external sources, analytics software and use of a semantic interoperability framework are planned for the next months.