How ‘smart’ is smart dentistry?

Introduction

Advances in information, communication and health technologies triggered a paradigm shift in modern medicine – the transition to so called smart medicine. Some of the first appearances of the term smart medicine in the above context appeared in the late eighties and ninetieths in relation to (1) smart medical systems in the Space Station¹, (2) nuclear medicine and surgery² and (3) advanced biomimetic materials³. In the beginning of the third millennia the research literature production on this subject started to grow. New smart application were introduced, like robotics surgery⁴, smart medical systems in nutrition⁵, smart medical records⁶ and smart sensors⁷. Recently, additional new smart health technologies including personalized and precision medicine, gamification based treatment, artificial intelligence, 3D printing, nanotechnology, Internet of Things and semantic health records have emerged^8–10. Recently, dentistry started to follow smart medicine trends¹¹ and the aim of this paper is to assess the ‘smartness’ of smart density using a bibliometric approach. Due to the lack of “gold standards” it is not yet clear what may be considered smart or not smart technology in medicine or dentistry. Thus, in the absence of better metrics, we assessed the dentistry “smartness” with the frequency of use of the above listed smart technologies.

Methods

To analyse the state of the art in smart dentistry, we analysed the corpus of papers retrieved from the Scopus bibliographical database (Elsevier, Netherlands). The search string was composed from representative keywords found in smart medicine research in the following manner: smart or personalized or precision or G4H or “artificial intelligence” or “3D print*” or nanotechnology or robotic* or IoT or “semantic health record”. The search was restricted to the subject area of dentistry (which in Scopus includes dentistry, endodontics, oral health, oral biology, orthodontics, prosthodontics and periodontology). We limited the search to the period beginning in 2001, when the growth of literature production on smart medicine began, up to 2018 (inclusive) and articles published in journals only. Using descriptive bibliometrics we identified the research literature production trends, most productive countries and most prolific journals.

To analyse and visualize the context of the smart density research literature we used a bibliometric mapping approach and a popular mapping tool called VOSViewer Version 1.6.9 (Leiden University, Netherlands)^12,13. The outputs from VOSViewer are various types of bibliometric maps, frequently called science landscapes. Landscapes can reveal different patterns and aspects of research literature like associated or related terms/keywords, timelines, citation, country or networks and similar. In our study, the author cluster keyword landscape was induced using “Create a map based on bibliographic data” option in the opening VOSViewer menu. After selecting Scopus as the bibliographic database used and defining the names of files to be analysed we selected “Co-occurrence Author Keywords” as the type of analysis and “Full counting” as the counting method. Then we set the “Minimum number of occurrences of a keyword” to 8 occurrences. For all other parameters the default values were used. The proximity of terms indicates keyword similarity and the coloured clusters represent strongly associated keywords. Using a customized VOSViewer thesaurus file, we excluded common and statistical keywords like systematic review or meta-analysis from the analysis. We also mapped synonyms into one entity (for example cone beam computer tomography, cone-beam computer tomography, cone beam computed tomography, cone-beam computed tomography and cbct into cone-beam computer tomography). The thesaurus file is consisting of two columns, first includes the synonym and second the keyword in which the synonym should be mapped. To omit a keyword from the analysis, the second column entry is left empty.

Results

The search was performed on 12^th of December 2018 and resulted in a corpus of 2470 papers. The research literature production exhibits the linear growing trend from 2001 till 2016, namely from 46 to 198 articles per year, with the average increase of nine papers per year. In last two years the growth was still linear, however with an average increase of 78 articles per year. The productivity reached its peak in 2018 with 353 articles.

The most productive countries were United Stated of America (USA) (n=627), Germany (n=298), Brazil (n=223), Italy (n=174), United Kingdom (UK) (n=168), India (n=1266), Japan (n=120), South Korea (n=111). Switzerland (n=110) and China (n=119). The top 10 productive countries are belonging either to the G8 group or are countries with highly developed economies and health systems. The most prolific journals are Journal of Prosthetic Dentistry (n=131), Dental Materials (n=83), Oral Oncology (n=79), Journal of Oral and Maxillofacial Surgery (n=75), Journal of Dental Research (n=71), American Journal of Orthodontics and Dentofacial Orthopaedics (n=59), Clinical Oral Implants Research (n=51) and Clinical Oral Investigation (n=49). Top journals belong to the most prestigious and highest-ranking journals in the dentistry field.

Nine clusters (Figure 1) emerged on the cluster landscape. We used the cluster keywords as codes in the thematic analysis¹⁴, focusing on “medical smartness”. In that manner the following smart dentistry themes were derived:

From a quantitative point of view, the most prolific smart medicine technologies used in dentistry are 3D printing occurring in 99 articles, nanotechnology occurring in 80 articles, robotic surgery occurring in 43 articles, digital impression occurring in 33 articles and artificial intelligence occurring in 13 articles.

Figure 1. The author keywords cluster science landscape of smart dentistry research.

Discussion

The above analysis showed that smart dentistry in general is following smart medicine “movement” especially in using 3D printing, nanotechnology and smart materials, robotics, IoT (i.e. sensors) technologies, personalised and precision medicine and artificial intelligence. The three examples of technologies/approaches which brought most smartness to the dentistry are smart materials, which can be altered in a controlled manner by pH, various fields, temperature, stress, etc to mimic for instance enamel and dentin or execute desired intelligent functions like diagnostic and regeneration; real time imaging and CAD/CAM systems setting foundations for increase in precision in robotic surgery; and artificial intelligence which aids quick diagnosis and customises treatment planning based on myriad of data gathered from 3D scans, cone beam computed tomography and similar complex devices.

Conclusion

Despite many applications of smart technologies in dentistry, there are still substantial gaps. Smart medicine technologies regarding gamification, deep-learning or semantic eHealth dentistry records are yet to be started to use. Thus, in absence of “gold standards” we may state that the smart dentistry is smart, but to be really successful it must become smarter.

Data availability