Rokhshad et al.

Maxillofacial Plastic and

Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14
https://doi.org/10.1186/s40902-023-00382-w Reconstructive Surgery

REVIEW Open Access

Artificial intelligence applications and ethical

challenges in oral and maxillo‑facial cosmetic
surgery: a narrative review
Rata Rokhshad1,2* , Seied Omid Keyhan3,4,5,6 and Parisa Yousefi5

Abstract
Artificial intelligence (AI) refers to using technologies to simulate human cognition to solve a specific problem. The
rapid development of AI in the health sector has been attributed to the improvement of computing speed, exponen-
tial increase in data production, and routine data collection. In this paper, we review the current applications of AI for
oral and maxillofacial (OMF) cosmetic surgery to provide surgeons with the fundamental technical elements needed
to understand its potential. AI plays an increasingly important role in OMF cosmetic surgery in various settings, and
its usage may raise ethical issues. In addition to machine learning algorithms (a subtype of AI), convolutional neural
networks (a subtype of deep learning) are widely used in OMF cosmetic surgeries. Depending on their complexity,
these networks can extract and process the elementary characteristics of an image. They are, therefore, commonly
used in the diagnostic process for medical images and facial photos. AI algorithms have been used to assist surgeons
with diagnosis, therapeutic decisions, preoperative planning, and outcome prediction and evaluation. AI algorithms
complement human skills while minimizing shortcomings through their capabilities to learn, classify, predict, and
detect. This algorithm should, however, be rigorously evaluated clinically, and a systematic ethical reflection should
be conducted regarding data protection, diversity, and transparency. It is possible to revolutionize the practice of
functional and aesthetic surgeries with 3D simulation models and AI models. Planning, decision-making, and evalua-
tion during and after surgery can be improved with simulation systems. A surgical AI model can also perform time-
consuming or challenging tasks for surgeons.
Keywords Artificial intelligence, Deep learning, Machine learning, Orthognathic surgery, Rhinoplasty

Background
AI (artificial intelligence) is the creation of machines
capable of performing tasks that usually require humans.
*Correspondence: Its use dates to the 1950s. As a subfield of AI, machine
Rata Rokhshad
[email protected] learning uses algorithms to learn intrinsic statistical
1
Topic Group Dental Diagnostics and Digital Dentistry, ITU/WHO Focus patterns and structures in data, allowing predictions of
Group AI on Health, Berlin, Germany
2
yet-unknown variables (Fig. 1). Data-driven algorithms
Department of Medicine, Boston University Medical Center, Boston, MA,
USA can be built by machines, and thus, they can solve pre-
3
College of Dentistry, Department of Oral & Maxillofacial Surgery, diction problems without human intervention. Artificial
Gangneung-Wonju National University, Gangneung, South Korea
4
neural networks (ANN) mimic the human brain nonlin-
Department of Oral & Maxillofacial Surgery, University of Florida, College
of Medicine, Jacksonville, FL, USA early using artificial neurons like human neural networks.
5
Maxillofacial Surgery & Implantology & Biomaterial Research The neural network can simulate human cognitive capa-
Foundation, Tehran, Iran
6
bilities, such as solving problems, making decisions, and
Iface Academy, Atlanta, GA, USA
learning new things [1, 2]. A “deep learning” architecture

This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2023. Open
Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation,
distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is
not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the
permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​
mmons.​org/​licen​ses/​by/4.​0/.
Rokhshad et al. Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14 Page 2 of 7

predictions, and evaluations of outcomes are all sup-

ported by algorithms. Algorithms also support treatment
decisions and preoperative procedures. Orthognathic
surgery, oral cancerology, and oral surgery are fields
where machine learning techniques are applied [10].
Moreover, AI models have been used to classify implants
on radiographs, predict osteointegration success and
peri-implantitis implant survival, and optimize implant
design parameters such as porosity, length, and diameter
to minimize stress at the implant-bone interface [10].
Technology advancements and digitization have also
made AI increasingly prevalent in OMF cosmetic surger-
ies. In many fields, computers can now provide second
opinions. Making diagnosis more accurate, rapid, and
efficient by using AI in OMF cosmetic surgeries is pos-
sible. This narrative review was prompted by the rapid
development of AI in OMF cosmetic surgeries, and the
emergence of new studies related to them. We aimed
Fig. 1 Deep learning, machine learning, and artificial intelligence
to present a general overview of how AI can be used in
modern OMF cosmetic surgeries in this study.

refers to a multilayered neural network. These models are Main text

beneficial for complex data structures, such as images, Throughout January 2023, the following databases
because they can represent the image and its hierarchi- were electronically searched: PubMed, Web of Sci-
cal features, such as edges, corners, shapes, and macro ence, Google Scholar, Arxiv, Embase, Scopus, IEEE, and
patterns [1–3]. The convolutional neural network (CNN) medRxiv (Table 1). Two authors conducted the screen-
is one of the most common subclasses of ANN in medi- ing procedure independently (RR, PY). The search was
cine and dentistry. For processing digital signals, such conducted using the following keywords: artificial intel-
as sound, images, and videos, CNNs use a special neu- ligence, machine learning, deep learning, neural network,
ron connection architecture and mathematical operation machine intelligence, cosmetic/aesthetic/facial, and sur-
convolution [4]. gery. An evaluation of publications focusing on AI in
In medicine, these technologies have been adopted plastic surgery was conducted.
widely, mainly in computer vision. MYCIN, a rule-based Based on the keywords used in the database query,
expert system trained to distinguish various bacterial all studies were reviewed for applicability. Publications
infections, was the first AI in healthcare to be imple- evaluating AI models segmentation, object detection, or
mented in the early 1980s [5]. Although AI was intro- classification task in plastic surgery were included fol-
duced into healthcare relatively early, it was in 2005 that lowing an abstract screening. We conducted a primary
the first neural network algorithm estimated burn recov- abstract screening to exclude any articles not related to
ery time in plastic surgery [6]. During the last 10 years, the application of AI in plastic surgery. Secondly, articles
AI has made remarkable progress in plastic surgery, espe- with full-text access or available in English were included
cially after face recognition algorithms and deep learn- for a secondary screening. Moreover, review articles were
ing were implemented. Three-dimensional models were excluded.
introduced for presurgical planning in the 2000s. Several
AI applications in maxillofacial surgery utilize digital
Rhinoplasty
imaging, 3D photography, intraoral scans, and 3D photo-
An essential feature of machine learning models, such as
graphs to predict results and plan surgeries, for example,
artificial neural networks, is the ability to classify as one
after skeletal trauma [7–9]. Using some of these models,
of its influencing factors efficiently and thus has been
it was possible to predict the implant size for augmenta-
considered to be a better option for detecting nasal bones
tion rhinoplasty. The postoperative morphology was also
because of their ability to rapidly depict the interdepend-
predicted using image processing algorithms and quanti-
ence between the nasal bone and facial landmark. Pre-
tative measurements of nasal changes [2]. A strong rep-
dicting fractures using CNNs and R-CNNs is essential to
resentation of convolutional neural networks is used in
early detection and well-planned surgery. In order to pre-
the algorithms related to machine learning. Diagnoses,
dict nasal problems, various machine learning techniques
Rokhshad et al. Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14 Page 3 of 7

Table 1 The specific search query for each database (till 30th December 2022)
Database Keywords Results

PubMed ("Artificial intelligence" OR "Machine learning" OR "deep learning" OR "neural network" OR "machine intelligence") AND 256
("cosmetic" OR "aesthetic" OR "facial") AND "surgery"
Google Scholar allintitle:("Artificial intelligence" OR "Machine learning" OR "deep learning" OR "neural network" OR "machine intelligence") 21
AND ("cosmetic" OR "aesthetic" OR "facial") AND "surgery"
Scopus TITLE-ABS-KEY ( ( "Artificial intelligence" OR "Machine learning" OR "deep learning" OR "neural network" OR "machine 245
intelligence" ) AND ( "cosmetic" OR "aesthetic" OR "facial" ) AND "surgery" )
Embase (’artificial intelligence’/exp OR ’artificial intelligence’ OR ’machine learning’/exp OR ’machine learning’ OR ’deep learning’/ 701
exp OR ’deep learning’ OR ’neural network’/exp OR ’neural network’ OR ’machine intelligence’/exp OR ’machine intelligence’)
AND (’cosmetic’/exp OR ’cosmetic’ OR ’aesthetic’ OR ’facial’) AND (’surgery’/exp OR ’surgery’)
Web of Science ("Artificial intelligence" OR "Machine learning" OR "deep learning" OR "neural network" OR "machine intelligence") AND 178
("cosmetic" OR "aesthetic" OR "facial") AND "surgery"
IEEE ("Artificial intelligence" OR "Machine learning" OR "deep learning" OR "neural network" OR "machine intelligence") AND 65
("cosmetic" OR "aesthetic" OR "facial") AND "surgery"

have been used, including back-propagation neural net- prediction algorithms. Machine learning algorithms can
works (BPNNs) for identifying nose bones, random for- predict the results of each surgeon if photographs taken
ests, and support vector machines [11–13]. during perioperative surgery are available. Computerized
There are lots of technical challenges associated with simulations can become more realistic by incorporating
rhinoplasty in plastic surgery. Computer-aided models AI.
would significantly benefit rhinoplasty over any other In a study by Dorfman et al. [14], a detailed photo-
cosmetic surgery because of its complexity and sig- graphic analysis involving 68 facial measurements and
nificant aesthetic and functional consequences for the 128 data points generated for each photo, combined with
patient. Considering the inherently visual nature of rhi- a ranking CNN algorithm (Microsoft Azure Face API),
noplasty, AI applications are a fertile field. A machine eliminates human layperson error and yields an accurate
learning algorithm can recognize hidden patterns and estimate of human age. A CNN algorithm has also been
accurately predict outcomes based on this visual nature, shown to outperform human references when estimat-
which can be converted into raw data. Moreover, as so ing age. The CNN algorithm was programmed to resize
many pre- and postoperative photos are available, it is and crop all included patient photos so that the eyes and
relatively easy to create a rich database of pre-and post- lips of every patient could be measured from a standard
operative photos to train the algorithms for a complete location. Not only were all included patient photos fron-
predictive accuracy that humans cannot match. AI mod- tal shots with the face in a neutral pose, but the CNN
els designed especially for rhinoplasty have been iden- algorithm was also specifically programmed to measure
tified; these models use different AI domains and are the eyes and lips of all included patients from a standard
implemented at various stages of preoperative planning location. As demonstrated by the correlation coefficient
and postoperative outcomes assessment. of 0.9, both smiling and frowning do not affect mood,
However, the gap between ideal simulation and actual self-perception, and, ultimately, age determination [6, 14,
outcomes in rhinoplasty prevents us from maximizing 15]. Moreover, through a combination of 3D image regis-
the benefits of technological implementation in rhino- tration technology and databases, Zeng et al. developed a
plasty. However, 3D simulations and computerized analy- virtual planning system to accurately estimate the dimen-
ses benefit surgeons in preoperative and postoperative sions of forehead flaps used for nasal defect reconstruc-
management. There is an apparent gap in patient satisfac- tion (Fig. 2) [16].
tion with rhinoplasty, as reflected in the relatively high Using perioperative photographs, Chinski et al. [15]
revision rates. However, machine learning can help fill developed an AI model that accurately simulates the out-
this gap with its predictive capabilities and pattern rec- comes of rhinoplasty surgeries. Residents and specialists
ognition abilities. It is possible to train machine learning in otolaryngology evaluated simulations created by the
algorithms using perioperative photographs to produce model. A surgeon’s simulation image and the AI model’s
more realistic simulations based on the 3D models. It image were shown randomly to the evaluators. Using a
is more accurate and realistic to make predictions with Likert scale, the participants expressed how much they
machine learning when combined with perioperative agreed with the simulations. AI simulations were agreed
photography databases rather than with ideal robotic upon by 68.4% of evaluators, while surgeon simulations
Rokhshad et al. Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14 Page 4 of 7

Fig. 2 Using 2D photographs to create 3D models [1]

were agreed upon by 77.3%. However, despite higher surgery. Based on the results, the models have shown a
agreement rates among experts in the surgeon’s simu- high degree of accuracy of 85 to 95.6% [20]. Addition-
lations, the model achieved promising results. Before ally, AI models predicted the perioperative blood loss
meeting in person with a particular surgeon, patients can and systemic infections following orthognathic surgery
generate a realistic simulation of the postoperative out- in addition to predicting the future need for orthognathic
come to form an accurate appraisal of the potential out- surgery [4, 19, 21].
come [6, 15]. In a study by Hong et al., the accuracy of AI-assisted
cephalometric landmark detection in jaw orthognathic
Orthognathic surgeries surgery cases had been reported to be 75%, even when
An orthognathic surgeon’s clinical experience is essential orthodontic brackets, surgical plates, screws, fixed
to creating a detailed treatment plan, and the plan plays retainers, genioplasty, and bone remodeling were present
a vital role in the outcome [17]. As a surgeon designs [22]. A CNN model was 94.4% accurate for diagnosing
and fabricates surgical splints based on CT (computed orthognathic surgery cases using both lateral and frontal
tomography scan) or CBCT (Cone-beam computed cephalograms [23]. According to a study by Jeong et al.,
tomography systems) models, 3D craniomaxillofacial deep learning CNN was able to accurately diagnose sur-
features are automatically registered [18, 19]. Thus, gical patients based on facial photographs (frontal and
measuring the amount and direction of hard and soft tis- lateral) [24]. In a study by Tanikawa et al., they evaluated
sue movements in 3D before orthognathic surgery can AI for predicting the 3D facial shape after orthodontic
be valuable for determining the amount and direction treatment and orthognathic surgery (Fig. 3) [25]. This
of surgical interventions. Due to cleft-related deformi- confirms the clinical acceptability of AI systems for pre-
ties and scar tissue, it is especially beneficial for treating dicting facial morphology after treatment.
patients with clefts as their soft tissues differ both mor-
phologically and behaviorally from those of non-cleft Future applications
patients [18]. AI can be used to identify precise land- AI has great potential for OMF cosmetic surgeries
marks, analyze rapid digital cephalometric data, make as one of many specialties. Thinking machines could
clinical decisions, and predict treatment outcomes using improve the efficiency and patient satisfaction of plastic
software enabled by AI. surgeons’ diagnostic, case-planning, and perioperative
Additionally, artificial intelligence has been applied to tasks. In order to make surgical decisions, the surgeon
presurgical orthopedics, speech pathology detection, and must be able to create an appropriate differential diag-
the need to predict the need for CLP (cleft lip and palate) nosis list, determine the best tests for establishing the
Rokhshad et al. Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14 Page 5 of 7

Fig. 3 In a study by Tanikawa et al., facial morphology in Japanese patients after orthognathic surgery and orthodontic treatment was predicted
using AI [2]. Surgery group pre-treatment actual facial changes (top-left) and the superimposition of pre-treatment and post-treatment actual facial
changes (blue and yellow). After surgery, the blue column indicates a downward displacement after treatment, and the yellow column indicates
an upward displacement after treatment. When viewed antero-posteriorly, yellow indicates protrusion, while blue indicates retrusion. A custom
MATLAB program was used to create the figures

diagnosis, and devise a plan to deal with the diagnosis Ethical issues
using heuristic techniques and informed judgment. AI- Multiple ethical concerns arise when AI is introduced
enabled decision-making tools combined with predic- into OMF cosmetic surgeries, especially in a plastic sur-
tive analytics and integrating human surgical intuition geon’s clinical practice. Many ethical dilemmas arise from
hold great promise for improving surgical outcomes. AI systems that claim to classify attractiveness objec-
The surgeon could make real-time decisions periopera- tively. Ethnicity and gender can be discriminated against.
tively based on 3D planning, anatomical localization, In isolation, AI could lead to the propagation of racial
and surgical navigation [15, 18, 21, 26]. Complex sur- division and a loss of diversity in cosmetic surgeries [27].
gical procedures cannot be performed with current AI Dataset size is also a significant limitation, particularly
tools. They may, however, become capable of perform- when training CNN, which is particularly data intensive.
ing more complex tasks in the future. OMF cosmetic Training data, algorithms, parameters, and quality influ-
surgeries could be improved through technological ence the training data needed [10, 28, 29]. Most of the
advancements, reducing the amount of time spent studies on applications of AI in OMF cosmetic surgeries
anesthetizing patients and decreasing their recovery had limited sources of datasets.
time after surgery. This technology also presents excit- A small dataset can be overcome in several ways. Spe-
ing opportunities for improving surgical outcomes in cific data augmentation techniques can partially address
low- and middle-income countries with a need for sur- the problem, especially in geometric deformation image
geons and their expertise and limited resources. processing. However, collecting datasets from different
Similarly, the armed forces may use AI surgical centers containing different genders, ages, and nation-
machines to treat injuries far away from medical cent- alities will increase the generalizability of the AI model.
ers. It is still being determined whether robot-assisted Black patients and providers are underrepresented in
surgery is cost-effective in plastic surgery, as in other rhinoplasty and blepharoplasty [30]. It is essential to
specialties. A national healthcare system should deter- discuss the validity of assessing attractiveness based
mine this before implementing it widely. on data obtained from a dating platform. According to
Rokhshad et al. Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14 Page 6 of 7

general definitions, attractiveness is the ability to create Author’s information

Not applicable.
interest and desire in observers. The definition is indeed
subject to subjectivity and cultural influences. AI-based Funding
measurements are only quantifiable representations of This research did not receive any specific grant from funding agencies in the
public, commercial, or not-for-profit sectors.
opinions, regardless of how precise they are [31]. As an
example of such bias, facial recognition systems may be Availability of data and materials
used in aesthetic practices. Using data sets from different Not applicable.
nationalities and countries might marginalize other cul-
tures’ values and perceptions of beauty. Furthermore, AI Declarations
should not take the place of shared decision-making to Ethics approval and consent to participate
achieve the best quality of patient care. Due to the limita- Not applicable.
tions that such technology imposes on its data sets, pro-
Consent for publication
viders should ensure that a biased view does not disrupt Not applicable.
shared decision-making.
Competing interests
This paper’s results and/or discussion are not influenced by any competing
Limitations interests of authors.
A narrative review reveals consolidated knowledge and
the need for additional research in a field of research, in
Received: 8 January 2023 Accepted: 1 March 2023
contrast to a systematic review which explains the quality
and reliability of existing knowledge (including the risk of
bias assessment). Furthermore, the data search was lim-
ited to English-language articles, so studies in other lan- References
guages were not included. 1. Flynn JP, Cha E, Flynn TJ, Kriet JD, Humphrey CD (2021) Machine learning
generated synthetic faces for use in facial aesthetic research. Facial Plast
Surg Aesthet Med. 23:339–343
Conclusions 2. Ossowska A, Kusiak A, Świetlik D (2022) Artificial intelligence in den-
It has been shown that AI accurately predicts rhino- tistry—narrative review. Int J Environ Res Public Health. 19:3449
plasty outcomes or whether orthognathic surgery will 3. Wang C, Zhang J, Lassi N, Zhang X (2022) Privacy protection in using
artificial intelligence for healthcare: Chinese regulation in comparative
be needed in the future for cleft patients. However, it is perspective. Healthcare. 10:1878
imperative to improve existing models by analyzing dif- 4. Siddiqui A, Sukhia RH, Ghandhi D (2022) Artificial intelligence in dentistry,
ferent datasets, surgical methods, and a variety of sur- orthodontics and Orthognathic surgery: a literature review. J Pak Med
Assoc. 72:91–96
geries and populations to get flawless and generalized 5. Swartout WR (1985) Rule-based expert systems: the mycin experiments
results. Identifying the study gaps will require more clini- of the stanford heuristic programming project. Artif Intell. 26:364–366
cal research incorporating larger sample sizes, cross-vali- 6. Eldaly AS, Avila FR, Torres-Guzman RA, Maita K, Garcia JP, Palmieri Ser-
rano L et al (2022) Simulation and artificial intelligence in Rhinoplasty: a
dation, and inter-model comparisons. By considering the systematic review. Aesthetic Plast Surg. 46:2368–2377
current limitations, developing the AI model can assist 7. Pereira KR, Sinha R (2020) Welcome the “new kid on the block” into the
clinicians in decision-making and provide an additional family: artificial intelligence in oral and maxillofacial surgery. Bri J Oral
Maxillofac Surg 58:83–84
tool for future diagnosis and treatment planning in OMF 8. McCullough M, Ly S, Auslander A, Yao C, Campbell A, Scherer S et al
cosmetic surgeries. (2021) Convolutional neural network models for automatic preop-
erative severity assessment in unilateral cleft lip. Plast Reconstr Surg.
148:162–169
Abbreviations 9. Spoer DL, Kiene JM, Dekker PK, Huffman SS, Kim KG, Abadeer AI et al
AI Artificial intelligence (2022) A systematic review of artificial intelligence applications in plastic
ANN Artificial neural networks surgery: looking to the future. Plast Reconstr Surg Glob Open. 10:4608
BPNN Back-propagation neural networks 10. Rasteau S, Ernenwein D, Savoldelli C, Bouletreau P (2022) Artificial intel-
CBCT Cone-beam computed tomography systems ligence for oral and maxillo-facial surgery: A narrative review. J Stomatol
CT Computed tomography scan Oral Maxillofac Surg. 123:276–282
CLP Cleft lip and palate 11. Franke-Gromberg C, Schüler G, Hermanussen M, Scheffler C (2010) Digital
CNN Convolutional neural network 2D- Photogrammetry and Direct Anthropometry A Comparing Study
OMF Oral and maxillo-facial on Test Accomplishment and Measurement Data. Anthropologischer
NN Neural networks Anzeiger. 68:11–20
12. Chowdhury NI, Li P, Chandra RK, Turner JH (2020) Baseline mucus
Acknowledgements cytokines predict 22- item sino- nasal outcome test results after endo-
Not applicable. scopic sinus surgery. Int Forum Allergy Rhinol. 10:15–22
13. Tuan HNA, Hai NDX, Thinh NT (2022) Shape prediction of nasal bones by
Authors’ contributions digital 2D-photogrammetry of the nose based on convolution and back-
The article was written by RR and SOK. RR and PY collected the data. SOK and propagation neural network. Comput Math Methods Med. 2022:1–18
PY edited the manuscript and performed the critical review. The authors read 14. Dorfman R, Chang I, Saadat S, Roostaeian J (2020) Making the subjective
and approved the final manuscript. objective: machine learning and rhinoplasty. Aesthet Surg J. 40:493–498
Rokhshad et al. Maxillofacial Plastic and Reconstructive Surgery (2023) 45:14 Page 7 of 7

15. Chinski H, Lerch R, Tournour D, Chinski L, Caruso D (2022) An artificial

intelligence tool for image simulation in rhinoplasty. Facial Plastic Surg
38:201–206
16. Zeng W, Chen G, Ju R, Yin H, Tian W, Tang W (2018) The combined applica-
tion of database and three-dimensional image registration technology in
the restoration of total nose defect. J Craniofacial Surg 29:484–487
17. Mohaideen K, Negi A, Verma DK, Kumar N, Sennimalai K, Negi A (2022)
Applications of artificial intelligence and machine learning in orthog-
nathic surgery: a scoping review. J Stomatol Oral Maxillofac Surg.
123:962–972
18. Bichu YM, Hansa I, Bichu AY, Premjani P, Flores-Mir C, Vaid NR (2021) Appli-
cations of artificial intelligence and machine learning in orthodontics: a
scoping review. Prog Orthod. 22:18
19. Bouletreau P, Makaremi M, Ibrahim B, Louvrier A, Sigaux N (2019) Artificial
intelligence: applications in orthognathic surgery. J Stomatol Oral Maxil-
lofac Surg. 120:347–354
20. Huqh MZU, Abdullah JY, Wong LS, Jamayet NB, Alam MK, Rashid QF et al
(2022) Clinical applications of artificial intelligence and machine learning
in children with cleft lip and palate—a systematic review. Int J Environ
Res Public Health. 19:10860
21. Patcas R, Bornstein MM, Schätzle MA, Timofte R (2022) Artificial intel-
ligence in medico-dental diagnostics of the face: a narrative review of
opportunities and challenges. Clin Oral Investig. 26:6871–6879
22. Hong M, Kim I, Cho JH, Kang KH, Kim M, Kim SJ et al (2022) Accuracy
of artificial intelligence-assisted landmark identification in serial lateral
cephalograms of Class III patients who underwent orthodontic treatment
and two-jaw orthognathic surgery. Korean J Orthod. 52:287–297
23. Shin W, Yeom HG, Lee GH, Yun JP, Jeong SH, Lee JH et al (2021) Deep
learning based prediction of necessity for orthognathic surgery of
skeletal malocclusion using cephalogram in Korean individuals. BMC Oral
Health. 21:130
24. Jeong SH, Yun JP, Yeom HG, Lim HJ, Lee J, Kim BC (2020) Deep learning
based discrimination of soft tissue profiles requiring orthognathic surgery
by facial photographs. Sci Rep. 10:16235
25. Tanikawa C, Yamashiro T (2021) Development of novel artificial intel-
ligence systems to predict facial morphology after orthognathic surgery
and orthodontic treatment in Japanese patients. Sci Rep. 11:15853
26. Jarvis T, Thornburg D, Rebecca AM, Teven CM (2020) Artificial intelligence
in plastic surgery. P last Reconstr Surg Glob Open. 8:3200
27. Murphy D, Saleh D (2020) Artificial intelligence in plastic surgery: what is
it? Where are we now? What is on the horizon? Ann Royal College Surg
England. 102:577–580
28. Sun W, Cai Z, Li Y, Liu F, Fang S, Wang G (2018) Data processing and text
mining technologies on electronic medical records: a review. J Healthc
Eng. 2018:4302425
29. Kohli MD, Summers RM, Geis JR (2017) Medical image data and datasets
in the era of machine learning—whitepaper from the 2016 C-MIMI Meet-
ing Dataset Session. J Digit Imaging. 30:392–399
30. Ullrich PJ, Garg S, Reddy N, Hassan A, Joshi C, Perez L et al (2022) The
racial representation of cosmetic surgery patients and physicians on
social media. Aesthet Surg J. 42:956–963
31. Patcas R, Bernini DAJ, Volokitin A, Agustsson E, Rothe R, Timofte R (2019)
Applying artificial intelligence to assess the impact of orthognathic
treatment on facial attractiveness and estimated age. Int J Oral Maxillofac
Surg. 48:77–83

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.
© This is a U.S. Government work and not under copyright protection in the
US; foreign copyright protection may apply 2023. This work is published
under http://creativecommons.org/licenses/by/4.0/(the “License”).
Notwithstanding the ProQuest Terms and Conditions, you may use this
content in accordance with the terms of the License.