Forensic odontology publications: insights into research trajectories and emerging trends through bibliometric study

INTRODUCTION

From a lexical perspective, the term “forensic odonto­logy” (FO) is a compound of two words: “forensic” and “odontology.” The term “forensic” has a unique history, originating from the Latin word “forensis”, which pertained to a forum. In ancient Rome, the forum was not merely a marketplace for transactions, but also a center for various activities, including those of public affairs. Over time, the meaning of “forensic” evolved to specifically denote legal matters, and the term made its way into the English language in 1659. Hence, in contemporary usage, “forensic” refers to the science dealing with the application of scientific knowledge to legal problems and legal proceedings.
The word “odontology” is derived from the Ancient Greek: ὀδούς (Romanized: odoús), meaning “tooth”, and the Greek word: λόγος (Romanized: logos), encompassing various meanings, such as speech, discourse, study, calculation, reason, and order. The term “odonto­logy” includes the suffix “-logy,” indicating a “branch of knowledge” or “science”. FO, also known as forensic dentistry and forensic odonto-stomatology, was de-fined by Keiser-Nielson [1] as the branch of forensic medicine dealing with the examination and evaluation of dental evidence.
The evolution of FO over the years has broadened its scope to encompass various areas, including dental human identification (DHI), bite mark (BM) analysis, dental age estimation (DAE), and evaluation of dental injuries in criminal and civil litigation, malpractice cases as well as instances of child abuse and neglect [2]. This evolution has sparked extensive discussions and debates within the scientific FO community, resulting in a wealth of scientific publications, such as original articles, case reports, technical notes, systematic reviews, meta-analyses, commentaries, and editorials. These publications have solidified FO’s foundation as a specialization on its own, establishing it as a distinct and valuable discipline within forensic sciences, and marked by a growing interest among researchers. Bibliometric studies serve as invaluable tools for under-standing the evolution, trends, and dynamics within scholarly fields by quantitatively analyzing bibliographic data [3].

OBJECTIVES

In this paper, we presented a comprehensive bibliometric analysis focusing on the tendencies in FO publications from 1945 to 2023 to analyze the trend of FO research. By systematically examining the scholarly output in this specialized area, we aimed to provide insights into the historical progression, thematic developments, and scholarly impact within FO. Through a comprehensive bibliometric analysis across 78 years, we seek to provide researchers and readers worldwide with an overview of the diverse facets of FO as well as insights into research trends and developments.

MATERIAL AND METHODS

DATA EXTRACTION

This observational descriptive study utilized retrospective data collection from the Scopus database. The search string employed aimed to capture all relevant research fields within FO. The search string used was TITLE-ABS-KEY ( (“Forensic Dentistry” OR “FO”) OR (“Dental Identification” OR “Dental Comparison” OR “dental profil*”) OR (“Dental Age Estimation” OR “Age Determination by Teeth”) OR (“bitemar*” OR “bite-mar*” OR “bite mar*”) OR (“dental malpractic*”) OR ( (“child abuse” OR “child-abuse”) AND (“teeth” OR “tooth”) ) OR (“cheiloscopy” OR “rugoscopy”) ) AND PUBYEAR > 1944 AND PUBYEAR < 2024 AND (LIMIT- TO (SRCTYPE “j”) ) AND (LIMIT-TO (DOCTYPE “ar”) OR LIMIT-TO (DOCTYPE “re”) ). This exploration was limited to the Scopus database to ensure consistency of the search and avoid difficulty in merging two different digital library databases, which have different meta-data indexing.
This search string consisted of a combination of common glossary terms used to index FO-related studies along with some restrictions, such as limiting the term “child abuse” to “teeth” to confine the search within dental-related studies only. Additionally, some query commands, such as wildcards (*), enabled the search string to be more flexible. For example, “dental profil*” can include any studies, which have “dental profile”, “dental profiles”, or “dental profiling”. The search was conducted on October 20^th, 2023. Data were extracted from the inception of the database to the search date, and only re-search and review articles were included. The meta-data obtained from Scopus was exported to .csv format and imported into RStudio version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria) for further analysis.

BIBLIOMETRIC ANALYSIS

Before conducting the main bibliometric analysis, author and university data were cleaned to ensure the optimal representation of current FO research. Authors’ names underwent cleaning by cross-referencing each author’s name with their corresponding Scopus ID. In cases where multiple names were detected under one Scopus ID, the names were merged. Affiliation names were cleaned using Levenshtein distance filtering, with a distance filter set at 0.1. If the distance fell below 0.1, names were combined with manual curation for each merging. The primary bibliometric analysis utilized the Bibliometrix package, encompassing citation analysis, colla­boration, and ranking. In the analysis of trending topics, key words were employed to represent trends in the field. To ensure comprehensiveness, a list of synonyms was created to combine two or more key words with similar meanings (e.g., “Panoramic”, “OPG”, and “Panoramic Radiograph” were merged into “Orthopan-tomograph”). For affiliation productivity, data were counted using R base code, crediting each affiliation once per paper (e.g., if there were 3 authors from affiliation X in one paper, the affiliation was credited with 1 article).

RESULTS

MAIN INFORMATION

The primary dataset comprised of 6,416 studies conducted between 1945 and 2023, demonstrating an annual growth rate of 7.17%. Over 13,646 unique authors contributed to FO studies, with 12.34% of these publications resulting from international collaborations. There are 1,574 single-authored documents, with an average of 3.44 authors per document.

PRIMARY SOURCES

The primary sources for FO studies were Forensic Science International (n = 349), Journal of Forensic Science (n = 340), and Journal of Forensic Odon-to-Stomatology (n = 310). Notably, due to distinct entries for the Journal of Forensic Odonto-Stomatology in Scopus, combined entries were utilized for accurate representation. All the primary sources are listed in Ta-ble 1.

AUTHORS AND INSTITUTIONS

The leading authors in FO research included Came­riere (n = 82), Franco (n = 68), Da Silva (n = 55), Willems (n = 51), and Thevissen (n = 50). These authors were affiliated with top institutions, such as Universidade de São Paulo (n = 115), Università di Macerata (n = 62), and Katholieke Universiteit Leuven (n = 59) (Tables 2 and 3). From a country perspective, India led in publications (n = 682), followed by the USA (n = 260) and Brazil (n = 258). Germany exhibited the highest average citations per article (n = 33.8), followed by Italy (n = 25.5) and the United Kingdom (n = 24.9) (Table 4).

ARTICLES AND TRENDING TOPICS

The identification of articles with the most significant impact on the field are described in Table 5. The top impactful articles in FO predominantly focused on DAE, with nine out of ten articles originating from this sub-specialty. Notably, Demirjian et al.’s “A new system of dental age assessment” garnered the most citations (n = 1,620), and exhibited significant influence over time.
“Age Estimation” stands out as the most used key word in FO studies. “Forensic Dentistry”, frequently employed in indexing terms, closely follows alongside “Identification”. Further analysis revealed connections between these top key words and other frequently used terms. For example, “rental features” and “disaster victim identification” co-occur with “Identification”, while “radiographic image analysis”, “panoramic radiographs”, and “Demirjian’s method”, exhibited an association with “Age Estimation”. The presence of key words, such as “Bite Marks”, “Cheiloscopy”, and “Sex Estimation”, further underscored the prominence of these sub-fields within FO research.
As expected from the key word occurrences, trending topic timeline analysis revealed the dominance of DAE-related studies, heavily involving panoramic cone beam computed tomography (CBCT), population data analysis, and other parameters, such as canines, molars, and third molar maturity index. Another dominant theme in FO was DHI, with recurrent terms including “denture marking”, “odontometrics”, and “disaster victim identification”. Aspartic acid racemization, a biochemical process used in FO to estimate dental age, emerged as a sustained trend from 1997 to 2018. Other sub-specialties, such as cheiloscopy and rugoscopy were also present. Additionally, machine learning emerged as a recent trend, indicating exciting developments in FO research (Figures 1-3).

DISCUSSION

The current study indicated that DAE is the most prominent sub-specialty in FO. Nine out of ten of the most cited papers in our analysis focused on methodologies related to DAE, highlighting its significance within the field. Moreover, DAE emerges as the most frequent authors’ key word, indicating its wide-spread use and importance. This prominence is further underscored by the evolving trends in FO studies. The terminology associated with DAE has evolved over time, transitioning from bio-molecular approaches to two-dimensional and three-dimensional radiography, and incorporating machine learning techniques.
The evolution of DAE research is driven by the aim to achieve precise age determination with a small error range. One of the most consistent methods is aspartic acid racemization, with the studies detected spanning from 1997 to 2016. Age estimation based on the quantification of aspartic acid racemization utilizes organic chemical compounds and protein amino acids. These compounds undergo a non-enzymatic reaction that results in a cumulative of these products, leading to an age-dependent protein product. Therefore, through the quantification of these proteins, an estimated chronological age can be determined with a relatively small error, even in adult age range [4]. However, the drawbacks of this method are its expense and limitation to postmortem scenarios, as the analysis requires destruction of the tooth [5].
Contrary to invasive and destructive methods, recent DAE research has increasingly focused on non-invasive approaches, particularly through the utilization of radiographs. Radiographs enable FOs to observe dental development in children and juveniles as well as regressive changes in adults. For instance, Demirjian et al.’s seminal work: “A new system of dental age assessment”, published in 1973, is one of the most cited studies in the field [6]. Despite its age, this method has experienced a resurgence in interest from 2010 to 2020. This renewed attention primarily stems from advancements in modelling techniques and the use of automation, facilitated by computer vision, to streamline the staging process [7]. These findings under-score the notion that older research in FO can still be valuable, provided that the formulation or model used for DAE is re-calibrated to account for modern or specific popu­lation characteristics. This emphasizes the importance of adapting older methodologies to contemporary contexts to ensure accurate and reliable results.
Another trend that emerged between 2012 and 2022 is the analysis of legal age. Legal age refers to the age at which an individual can be held responsible for criminal acts, typically set at 18 years old in many jurisdictions [8]. This concept of legal age also holds significance in cases involving asylum seekers, where individuals below a certain age are entitled to various protections from the receiving country [9]. However, to benefit from these protections, some asylum seekers intentionally misrepresent their age, claiming to be younger than they really are [10]. In response to this challenge, several methodologies and guidelines have been developed, including the third molar maturity index (I3M) [11]. The popularity of the I3M method has surged alongside the trend of legal age analysis. This can be attributed to the method’s reliance on numeric data that lends itself to a straightforward analysis of cut-off values for determining legal age through simple logistic regression. In contrast, staging methods require Bayesian’s probability analysis to achieve optimal results in determining legal age, making logistic regression cut-off value analysis a more accessible alternative [12]. The latest trend in DAE research involves the utilization of CBCT and machine learning techniques. CBCT offers FOs a three-dimensional representation of a tooth, enabling volumetric analysis. This detailed volumetric information is crucial for assessing dental morphology changes, particularly in estimating adult dental age, where thorough examination of regressive tooth changes is necessary [13]. Additionally, machine learning methodologies have been increasingly integrated into FO, ranging from modelling to computer vision ap-plications. As explained in the previous paragraph, machine learning models allow FOs to not only create more accurate models, but also enhance efficiency [14]. This effectiveness can stem from better utilization of multiple variables, lower error rates, or improved performance in real-life scenario simulations through cross- validation.
DHI has been a crucial tool in forensic investigations, owing to the uniqueness and durability of dental structures, and is continually evolving to meet the demands of forensic investigation [15]. When conducting a bibliometric study, it is essential to undertake a comprehensive examination of the historical advancement of dental identification techniques and concurrent evolution of scientific publications in this domain. As a direct result of this, the terminology related to DHI has undergone evolution over time. Within the top 10 most cited papers in FO, our analysis uncovered a conspicuous absence of studies specifically dedicated to DHI, with one notable exception [16]. This landmark study titled: “A look at forensic dentistry – Part 1: The role of teeth in the determination of human identity”, has served as a foundational reference for sub-sequent research efforts in the field of DHI [16]. The roots of DHI can be traced back to ancient civi­lizations, where rudimentary methods of dental ex-ami­nation were employed for identification purposes [17]. Despite the lack of systematic approaches, early observations laid the foundation for future developments in FO. The beginning of the 20th century witnessed significant advancements in FO, with pioneers, such as Dr. Oscar Amoedo, advocating for the use of dental records in identification [18]. Standardized dental charts and record-keeping practices emerged, facilitating the documentation and retrieval of dental information for forensic purposes. Early publications focused on identification based on dental evidence, often mentioning the dental treatment chart as an identification document [19, 20].
The introduction of radiography allowed for detailed imaging of dental structures, enhancing the accuracy of identification without damaging the tissues, especially in regions where invasive dental autopsy is not possible due to religious or ethical reasons [21]. Postmortem exa­mination of teeth and supporting structures as an aid in personal identification [22], and the use of bite-wing X-rays for dental comparisons have monopolized first publications [23]. Technological innovations in the latter half of the 20th century revolutionized dental identification practices. Digital imaging technologies and computerized databases stream-lined data management processes, facilitating the analysis of dental records on a larger scale. In recent years, it appears that 3D imaging techniques, such as CBCT and intra-oral three-dimensional optical scanning, have gathered significant interest among researchers [24, 25]. This is likely due to their ability to provide highly detailed and accurate representations of dental structures, facilitating more precise identifications. The topic of disaster victim identification (DVI) also has an increased attention spanning from 2006 to 2019. During this period, researchers and FOs contributed to the discourse surrounding DVI as well as the development of relevant software designed to aid in compute­rized dental comparisons [26, 27]. This reflects the growing importance of digital tools in dental identification processes. Additionally, the increasing frequency and severity of natural disasters and humanitarian crises, have highlighted the importance of effective victim identification methods [28].
Over the decades, FO researchers continually innovated techniques and technologies to enhance accuracy and reliability. Concurrently, the introduction of denture marking techniques has fortified identification processes, with research detected spanning from 1998 to 2013. Denture marking involves the incorporation of identifiable marks or labels on dentures. Unique marks or labels on dentures can provide vital information for establishing an individual’s identity [29]. Recent trends in denture marking research reflect advancements in materials science and technology [30]. Our analysis revealed that “odontometrics” was consistently trending as a common key word over the years 2008-2018. Odontometrics refers to the measurement and analysis of dental characteristics, including the size, shape, and arrangement of teeth. These studies are crucial in FO for DHI purposes. Over the years, odontometric studies have evolved from traditional manual measurements to more advanced digital techniques. Three-dimensional imaging techniques, such as CBCT [31], 3D scanning [32], and photo-grammetry [33], offer enhanced visualization of dental structures in three dimensions. These techniques provide detailed anatomical information, facilitating more comprehensive odontometric analysis and measurement. The integration of machine learning and artificial intelligence algorithms holds promise for improving odontometric analysis [34], assisting in pattern recognition, classification, and predictive modeling; thereby, enhancing the efficiency and accuracy of dental measurements, analysis as well as comparisons for DHI scenarios. These advancements not only enhanced the precision of dental morphological assessments, but also facilitated cross-disciplinary collaborations, particularly with molecular biologists in the realm of PCR and DNA analysis. This has been identified as an emerging trend in the research over the years 2001-2018 [35, 36]. The increasing utilization of molecular techniques underscores FO’s adaptability to emerging technologies and its commitment to refining identification methodologies.
Key words, such as “Cheiloscopy” were identified in our analysis as they have been trending in FO publications between 2014 and 2020. Cheiloscopy refers to the distinct patterns present on the vermilion border of the lips, analogous to fingerprints in uniqueness. Cheiloscopy specifically deals with the examination and analysis of these patterns [37]. BM analysis, a routine aspect of daily work in FO, involves examining patterns created by the teeth and related structures on various surfaces. This forensic technique is valuable in criminal investigations, aiding in identifying individuals or linking suspects to crime scenes, based on unique BM patterns [38].
Influential studies, such as those conducted by Dr. Pretty and Dr. Sweet (2001) [39, 40], have contri­buted to discussions on the reliability and admissibility of BM evidence in legal proceedings, promoting a higher standard of scientific consistency and evidence-based practices within forensic science. The “Standards and guidelines for evaluating bite marks”, published by the American Board of FO (ABFO) in February 2018 [41], serve as a comprehensive guide for FOs and professionals involved in analyzing BM evidence. With the evolution of technology and the advent of more sophisticated analytical methods, such as digital imaging and computer-aided techniques, the literature on BMs in FO has expanded significantly [42, 43]. The use of high-resolution photography, 3D scanning, and software-based analysis, has enhanced the accuracy and reliability of BM assessments.
Despite the potential of BM analysis in forensic investigations, there are significant limitations to consider, including possibility of distortion in BMs due to skin elasticity and movement, variability in human skin charac­teristics, and subjectivity involved in interpreting BM evidence [44]. Instances of wrongful convictions based on flawed or misleading forensic analysis [45], particularly highlighted by organizations, such as the Innocence Pro­ject, have raised concerns about the reliability and scienti­fic validity of BM analysis in criminal investigations [46]. This scrutiny may explain the decline in the trend of “BM Analysis” as key word from 2020 onwards.

LIMITATIONS AND FUTURE DIRECTIONS

While our bibliometric analysis provides valuable insights into trends and patterns within FO research, it is important to acknowledge certain limitations. The limi­tations of this study stem from the inherent characteristics of the Scopus database itself. While Scopus is widely recognized as a high-quality academic digital library, its lack of standardization in data input may lead to potentially inaccurate results when conducting bibliometric studies. As previously noted by Boedi et al. (2023) [47], multiple errors can arise from variations in data inputs, particularly in authors’ names and affiliations. These issues were addressed in our study through data cleaning procedures. This is apparent in previous bibliometric research, which has a duplicate naming, wherein one institution was counted as two different entities [48]. Furthermore, the data provided by Bibliometrix, a popular R package for bibliometric analysis, may utilize different counting methods for measuring institution or country productivity [49]. While Bibliometrix calculates university or country productivity based on the frequency of mentions in articles, this approach can sometimes be mistakenly interpreted as inflated numbers, making an institution or country with relatively low publication counts, but higher authorships per paper appear higher in the ranking. For instance, if four authors from Institution X contribute to one paper, Bibliometrix would credit Institution X with four points. In our study, we implemented a custom code to ensure the uniqueness of university productivity counts, whereby Institution X would only be credited with one point in such a scenario. This approach provides a clearer and fairer depiction of the actual research output by institutions and countries, preventing inflation of metrics due to multiple authorship from the same institution or country.
Another limitation of this study is the exclusive use of Scopus as the primary database for bibliometric analysis. While Scopus was selected for its extensive multi-disciplinary coverage, comprehensive citation da-ta, and advanced analytical tools, this choice may limit the inclusiveness of our dataset. For instance, data-bases, such as PubMed could provide a greater representation of clinically oriented FO publications, while Web of Science might capture additional highly cited and historical works that Scopus may not index com-prehensively. These differences in database coverage could impact metrics, including publication counts, citation frequencies, and identification of key journals and trends in FO. Future studies can consider integrating multiple databases to broaden the scope and potentially enhance the representativeness of bibliometric insights in this field. It should be noted that the integration between various databases may create a difficulty in unifying a dataset, since every database uses a different meta-data to log their publications. In addition, future bibliometric studies in FO could benefit from customized search strategies tailored to each sub-specialty field. This approach would enable a more nuanced analysis of micro-trends within each area of FO, offering valuable insights into the dynamics of knowledge production and dissemination within the field. Moreover, enhancing data standardization within academic databases, such as Scopus, can improve the accuracy and reliability of bibliometric studies. Collaboration among database providers, researchers, and software developers, may possibly lead to standardized protocols for data input and analysis, elevating the quality of bibliometric research across academic disciplines, including FO. Additionally, future research in FO can benefit from a more focused investigation into the application of emerging technologies, particularly artificial intelligence (AI). AI holds potential to revolutionize the analysis of dental records, BM analysis, and DAE processes. Expanding research in these areas would not only broaden the scientific base of FO, but also enhance its practical applications in law enforcement and disaster response. However, it is important to perceive AI as a tool to assist in FO, rather than as a decision maker.

CONCLUSIONS

The analysis of the landscape of FO research offers valuable insights into the evolution and trends within the field. The trending topic analysis underscores the dynamic nature of FO research, with certain themes and methodologies consistently appearing over time. Our bibliometric study offers insights into the trends and evolution of FO publications from 1945 to 2023. We observed a steady growth in research output and the prominence of DAE of research topics. The analysis provides a comprehensive overview of the current landscape of FO research, highlighting key themes, metho­dologies, and areas of focus. By identifying trends and impactful contributions, this analysis provides valuable guidance for future research directions and potential areas for further exploration within the field of FO.

DISCLOSURES

1. Institutional review board statement: Not applicable.
2. Assistance with the article: None.
3 Financial support and sponsorship: None.
4. Conflicts of interest: The authors declare no potential conflicts of interest concerning the research, authorship, and/or publication of this article.

References