Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
Thomas T. Nguyen, DMD, MSc, FRCD(C); Naomie Larrivée; Alicia Lee; Olexa Bilaniuk, BASc MSc;
Robert Durand, DMD, MSc, FRCD(C)
Cite this as: J Can Dent Assoc 2021;87:l7
ABSTRACT
The field of artificial intelligence (AI) has experienced spectacular
development and growth over the past two decades. With
recent progress in digitized data acquisition, machine learning
and computing infrastructure, AI applications are expanding
into areas that were previously thought to be reserved for
human experts. When applied to medicine and dentistry, AI has
tremendous potential to improve patient care and revolutionize
the health care field. In dentistry, AI is being investigated for a
variety of purposes, specifically identification of normal and
abnormal structures, diagnosis of diseases and prediction of
treatment outcomes. This review describes some current and
future applications of AI in dentistry.
Published: May 3, 2021
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 1 of 7
Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
J Can Dent Assoc 2021;87:l7
May 3, 2021
W
hat once seemed like science fiction is now single, indivisible step of pattern-matching; rather, the child first sees
becoming reality in health care. Artificial intelligence the edges of the object, a particular grouping of which defines a
(AI) is a fast-moving technology that enables machines textured outline with simple shapes, such as eyes and ears. Among
to perform tasks previously exclusive to humans.1 Advances in these components, larger groups such as heads and legs arise, and a
AI offer a glimpse of such health care benefits as decreasing particular grouping of these defines the whole cat.
postoperative complications, increasing quality of life, improving
decision‑making and decreasing the number of unnecessary An extremely popular class of DL algorithms is the artificial neural
procedures.2 When applied to the fields of medicine and dentistry, network (ANN), a structure composed of many small communicating
AI can play a crucial role in improving diagnosis accuracy and units called neurons organized in layers. A neural network is
revolutionizing care. AI is currently used for a variety of purposes composed of an input layer, an output layer and hidden layers in
in dentistry: identification of normal and abnormal structures, between.7 It is possible to have 1 or a few hidden layers (shallow
diagnosis of diseases and prediction of treatment outcomes. neural network) or multiple/many hidden layers (deep neural
Furthermore, AI is used extensively in dental laboratories and is network, DNN) (Figure 1, a and b). These layers are called hidden
playing a growing role in dental education. The following review because their values are not pre-specified or visible to the outside.
describes current and future applications of AI in the clinical Their aim is to make it possible to build hierarchically on information
practice of dentistry. retrieved from the visible input layer to compute the correct value of
the visible output layer. The pattern of connections between neurons
defines the particular neural network’s architecture, and the fine-
What Is Artificial Intelligence? tunable strengths of those connections are called the weights of the
neural network.
AI is a branch of computer science that aims to understand and build
intelligent entities, often instantiated as software programs.3 It can be In medicine and dentistry, one of the most commonly used subclasses
defined as a sequence of operations designed to perform a specific of ANN is the convolutional neural network (CNN) (Figure 1c).
task.4 Historically, artificially intelligent systems applied hand- A CNN uses a special neuron connection architecture and the
crafted rules to the specific tasks they were meant to solve. Each mathematical operation, convolution, to process digital signals such
task required domain-specific knowledge, engineering and manual as sound, image and video. CNNs use a sliding window to scan a
fine-tuning of the system by subject-matter experts. For instance, a small neighbourhood of inputs at a time, from left to right and top to
system designed to detect lesions in medical imaging might look for bottom, to analyze a wider image or signal. They are extremely well
abnormally coloured lumps of a given shape. The fine-tunable parts adapted to the task of image classification and are the most-used
of the system might be a range of healthy tissue colours or minimum algorithm for image recognition.7
lengths and widths for a potential lump. Nowadays, medicine most
commonly uses a branch of AI called machine learning5 and, more
recently, deep learning.6 Clinical Application of AI in Dentistry
Machine learning (ML) is a branch of AI in which systems learn to Radiology
perform intelligent tasks without a priori knowledge or hand-crafted
rules. Instead, the systems identify patterns in examples from a CNNs have shown promising ability to detect and identify anatomical
large dataset, without human assistance. This is accomplished by structures. For example, some have been trained to identify and
defining an objective and optimizing the system’s tunable functions label teeth from periapical radiographs. CNNs have demonstrated
to reach it. In this process, known as training, an ML algorithm gains a precision rate of 95.8–99.45% in detecting and identifying teeth,
experience through exposure to random examples and gradual almost rivaling the work of clinical experts, whose precision rate was
adjustments of the “tunables” toward the correct answer. As a result, 99.98%.8,9
the algorithm identifies patterns that it can then apply to new images.
This technique is analogous to an adult showing several photos of CNNs have also been used for the detection and diagnosis of dental
cats to a child. The child eventually learns the patterns involved in caries.10 In 3000 periapical radiographs of posterior teeth, a deep
recognizing a cat and identifying one in new images. CNN algorithm was able to detect carious lesions with an accuracy
of 75.5–93.3% and a sensitivity of 74.5–97.1%. This is a considerable
Deep learning (DL) is a sub-branch of ML wherein systems attempt improvement over diagnosis by clinicians using radiographs alone,
to learn, not only a pattern, but also a hierarchy of composable with sensitivity varying from 19% to 94%.11 Deep CNNs have great
patterns that build on each other. The combination and stacking potential for improving the sensitivity of dental caries diagnosis
of patterns create a “deep” system far more powerful than a plain, and this, combined with their speed, makes them one of the most
“shallow” one. For instance, a child does not recognize a cat in a efficient tools used in this domain.
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 2 of 7
Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
J Can Dent Assoc 2021;87:l7
May 3, 2021
Figure 1: Schematic representation of the architecture of neural networks. Artificial neural networks are structures used in machine
learning. They contain many small communicating units called neurons, which are organized in layers. a. Shallow neural networks
are composed of an input layer, a few hidden layers and an output layer. b. Deep neural networks have an input layer, multiple
hidden layers and an output layer. c. Convolutional neural networks use filters to scan a small neighbourhood of inputs.
Orthodontics Periodontics
ANNs have immense potential to aid in the clinical decision-making According to the 1999 American Academy of Periodontology
process. In orthodontic treatments, it is essential to plan treatments classification of periodontal disease, 2 clinical types of periodontitis
carefully to achieve predictable outcomes for patients. However, it is are recognized: aggressive (AgP) and chronic (CP) forms.14 Because
not uncommon to see teeth extractions included in the orthodontic of the complex pathogenesis of the disease, no single clinical,
treatment plan. Therefore, it is essential to ensure that the best clinical microbiological, histopathological or genetic test or combination
decision is made before initiating irreversible procedures. An ANN of them can discriminate AgP from CP patients.15 Papantanopoulos
was used to help determine the need for tooth extraction before and colleagues16 used an ANN to distinguish between AgP and CP
orthodontic therapy in patients with malocclusion.12,13 The four in patients by using immunologic parameters, such as leukocytes,
constructed ANNs, taking into consideration several clinical indices, interleukins and IgG antibody titers. The one ANN was 90–98%
showed an accuracy of 80–93% in determining whether extractions accurate in classifying patients as AgP or CP. The best overall
were needed to treat patients’ malocclusions.12,13 prediction was made by an ANN that included monocyte, eosinophil,
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 3 of 7
Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
J Can Dent Assoc 2021;87:l7
May 3, 2021
neutrophil counts and CD4+/CD8+ T-cell ratio as inputs. The study prognosis. As some oral lesions can be precancerous or cancerous
concluded that ANNs can be employed for accurate diagnosis of AgP in nature, it is important to make an accurate diagnosis and prescribe
or CP using relatively simple and conveniently obtained parameters, appropriate treatment of the patient. CNN has been shown to be
such as leukocyte counts in peripheral blood. a promising aid throughout the process of diagnosis of head and
neck cancer lesions. With specificity and accuracy at 78–81.8% and
Various non-surgical and surgical methods have been devised for the 80–83.3%, respectively (compared with those of specialists, which
treatment of periodontally compromised teeth (PCT) and supporting were 83.2% and 82.9% respectively), CNN shows great potential for
structures.17 Despite advances in treatment modalities, no significant detecting tumoural tissues in tissue samples or on radiographs.25,26
improvement has been made in the method for diagnosing and
predicting the prognosis of PCT. Clinical diagnostic and prognostic One study used a CNN algorithm to distinguish between 2 important
judgement depends heavily on empirical evidence.18 Lee and maxillary tumours with similar radiologic appearance but different
coworkers19 evaluated the potential utility and accuracy of deep clinical properties: ameloblastomas and keratocystic odontogenic
CNN algorithms for diagnosing and predicting PCT. Using the CNN tumours.26 The specificity and the accuracy of diagnosis by the
algorithm, the accuracy of PCT diagnosis proved to be 76.7–81.0%, algorithm were 81.8% and 83.3%, respectively, comparable with
while the accuracy of predicting the need for extraction was those of clinical specialists at 81.1% and 83.2%. However, a more
73.4–82.8%. The noted difference in accuracy seemed to occur significant difference was observed in terms of diagnostic time:
between different types of teeth, with premolars more accurately specialists took an average of 23.1 minutes to reach a diagnosis,
diagnosed as PCTs than molars (accuracies were 82.8% and 73.4%, while the CNN achieved similar results in 38 s.26
respectively). This could be explained by the fact that premolars
normally have a single root, whereas molars have 2 or 3 roots, thus
exhibiting a more complex anatomy for a CNN to interpret. Challenges of AI
Endodontics The management and sharing of clinical data are major challenges
in the implementation of AI systems in health care. Personal data
Although mandibular molars tend to have similar root canal from patients are necessary for initial training of AI algorithms, as
configurations, several atypical variations may occur.20 To minimize well as ongoing training, validation and improvement. Furthermore,
treatment failures related to morphological differences and to the development of AI will prompt data sharing among different
optimize the clinical outcomes of endodontic therapy, cone-beam institutions and, in some cases, across national boundaries. To
computed tomography (CBCT) has become the gold standard. integrate AI into clinical operations, systems must be adapted to
However, because of its higher dose of radiation compared with protect patient confidentiality and privacy.27 Thus, before considering
conventional radiographs,21 CBCT is not used systematically. To broader distribution, personal data will have to be anonymized.28
overcome such challenges, AI has been introduced to classify Even with the ability to take these precautions, there is skepticism
the given data using a CNN22 to determine whether the distal in the health care community about secure data sharing.
root of the first mandibular molar has 1 or more extra canals.
Radiographs of 760 mandibular first molars taken with dental AI systems are also associated with safety issues. Mechanisms
CBCT were analyzed. Once the presence or absence of the must be created to control the quality of the algorithms used in
atypia was determined, image patches of the roots obtained from AI. To remedy this situation, the United States Food and Drug
corresponding panoramic radiographs were processed by a deep- Administration has created a new drug category, “Software as
learning algorithm to classify morphology. Medical Device,” through which it regulates safe innovation and
patient safety.29 Ambiguous accountability in the use of AI systems
Although the CNN had a relatively high accuracy of 86.9%,20 several is another concern. Who will be held responsible for a patient who
limitations exist regarding its clinical integration. The images must be faces unintentional consequences resulting from an error or adverse
segmented manually,23 which consumes a considerable amount of event caused by the AI technology? Is it the professional’s fault, or
time. Furthermore, the obtained images must be of adequate size and is it the fault of the developer who built the algorithm? Given that
should focus on a small region to allow the system to concentrate our legal system is based on the fundamental assumption that fault
on the object being studied, while covering enough area to include and crime are ultimately attributable to humans, substituting humans
pertinent information.24 with autonomous agents raises numerous questions of legal and
ethical order. These issues will continue to represent a considerable
Oral Pathology challenge to our legal system for the foreseeable future.
Detection and diagnosis of oral lesions is of crucial importance Finally, the transparency of AI algorithms and data is a substantial
in dental practices because early detection significantly improves issue. The quality of predictions performed by AI systems relies
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 4 of 7
Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
J Can Dent Assoc 2021;87:l7
May 3, 2021
heavily on the accuracy of annotations and labeling of the dental professionals. Rather, the use of AI should be viewed as a
dataset used in training. Poorly labeled data can lead to poor complementary asset, to assist dentists and specialists. It is crucial to
results.30 Clinic-labeled datasets may be of inconsistent quality, ensure that AI is integrated in a safe and controlled manner to assure
thus limiting the efficacy of the resultant AI systems. Furthermore, that humans retain the ability to direct treatment and make informed
health care professionals should possess a full understanding of decisions in dentistry.
the decisions and predictions made by an AI system, as well as
the capability to defend them.31 Interpretability of AI technology The road to successful integration of AI into dentistry will necessitate
is a known problem, and major advances are required before training in dental and continuing education, a challenge that most
certain classes of algorithms, such as neural networks, can institutions are not currently prepared for. In addition, AI plays a
make clinical diagnoses or treatment recommendations with full critical role in virtual reality (VR) and augmented reality (AR). A
transparency.29 new term, mixed reality, incorporates aspects of generative AI,
VR and AR into computer-superimposed information overlays to
enhance learning and surgical planning.32 As various AI systems for
Conclusions diverse dental disciplines are being developed and have produced
encouraging preliminary results, a future for AI in the health care
Although multiple studies have shown potential applications of system cannot be discounted. AI systems show promise as a great aid
AI in dentistry, these systems are far from being able to replace to oral health professionals.
THE AUTHORS
Dr. Nguyen Mr. Bilaniuk
is an Assistant Professor, Faculty of Dentistry, is a Research Software Developer, Mila –
McGill University, Montreal, Quebec. Quebec AI Institute, Montreal, Quebec.
Ms. Larrivée Dr. Durand
is a 4th year dental student, Faculty of is an Associate Professor, Faculty of
Dental Medicine, Université de Montréal, Dental Medicine, Université de Montréal,
Montreal, Quebec. Montreal, Quebec.
Ms. Lee
is a 4th year dental student, Faculty of Correspondence to: Dr. Thomas T. Nguyen, Assistant Professor,
Dental Medicine, Université de Montréal, Division of Periodontics, McGill Faculty of Dentistry, 2001
Montreal, Quebec. McGill College Avenue, Montreal, QC, H3A 1G1.
Email: thomas.nguyen@mcgill.ca
Acknowledgement: The authors thank Dr. John Syrbu and Dr. Borys
Bilaniuk for their expertise and contribution to this review.
The authors have no declared financial interests.
This article has been peer reviewed.
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 5 of 7
Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
J Can Dent Assoc 2021;87:l7
May 3, 2021
References
1. Yu KH, Beam AL, Kohane IS. Artificial intelligence in healthcare. Nat Biomed Eng. 2018;2(10):719-31.
2. Topol EJ. Deep medicine: how artificial intelligence can make healthcare human again. 1st ed. New York: Basic Books; 2019.
3. Russell SJ, Norvig P. Artificial intelligence: a modern approach. 3rd ed. Hoboken, N.J.: Prentice Hall; 2010.
4. Muller J, Massaron L. Artificial intelligence for dummies. Hoboken, N.J.: John Wiley & Sons; 2018.
5. James G, Witten D, Hastie T, Tibshirani R. An introduction to statistical learning with applications in R. New York: Springer; 2013.
6. Goodfellow I, Bengio Y, Courville A. Deep learning. 1st ed. Cambridge, Mass.: MIT Press; 2016.
7. Nielsen MA. Neural networks and deep learning. Determination Press; 2015. Available: http://neuralnetworksanddeeplearning.com/
(accessed 2021 April 16).
8. Zhang K, Wu J, Chen H, Lyu P. An effective teeth recognition method using label tree with cascade network structure. Comput Med
Imaging Graph. 2018;68:61-70.
9. Tuzoff DV, Tuzova LN, Bornstein MM, Krasnov AS, Kharchenko MA, Nikolenko SI, et al. Tooth detection and numbering in
panoramic radiographs using convolutional neural networks. Dentomaxillofac Radiol. 2019;48(4):20180051.
10. Lee JH, Kim DH, Jeong SN, Choi SH. Detection and diagnosis of dental caries using a deep learning-based convolutional neural
network algorithm. J Dent. 2018;77:106-11.
11. Bader JD, Shugars DA, Bonito AJ. Systematic reviews of selected dental caries diagnostic and management methods. J Dent Educ.
2001;65(10):960-8.
12. Xie X, Wang L, Wang A. Artificial neural network modeling for deciding if extractions are necessary prior to orthodontic treatment.
Angle Orthod. 2010;80(2):262-6.
13. Jung SK, Kim TW. New approach for the diagnosis of extractions with neural network machine learning. Am J Orthod Dentofacial
Orthop. 2016;149(1):127-33.
14. Armitage GC. Development of a classification system for periodontal diseases and conditions. Ann Periodontol. 1999;4(1):1-6.
15. Armitage GC. Learned and unlearned concepts in periodontal diagnostics: a 50-year perspective. Periodontol 2000. 2013;62(1):20-36.
16. Papantonopoulos G, Takahashi K, Bountis T, Loos BG. Artificial neural networks for the diagnosis of aggressive periodontitis
trained by immunologic parameters. PLoS One. 2014;9(3):e89757.
17. Papantonopoulos G, Takahashi K, Bountis T, Loos BG. Using cellular automata experiments to model periodontitis: a first step
towards understanding the nonlinear dynamics of the disease. Int J Bifurcation Chaos. 2013;23(3):1350056.
18. Papantonopoulos G, Takahashi K, Bountis T, Loos BG. Mathematical modeling suggests periodontitis behaves as a non-linear
chaotic dynamical process. J Periodontol. 2013;84(10):e29-39.
19. Lee JH, Kim DH, Jegon SN, Choi SH. Diagnosis and prediction of periodontally compromised teeth using a deep learning-based
convolutional neural network algorithm. J Periodontal Implant Sci. 2018;48(2):of114-23.
20. Hiraiwa T, Ariji Y, Fukuda M, Kise Y, Nakata K, Katsumata A, et al. A deep-learning artificial intelligence system for assessment of
root morphology of the mandibular first molar on panoramic radiography. Dentomaxillofac Radiol. 2019;48(3):20180218.
21. Zhang X, Xiong S, Ma Y, Han T, Chen X, Wan F, et al. A cone-beam computed tomographic study on mandibular first molars in a
Chinese subpopulation. PLoS One. 2015;10(8):e0134919.
22. Xue Y, Zhang R, Deng Y, Chen K, Jiang T. A preliminary examination of the diagnostic value of deep learning in hip osteoarthritis.
PLoS One. 2017;12(6):e0178992.
23. Wang X, Yang W, Weinreb J, Han J, Li Q, Kong X, et al. Searching for prostate cancer by fully automated magnetic resonance
imaging classification: deep learning versus non-deep learning. Sci Rep. 2017;7(1):15415.
24. Trebeschi S, van Griethuysen JJM, Lambregts DMJ, Lahaye MJ, Parmar C, Bakers FCH, et al. Deep learning for fully-automated
localization and segmentation of rectal cancer on multiparametric MR. Sci Rep. 2017;7(1):5301.
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 6 of 7
Use of Artificial Intelligence in Dentistry:
Current Clinical Trends and Research Advances
J Can Dent Assoc 2021;87:l7
May 3, 2021
25. Halicek M, Lu G, Little JV, Wang X, Patel M, Griffith CC, et al. Deep convolutional neural networks for classifying head and neck
cancer using hyperspectral imaging. J Biomed Opt. 2017;22(6):60503.
26. Poedjiastoeti W, Suebnukarn S. Application of convolutional neural network in the diagnosis of jaw tumors. Healthc Inform Res.
2018;24(3):236-41.
27. Char DS, Shah NH, Magnus D. Implementing machine learning in healthcare — addressing ethical challenges. N Eng J Med.
2018;378(11):981-3.
28. He J, Baxter SL, Xu J, Xu J, Zhou X, Zhang K. The practical implementation of artificial intelligence technologies in medicine. Nat
Med. 2019;25(1):30-6.
29. Software as medical device (SaMD). Maryland: United States Food & Drug Administration; 2018. Available: https://www.fda.gov/
medical-devices/digital-health/software-medical-device-samd (accessed 2019 June 7).
30. Redman TC. If your data is bad, your machine learning tools are useless. Harv Bus Rev. 2018;2 April. Available: https://hbr.
org/2018/04/if-your-data-is-bad-your-machine-learning-tools-are-useless (accessed 2019 June 12).
31. Murphy KP. Machine learning: a probabilistic perspective. Cambridge, Mass.: MIT Press, 2012.
32. Ferro AS, Nicholson K, Koka S. Innovative trends in implant dentistry training and education: a narrative review. J Clin Med.
2019;8(10):1618.
J Can Dent Assoc 2021;87:l7 ISSN: 1488-2159 7 of 7
