The landscape of virtual simulation in undergraduate dental education with implications for artificial intelligence incorporation

Abstract

Virtual simulation (VS) is pivotal in undergraduate dental education for cognitive and preclinical training. Expanding VS adoption and artificial intelligence (AI) incorporation necessitate an evidence synthesis to clarify current landscapes, challenges, and future directions. Following the PRISMA-ScR guidelines and Arksey and O’Malley’s framework, this scoping review included 57 studies (2007–2025) to map the landscape of VS and its AI-driven future. The included VS systems could be broadly categorized into two generations, with first-generation systems supporting knowledge acquisition and second-generation systems supporting procedural skill development. VS applications spanned multiple dental subspecialties with uneven technological maturity and depth. With assessment of examinations, expert-rated performance, simulator-derived metrics, and learner-reported questionnaires, most studies reported that VS improved learner performance and confidence. Challenges persist regarding uncertain long-term knowledge retention, incomplete replication of clinical settings, resource-intensive implementation, and the limited role in education. This review further explored the implications of future AI–VS incorporation, including AI-driven virtual patients and tutors, fidelity enhancement, adaptive feedback, and personalized competency-based assessment.

Introduction

Simulation-based education has been consistently shown to enhance learning outcomes in medical education^1,2. As a discipline emphasizing both theoretical knowledge and technical skill development equally, dental education has long relied on various simulation aids to support undergraduate education^3,4. Since the establishment of the first dental surgery school in Baltimore, MD, USA, in 1840, simulation technology has been gradually integrated into dental education⁵. With the growth of educational demands, phantom-head simulators emerged in 1894^1,6. Nowadays, phantom-head simulators equipped with plastic teeth, handpieces, and water spray systems to simulate clinical scenarios have become an essential aid in dental education¹. In the late 1990s, computer-supported virtual simulation (VS) was introduced into dental schools⁷. Due to its high fidelity in replicating clinical scenarios, VS has pioneered a new approach to learning dental motor skills and has become an increasingly important component of undergraduate dental education^1,8.

From a technological perspective, VS has developed in an evolutionary path consisting of two distinct generations. The first generation focuses on basic cognitive skills and utilizes interactive video technology to enhance knowledge acquisition and establish standardized learning environments^9,10. The second generation further optimizes standardized learning environments and emphasizes practical skills training. It integrates highly realistic head models with force feedback devices and partially incorporates haptic feedback and real-time assessment mechanisms^7,11. However, both generations are still constrained by traditional technological frameworks and present apparent development bottlenecks. From an application perspective, VS has been integrated into all subspecialties and foundational disciplines of dental medicine¹. It addresses core challenges in traditional teaching through three key approaches: anatomical visualization, basic procedural training, and high-risk surgical simulation^12–14. However, current VS implementations still have limitations: they primarily focus on the development of fundamental clinical skills, while neglecting both advanced technologies and basic medical mechanisms. Additionally, they provide insufficient coverage of rare diseases and complex interdisciplinary cases^1,15. From an outcome perspective, VS in dental education has been shown to improve students’ performances via objective and subjective assessments^16,17. For objective assessments, some VS systems incorporate built-in assessment tools to evaluate students’ performance based on accuracy, while most assessments rely on students’ exam scores. Subjective assessments have mainly relied on questionnaires regarding students’ learning experiences. However, current VS assessment lacks the continuous trace for students’ performances, outcome-feedback adjustments and subsequent individualized learning settings, which fails to achieve competency-based learning trajectories^11,18.

In recent years, the rapid development of artificial intelligence (AI) technology has provided a core driving force for innovation and transformation across all industries, and it has demonstrated enormous potential in optimizing teaching models and improving training quality^19,20. Modern dental education has a growing imperative to equip students with advanced clinical skills, support individualized competency-based learning trajectories, and prepare them to manage complex and multidisciplinary clinical scenarios²¹. Parallel to these educational demands, the rapid evolution of AI technology is reshaping VS through AI integration. The integration of AI into VS systems could enable advanced teaching approaches, including adaptive learning, automated performance analytics, realistic scenario generation, and personalized feedback, collectively positioning VS implications for the incorporation of AI as a powerful driver of innovation in medical and dental education^8,22,23. However, directly implemented empirical studies on AI-enabled VS in undergraduate dental education remain very limited, as advanced technologies like generative AI have not yet been fully implemented^19,24. Thus, to meet the evolving demands of modern dental education, it is essential to develop a comprehensive understanding of the current landscape and the challenges of VS from an AI perspective.

This scoping review primarily aims to outline the landscape and challenges of VS in undergraduate dental education, while also identifying areas in which future AI incorporation may further enhance simulation-based teaching. In line with the objectives of this scoping review, our research questions (RQ) were listed as follows:

RQ1: What are the core technologies and methodologies of VS systems?

RQ2: What are the applied subdisciplines and contents of VS?

RQ3: What are the educational outcomes based on VS assessment systems?

RQ4: What specific challenges impede the implementation of VS systems?

Results

After conducting a thorough search, we found 1420 articles, which were carefully assessed for relevance and eligibility. Ultimately, 57 articles were included in this scoping review based on strict inclusion and exclusion criteria, published between 2007 and 2025 (Fig. 1). From the 57 published studies, VS was reported to be applied in the educational practice involving undergraduate dental students across 18 countries including China (24 studies), the United Kingdom (4), Saudi Arabia (4), Germany (4), Thailand (3), Qatar (3), Sweden (2), France (2), Malaysia (2), Brazil (1), Egypt (1), Greece (1), Ireland (1), Israel (1), Korea (1), Romania (1), Spain (1), and the United Arab Emirates (1). Collectively, these articles provide diverse insights into the application of VS in stomatology undergraduate education (Fig. 2).

Fig. 1. PRISMA flow diagram.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram shows the detailed process of information retrieval and literature screening.

Fig. 2. Geographic distribution and publication trends of included studies.

A Distribution of studies reporting the use of VS in undergraduate dental education by continent. B Cumulative number of publications by year (2007–2025).

Overall, the methodological quality of the included studies was generally acceptable. Most studies showed reasonable clarity in their aims, basic design, and outcome reporting. However, a minority of studies had insufficiently reported methodological details in certain domains, which may introduce some risk of bias. These features should be considered when interpreting the reported educational benefits of VS.

Technologies and methodologies

While the types of VS used in undergraduate dental education have become increasingly diverse, they can be broadly categorized into four main types based on their core learning scenarios and practical orientation: virtual clinical scenarios, virtual dental models, virtual treatment planning, and practical operation simulators. Based on the relevant studies included in this scoping review, we further outline the developmental trajectory of VS in undergraduate dental education, which can be divided into two evolutionary stages. This generational classification was not based on publication year or proprietary platform names, but on the dominant technological architecture, interaction mode, and primary educational function reported across the included studies (Table 1 and Supplementary Table 1).

Table 1.

Condensed characteristics of virtual simulation applications in the included studies

Author, year	Country/region	VS-type	Application scenario	Real instruments	Force feedback	Real-time feedback	Evaluation
Rees, 200738	UK	Practical operation simulators	Cavity preparation	Yes	Not clear	Yes	Yes
Casap, 201155	Israel	Practical operation simulators	Implant surgery	Yes	No	Yes	Yes
Lund, 201161	Sweden	Virtual clinical scenarios, Practical operation simulators	Oral and maxillofacial surgery	Yes	Yes	Yes	Yes
Papadopoulos, 201370	Greece	Virtual clinical scenarios	Behavior management in pediatric dentistry	No	No	Yes	Yes
Rosen, 201465	Sweden	Practical operation simulators	Oral and maxillofacial surgery	Yes	Yes	Yes	Yes
Joseph, 201414	France	Practical operation simulators	Implant surgery	Yes	Yes	Yes	Yes
Juan, 201631	Spain	Virtual dental models	Dental morphology	No	No	Not clear	Yes
Tubelo, 201649	Brazil	Practical operation simulators	Prosthodontic cementation	Yes	No	Yes	Yes
Al-Saud, 201739	UK	Practical operation simulators	Preclinical operative dentistry	Yes	Yes	Yes	Yes
Shahriari-Rad, 201729	UK	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Liu, 2018112	China	Practical operation simulators	Crown preparation	Yes	No	Yes	Yes
Marei, 201887	Saudi Arabia	Virtual clinical scenarios	Oral and maxillofacial surgery	No	No	Yes	Yes
Mirghani, 2018101	UK	Practical operation simulators	Preclinical operative dentistry	Yes	Yes	Yes	Yes
Lone, 201932	Ireland	Virtual dental models	Dental morphology	No	No	Yes	Yes
Zhang, 202056	China	Virtual clinical scenarios, Practical operation simulators	Implant surgery	Yes	Not clear	Yes	Yes
Al-Saud, 202025	Saudi Arabia	Practical operation simulators	Crown preparation	Yes	Yes	Yes	Yes
Liebermann, 202033	Germany	Virtual dental models	Dental morphology	Yes	No	No	Yes
Mai, 202048	Korea	Virtual dental models	Dental occlusion	No	No	No	Yes
Murbay, 202036	China	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Reymus, 202034	Germany	Virtual dental models	Root canal anatomy	No	No	No	Yes
Vincent, 202037	France	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Yin, 202111	Thailand	Practical operation simulators	Root canal treatment	Yes	Yes	Yes	Yes
Zhang, 202126	China	Practical operation simulators	Periodontal treatment	Yes	Not clear	Yes	Yes
Ho, 202266	China	Virtual dental models	Orthodontic treatment	No	No	No	Yes
Li, 202252	China	Practical operation simulators	Crown preparation	Yes	Yes	Yes	Yes
Tee, 202269	Malaysia	Virtual clinical scenarios	Oral mucosal diseases	No	No	Yes	Yes
Shetty, 202373	United Arab Emirates	Virtual treatment planning	Implant surgery	No	No	No	Yes
Huang, 202357	China	Practical operation simulators	Implant surgery	Yes	No	Yes	Yes
Ba-Hattab, 202344	Qatar	Practical operation simulators	Root canal treatment	Yes	Yes	Yes	Yes
Alsufyani, 202312	Saudi Arabia	Virtual dental models	Panoramic radiographic anatomy	Yes	No	Yes	Yes
Luo, 202350	China	Virtual treatment planning	Removable prosthodontics	No	No	No	Yes
Petre, 202354	Romania	Virtual dental models	Prosthodontics	No	No	No	Yes
Daud, 202378	Qatar	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Liebermann, 2023113	Germany	Virtual dental models	Dental morphology	No	No	No	Yes
Hamama, 202441	Egypt	Practical operation simulators	Caries removal	Yes	Yes	No	Yes
Fu, 202413	China	Practical operation simulators	Periodontal treatment	Yes	Yes	Yes	Yes
Duan, 202427	China	Practical operation simulators	Root canal treatment	Yes	Not clear	Yes	Yes
Daud, 202417	Qatar	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Buajeeb, 202477	Thailand	Virtual clinical scenarios	Oral mucosal diseases	No	No	Yes	Yes
Li, 202442	China	Virtual clinical scenarios	Rubber dam isolation	No	No	Yes	Yes
Wang, 202458	China	Practical operation simulators	Periodontal treatment	Yes	Yes	Yes	Yes
Wei, 202443	China	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Wu, 202459	China	Practical operation simulators	Periodontal treatment	Yes	Yes	Yes	Yes
Yao, 202451	China	Virtual clinical scenarios	Reception in prosthodontics	Yes	No	No	Yes
Nassief, 202435	Saudi Arabia	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Peanchitlertkajorn, 202471	Thailand	Virtual clinical scenarios	Dental clinic management	No	No	No	Yes
Liu, 202546	China	Virtual dental models, Virtual clinical scenarios, Practical operation simulators	Apicoectomy	Yes	No	Yes	Yes
Hu, 202547	China	Virtual clinical scenarios	Regenerative endodontics	No	No	Yes	Yes
Lv, 202540	China	Practical operation simulators	Cavity preparation	Yes	Yes	Yes	Yes
Tang, 202574	China	Practical operation simulators	Crown preparation	Yes	Yes	Yes	Yes
Cheng, 202560	China	Practical operation simulators	Periodontal treatment	Yes	Yes	Yes	Yes
Huang, 202545	China	Virtual clinical scenarios	Root canal treatment	No	No	Yes	Yes
Prinz, 202530	Germany	Virtual clinical scenarios	Diagnosis and treatment in endodontics	No	No	Yes	Yes
Guo, 202563	China	Virtual clinical scenarios	Oral and maxillofacial surgery	No	No	Not clear	Yes
Rehman, 202572	Malaysia	Virtual clinical scenarios	Typical multidisciplinary clinical cases	No	No	Yes	Yes
Chen, 202567	China	Virtual clinical scenarios	Orthodontic treatment	Not clear	Not clear	Yes	Yes
Liu, 202564	China	Virtual dental models, Virtual clinical scenarios, Practical operation simulators	Orthognathic surgery	Yes	Not clear	Yes	Yes

First-generation VS systems were defined as cognition-oriented environments that primarily relied on interactive video technology, standard 2D displays, or immersive virtual reality (VR) headsets to support knowledge acquisition, spatial understanding, and standardized procedural cognition. Their applications include: (1) Virtual clinical scenarios simulating disease-specific diagnostic and treatment pathways (e.g., endodontic procedures from diagnosis to treatment, or orthodontic workflows from initial consultation to staged therapy), thereby facilitating the development of clinical reasoning and standardized procedural skills; (2) Virtual dental models for teaching dental anatomy, morphology, and radiographic interpretation (e.g., 3D visualization of dental structures and identification of anatomical landmarks in panoramic and periapical radiographs), enabling multi-angle observation and spatial memory formation; (3) Virtual treatment planning simulating clinical decision-making processes (e.g., restoration design and implant placement planning), enabling students to practice treatment planning and procedural skills in risk-free environments while developing core clinical competencies.

Second-generation VS systems were defined as procedure-oriented simulation environments that built upon interactive video technology and further integrated virtual instrument operation modules with high-fidelity head models, force feedback devices and real-time performance assessment. Representative platforms in this category include Simodont²⁵, Virteasy¹⁴, and UniDental²⁶. These systems focus on developing students’ hand–eye–brain coordination through core clinical procedures such as carious tissue removal, dental preparation, periodontal scaling, root canal preparation, and filling^14,25,27. By incorporating virtual instrument modules with realistic head models or force feedback systems, these advanced training environments enable comprehensive skill development. The latest advancements feature tactile feedback and real-time evaluation systems within virtual instruments, allowing students to perceive simulated dental tissue hardness and resistance during operation. The system simultaneously tracks operational trajectories, force control, and movement precision, while providing playback functionality, error detection, and instant correction prompts^11,28. This enables students to familiarize themselves with procedures, instruments, and equipment before actual patient treatment, facilitating proactive, independent, and repeatable learning that significantly enhances the relevance and effectiveness of pre-clinical skill training (Fig. 3).

Fig. 3. Distribution of VS system types across different dental subspecialties in the included studies.

Functional categories of VS systems are stratified by subspecialty: cariology and endodontics, prosthodontics and implantology, periodontology, oral and maxillofacial surgery, dental morphology, orthodontics, oral mucosal diseases, and other specialities (pediatric dentistry and basic dental sciences).

Overall, the two generations differed more in their dominant teaching roles than in chronology alone. First-generation VS systems were primarily applied to visualization, case-based learning, and structured treatment planning, thereby supporting knowledge acquisition and standardized procedural cognition^9,10. By contrast, second-generation systems were more closely associated with operative training tasks such as cavity preparation, periodontal scaling, root canal procedures, and implant simulation, with greater reliance on haptic feedback, realistic head models, and built-in performance evaluation^11,25,29. These patterns suggest that the evolution of undergraduate dental VS has been marked by a shift from cognition-oriented digital support toward more procedure-oriented and feedback-rich simulation environments. Notably, empirical evidence of AI-integrated VS is exemplified by the pioneering work of Prinz et al. (2025). Their study successfully implemented a large language model (GPT) to act as a virtual patient for endodontic diagnostics, enabling students to engage in realistic, conversational clinical reasoning³⁰. This specific application provides a tangible demonstration of how AI can transform traditional VS from static modules into dynamic, interactive learning environments^8,22,23.

Applied subdisciplines and contents

VS is being increasingly integrated across dental subspecialties. The included studies report applications spanning dental morphology, cariology and endodontics, prosthodontics and implantology, periodontology, oral and maxillofacial surgery, orthodontics, oral mucosal diseases, pediatric dentistry, and basic dental sciences. Across these disciplines, the reported training items consistently serve several shared educational purposes, including anatomical visualization, clinical reasoning and decision-making, treatment planning, and psychomotor skills training (Table 2 and Supplementary Table 2).

Table 2.

Applications, educational outcomes, and challenges of virtual simulation across dental subspecialties

Author, year	Subdiscipline	Training item	Assessment methods	Challenges
Rees, 200738	Cariology and endodontics	G. V. Black Class I and II cavity preparation	Questionnaire	Incomplete equivalence to real-life clinical settings
Casap, 201155	Prosthodontics and implantology	Implant surgery	Questionnaire, Operation test	Uncertain long-term retention of knowledge
Lund, 201161	Oral and maxillofacial surgery	Mandibular third molar surgery	Questionnaire	Incomplete equivalence to real-life clinical settings
Papadopoulos, 201370	Other	Pediatric dental communication and behavior management technology	Questionnaire	Incomplete equivalence to real-life clinical settings, non-substitutive supplement to traditional training
Rosen, 201465	Oral and maxillofacial surgery	Mandibular third molar surgery	Operation test, Questionnaire	Non-substitutive supplement to traditional training
Joseph, 201414	Prosthodontics and implantology	Implant surgery	Operation test	Uncertain long-term retention of knowledge
Juan, 201631	Dental morphology	Morphological structures of the mandibular first molar	Theory test, Questionnaire	Incomplete equivalence to real-life clinical settings, non-substitutive supplement to traditional training
Tubelo, 201649	Prosthodontics and implantology	Zinc phosphate cement manipulation	Theory test, Operation test	Uncertain long-term retention of knowledge
Al-Saud, 201739	Cariology and endodontics	Hand skills training	Operation test	Non-substitutive supplement to traditional training
Shahriari-Rad, 201729	Cariology and endodontics	Caries removal and cavity preparation	Theory test, Operation test	Non-substitutive supplement to traditional training, with a resources-intensive nature
Liu, 2018112	Prosthodontics and implantology	Ceramic crown preparation	Questionnaire, Operation test	Incomplete equivalence to real-life clinical settings
Marei, 201887	Oral and maxillofacial surgery	Impacted wisdom teeth and maxillofacial trauma	Theory test, Operation test	Incomplete equivalence to real-life clinical settings, resource-intensive nature, non-substitutive supplement to traditional training
Mirghani, 2018101	Cariology and endodontics	Hand skills training	Operation test	Non-substitutive supplement to traditional training, Incomplete equivalence to real-life clinical settings
Lone, 201932	Dental morphology	Identification of permanent teeth	Questionnaire, Theory test	Incomplete equivalence to real-life clinical settings, non-substitutive supplement to traditional training
Zhang, 202056	Prosthodontics and implantology	Implant surgery	Theory test, Operation test, Questionnaire	Uncertain long-term retention of knowledge, Incomplete equivalence to real-life clinical settings
Al-Saud, 202025	Cariology and endodontics	Clinical crown preparation	Operation test	Incomplete equivalence to real-life clinical settings
Liebermann, 202033	Dental morphology	Tooth morphology	Questionnaire	Resources-intensive nature, non-substitutive supplement to traditional training
Mai, 202048	Prosthodontics and implantology	Occlusal adjustment	Questionnaire, Theory test	Resources-intensive nature
Murbay, 202036	Cariology and endodontics	SISTA Classification-Site 1 cavity preparation	Operation test	Non-substitutive supplement to traditional training
Reymus, 202034	Dental morphology	Root canal anatomy	Theory test, Operation test, Questionnaire	Non-substitutive supplement to traditional training
Vincent, 202037	Cariology and endodontics	G. V. Black Class II cavity preparation	Operation test	Incomplete equivalence to real-life clinical settings, non-substitutive supplement to traditional training
Yin, 202111	Cariology and endodontics	Access cavity preparation	Operation test	Non-substitutive supplement to traditional training
Zhang, 202126	Periodontology	Supragingival scaling	Theory test, Operation test, Questionnaire	Uncertain long-term retention of knowledge
Ho, 202266	Orthodontics	Orthodontic diagnostic	Theory test	Incomplete equivalence to real-life clinical settings, non-substitutive supplement to traditional training
Li, 202252	Prosthodontics and implantology	Ceramic crown preparation	Questionnaire	Incomplete equivalence to real-life clinical settings
Tee, 202269	Oral medicine	Management of oromucosal lesions	Questionnaire, Theory test	Non-substitutive supplement to traditional training
Shetty, 202373	Prosthodontics and implantology	Implant planning	Theory test	Uncertain long-term retention of knowledge
Huang, 202357	Prosthodontics and implantology	Implant surgery	Questionnaire	Non-substitutive supplement to traditional training
Ba-Hattab, 202344	Cariology and endodontics	Access cavity preparation	Questionnaire	Uncertain long-term retention of knowledge
Alsufyani, 202312	Dental morphology	Panoramic radiographic anatomy	Questionnaire	Uncertain long-term retention of knowledge, non-substitutive supplement to traditional training
Luo, 202350	Prosthodontics and implantology	Removable partial denture framework design	Theory test, Operation test, Questionnaire	Resources-intensive nature
Petre, 202354	Prosthodontics and implantology	Partial edentulism cases treatment	Questionnaire	Uncertain long-term retention of knowledge
Daud, 202378	Cariology and endodontics	G. V. Black Class I cavity preparation	Questionnaire, Operation test	Uncertain long-term retention of knowledge, Incomplete equivalence to real-life clinical settings, resources-intensive nature
Liebermann, 2023113	Dental morphology	Tooth morphology	Questionnaire	Non-substitutive supplement to traditional training
Hamama, 202441	Cariology and endodontics	Caries removal	Questionnaire	Incomplete equivalence to real-life clinical settings, resource-intensive nature, non-substitutive supplement to traditional training
Fu, 202413	Periodontology	Ultrasonic periodontal scaling	Operation test, Questionnaire	Non-substitutive supplement to traditional training
Duan, 202427	Cariology and endodontics	Access cavity preparation	Operation test, Questionnaire	Uncertain long-term retention of knowledge
Daud, 202417	Cariology and endodontics	G. V. Black Class I cavity preparation	Focus group interview	Incomplete equivalence to real-life clinical settings
Buajeeb, 202477	Oral medicine	Management of oromucosal lesions	Theory test, Questionnaire	Incomplete equivalence to real-life clinical settings
Li, 202442	Cariology and endodontics	Rubber dam isolation technique	Operation test, Questionnaire	Uncertain long-term retention of knowledge
Wang, 202458	Periodontology	Periodontal probing	Operation test, Questionnaire	Uncertain long-term retention of knowledge, Incomplete equivalence to real-life clinical settings
Wei, 202443	Cariology and endodontics	Access and coronal cavity preparation	Operation test, Questionnaire	Non-substitutive supplement to traditional training, with uncertain long-term retention of knowledge
Wu, 202459	Periodontology	Scaling and root planing	Questionnaire, Operation test	Uncertain long-term retention of knowledge
Yao, 202451	Prosthodontics and implantology	Clinical reception	Operation test	Incomplete equivalence to real-life clinical settings
Nassief, 202435	Cariology and endodontics	Hand skills training	Focus group interview	Uncertain long-term retention of knowledge
Peanchitlertkajorn, 202471	Other	Dental practice management	Theory test, Questionnaire	Non-substitutive supplement to traditional training
Liu, 202546	Cariology and endodontics	Apicoectomy	Theory test, Operation test, Questionnaire	Non-substitutive supplement to traditional training
Hu, 202547	Cariology and endodontics	Regenerative endodontics	Theory test, Operation test, Questionnaire	Non-substitutive supplement to traditional training
Lv, 202540	Cariology and endodontics	G. V. Black Class I, II and V cavity preparations	Questionnaire, Operation test	Incomplete equivalence to real-life clinical settings, resources-intensive nature
Tang, 202574	Prosthodontics and implantology	Metal-ceramic crown preparation	Questionnaire, Operation test	Incomplete equivalence to real-life clinical settings
Cheng, 202560	Periodontology	Subgingival scaling and root planing	Questionnaire, Operation test	Incomplete equivalence to real-life clinical settings
Huang, 202545	Cariology and endodontics	Root canal filling	Questionnaire, Operation test, Theory test	Incomplete equivalence to real-life clinical settings
Prinz, 202530	Cariology and endodontics	Endodontic diagnostics and emergency treatment	Questionnaire	Incomplete equivalence to real-life clinical settings
Guo, 202563	Oral and maxillofacial surgery	Temporomandibular joint surgery	Questionnaire, Theory test	Incomplete equivalence to real-life clinical settings
Rehman, 202572	Other	General clinical learning, patient communication and diagnosis	Questionnaire	Uncertain long-term retention of knowledge
Chen, 202567	Orthodontics	Orthodontic treatment process	Questionnaire	Non-substitutive supplement to traditional training
Liu, 202564	Oral and maxillofacial surgery	Orthognathic surgery	Questionnaire, Theory test	Incomplete equivalence to real-life clinical settings, resources-intensive nature

In dental morphology education, VS supports tooth morphology learning through digital visualization tools. Mobile-based instruction was reported as non-inferior to physical models and offers practical advantages in portability and cost³¹. Augmented reality (AR) applications enable the overlay of 3D tooth models with touch-based manipulation³¹. Desktop tools support tooth identification aligned with the FDI tooth-numbering system and provide real-time feedback³². VR headsets enable immersive virtual laboratories for macroscopic anatomy³³, root canal observation³⁴, and anatomical landmark identification on panoramic radiographs¹². Stereoscopic visualization may further support spatial learning beyond conventional 2D instruction³⁵.

In cariology and endodontics, VR-based training enables cavity preparation using haptic-enabled handpieces and foot pedals^36–38. Virtual drilling and material-removal tasks are used to develop fine motor skills and hand–eye coordination^29,39,40. Clinical simulations address caries management⁴¹ and the complete endodontic workflow, including rubber dam isolation⁴², access cavity preparation^27,43, pulp chamber preparation⁴⁴, and obturation⁴⁵. Advanced techniques such as apicoectomy and pulp regeneration have also been incorporated^46,47. Generative AI-powered virtual patients (GPT-4) have been applied to teach differential diagnosis of pulpal versus periodontal pain³⁰.

In prosthodontics, VS supports theoretical instruction on occlusion⁴⁸, adhesive materials⁴⁹, removable partial denture design⁵⁰, and patient reception⁵¹. Haptic simulators enable training in crown, inlay, and veneer preparation^52,53. Integration with 3D printing extends hands-on practice⁵⁴. In oral implantology, virtual navigation and haptic feedback facilitate surgical planning⁵⁵ and support site localization, positioning, drilling, and osteotomy^14,55,56. VR headsets enable systematic instruction of the complete implant workflow⁵⁷.

In periodontology, instruction focuses on probing⁵⁸, ultrasonic scaling, and root planing^13,26,59,60. Haptic styluses guide standardized six-step protocols. Real-time evaluation assesses instrument angulation and force control, supporting precision training in periodontal therapy^59,60.

In oral and maxillofacial surgery, VS addresses complex anatomy and higher-risk procedures. Systems simulate clinical decision-making using authentic cases⁶¹. Reported applications include inferior alveolar nerve block⁶², temporomandibular joint management⁶³, and orthognathic surgery⁶⁴. Extraction simulators train core skills for third molar removal, including bone removal, sectioning, and elevation, with automated quantification of tissue resection^61,65.

In orthodontics, VS-based instruction covers diagnosis and treatment planning. Students use electronic models for occlusal analysis and tooth-size measurement⁶⁶ and learn complete treatment sequences for Angle Class II Division 1 malocclusion⁶⁷.

In oral medicine, virtual cases and simulated clinics reinforce diagnostic knowledge and clinical reasoning. Gamified scenarios enhance recognition of typical mucosal conditions⁶⁸. Virtual clinics simulate patient–clinician communication in safe, repeatable environments⁶⁹.

Other applications include pediatric dentistry, clinic management, and multidisciplinary training. One study used virtual scenarios involving an 8-year-old patient and a parent to develop pediatric communication skills⁷⁰. A gamified virtual clinic taught management skills such as infection control and spatial planning⁷¹. A virtual dental clinic (VDC) integrating multidisciplinary cases supports pre-clinical training and complex case learning⁷².

Overall, VS has already been broadly integrated into undergraduate teaching across dental subspecialties and basic disciplines. However, its technological maturity remains uneven. Fields such as orthodontics, oral mucosal diseases, pediatric dentistry, and some basic dental sciences areas still rely mainly on first-generation applications, which are largely centered on theoretical learning and doctor–patient communication through virtual case libraries and three-dimensional models, with limited practical operation modules^66,68,70. In contrast, more procedure-intensive disciplines, including endodontics, prosthodontics and implantology, periodontology, and oral and maxillofacial surgery, have more often adopted second-generation systems based on VR/AR integration, force feedback, and full-process procedural simulation^11,13,52,65. These inter-disciplinary differences appear to be mainly related to the complexity of the target skills, the degree of haptic or immersive support required, and the maturity of the available simulation systems⁷.

Assessment methods and outcomes

The effectiveness of VS-based dental education is appraised to ensure it meets the evolving needs of undergraduate trainees. Evaluations are typically organized into four complementary strands: (1) Knowledge acquisition and retention: Most trials employ pre- and post-training theoretical examinations, quizzes or MCQ batteries to quantify learning gains^26,36,67,73. (2) Practical competence: Operative competence is measured with expert-scored practical tests on synthetic resin teeth²⁷, pig mandibles⁴⁶, periodontal pathology models^13,59,60 or phantom-head crown preparations⁷⁴, frequently supplemented by simulator-generated metrics such as residual calculus rate^59,60, drilling accuracy verified by CBCT/fluoroscopy⁵⁶, and number of attempts or total preparation time⁶⁵. (3) Skill transfer and clinical readiness: transfer is assessed through laboratory transfer tests^26,56, standardized-patient Mini-CEX encounters⁵¹, and case-design exercises⁵⁰. (4) User experience and program satisfaction: usability is captured via the System Usability Scale³⁰, iterative questionnaires^33,38,57,64, online or face-to-face focus groups, and attitudinal surveys. Several implementations also apply Miller’s pyramid¹⁸ or formative feedback algorithms to guide reflective practice¹⁸.

Across studies, undergraduate competencies improved after VS training, which reinforced both theory and practice. Virtual dental models visualize teeth, root canals, occlusion, and edentulous or orthodontic cases, enabling students to explore anatomy interactively without entering a dissection laboratory and thereby improving comprehension and retention^32,34,66. Virtual treatment planning systems allowed undergraduates to repetitively design implants, removable partial dentures, or rubber-dam protocols under variable clinical conditions, thereby standardizing scheme formulation training^42,50,73. Virtual clinical scenario systems embedded VS within simulated patient encounters, offering case-based training that mirrors real settings, fosters communication skills, and sharpens clinical reasoning^30,69,75. VS-based case learning enhances knowledge acquisition, retention, and transfer to clinical contexts, ultimately improving diagnostic accuracy and clinical-reasoning instruction^30,75. Practical operation simulators offered a risk-free, repeatable, and real-time feedback environment in which students master procedures, instruments, and equipment before treating patients^36,43. Deployed across disciplines, these systems train hand–eye coordination, fine motor skills, access cavity preparation, endodontics, periodontics, extractions, and implant or prosthodontic procedures, ensuring thorough and standardized pre-clinical training^13,46,74.

VS provides repeatable, immersive and instant feedback training, which significantly increases undergraduate dental students’ confidence in pre-clinical skills^3,36,43. Most studies employed five-point Likert scales to evaluate satisfaction and educational impact; students consistently rated the approach as useful, informative and understanding-enhancing^17,57,74. In implant planning and dental anatomy modules, students overwhelmingly selected VR as their preferred future learning tool, indicating strong perceived efficacy^33,73. Platforms such as Simodont reinforce students’ sense of control over procedural details through multidimensional visualization and unlimited practice, thereby boosting confidence^36,76. Although some students noted that haptic fidelity remains inferior to conventional phantom heads and advocated for closer integration⁶, VS has already significantly enhanced self-efficacy in aseptic technique, access cavity preparation, periodontal therapy and other core skills, providing a solid psychological and technical bridge to the clinical setting^3,42,59 (Table 2).

Challenges

In pursuit of advanced undergraduate dental pre-clinical training, the application of VS is faced with multifaceted challenges, including uncertain long-term retention of knowledge^12,42,59,72, incomplete equivalence to actual clinical settings^25,51,63,77, heavy reliance on digital and instructional resources^29,48,50,78 and non-substitutive supplement to traditional training^11,13,47,71 (Table 2).

Uncertain sustained knowledge retention

As previously described, contemporary VS systems have been increasingly involved in core knowledge and skills across subdisciplines, facilitating students’ rapid, short-term acquisition of foundational knowledge. Most studies found that VS improved undergraduates’ initial knowledge acquisition through short-time performance assessment. However, uncertainties remain regarding its long-term educational effectiveness, particularly in terms of sustained knowledge retention^36,49,56,73. As novel teaching methods, VS generated potential novelty effects, which may help students to improve the instant performance in pre-clinical training⁷⁹. A study on digital learning methods has found that the mobile VR group demonstrated a drop in perceived usefulness and behavior intention after 3 months of use⁸⁰. This phenomenon may be attributed to the limited and isolated implementation of VS, which fails to integrate knowledge and skills into a coherent developmental system^72,81. Consequently, the fragmented learning pattern undermines learners’ deep cognitive engagement and raises concerns about sustained knowledge retention across the entire professional life cycle of dentists^27,54,81.

Incomplete replication to real-world clinical settings

One of the major challenges of VS lies in its incomplete replication of the complexities and unpredictability of real-world clinical settings^17,29,39,41. Real-world clinical situations involve a wide spectrum of disease presentations, anatomical variations, resource limitations, personalized patient communication, and multidisciplinary treatment (MDT) coordination, among numerous other dynamic factors^30,61,69. However, most existing VS systems operate within a closed ecosystem characterized by limited databases, restricted interoperability, and delayed updates in treatment guidelines and research advancements^1,81. Consequently, these systems predominantly rely on standardized, pre-set models designed for repetitive and deliberate practice^40,53,57,81. While such an ecosystem effectively enables students to refine procedural techniques in a controlled, risk-free environment that safeguards patient safety, it simultaneously raises concerns regarding the translatability of simulation-acquired skills to authentic clinical practice^25,37. The absence of continuously updated clinical data and diversified case repositories constrains students’ exposure to the variability intrinsic to real-world scenarios⁸². As a result, the current closed ecosystem of VS may impede the development of students’ adaptive problem-solving abilities, critical thinking, and competence in managing complex clinical challenges that evolve alongside contemporary dental practice^51,70,82. These limitations may be further shaped by differences in local technological ecosystems and the availability of mature simulation platforms.

Resources-intensive nature

The resource-intensive nature of VS is one of the most prominent challenges for wide and continuous application, which adds financial and educator burden. Implementing VS requires digital resources involving equipment, software, network and multimedia materials, which brings a large financial burden to establish VS programs. For instance, a single high-fidelity mannequin can cost upwards of $50,000, without the additional expenses for software, maintenance, and upgrades^29,33,83. Devices with inferior features and lower costs could damage the teaching outcomes with unfavorable interactive experience, including lower resolution visual performance, lagged feedback and poor availability^1,84. Limited by adaptations between teaching platforms and subdisciplines, schools have to build specialized platforms for each subspecialty, adding extra teaching costs^40,64. Moreover, studies have indicated that the VS program necessitates professional instructions from teachers to achieve optimal teaching goals, which demands ongoing development and support for teachers^85,86. Dongru Chen et al. proposed that teachers devoted a large amount of time and energy to developing the VR software by communicating with the engineers⁶⁷. Due to the resource-intensive nature, VS training called for the development of a cost-effective solution and highly compatible universal platforms, ensuring access for high-quality pre-clinical training^1,48,50. In addition, differences in institutional resources, infrastructure readiness, and platform accessibility may also contribute to uneven adoption and implementation across educational settings.

Non-substitutive supplement to traditional training

Numerous studies have concluded that VS could not completely replace traditional training but serves as a supplement^13,32,47,71. The main reason is that VS systems cannot fully replicate the clinical practice^32,41. Thus, how to facilitate effective VS training programs remains a great challenge in the development of high-quality pre-clinical training. Since global adoption of VS has been a growing tendency, it is a novel teaching paradigm to integrate simulation training with other training modalities, including problem-based learning, interdisciplinary training programs, multimedia materials and inverted classroom^49,50,69,87.

Discussion

With the increasing maturity of VS, its role in undergraduate dental education will evolve from an auxiliary teaching tool to a core infrastructure, reshaping curriculum design, teaching implementation, and competency assessment^1,7,19. Based on the current evidence synthesized in this review, a forward-looking agenda for the future incorporation of AI into dental VS is proposed^19,24,88. It outlines the development directions of dental AI–VS incorporation from the perspectives of technology⁷, application¹³, and assessment methods³⁶, aiming to provide guidance for educators and institutions in constructing a sustainable, scalable, and student-centered virtual learning ecosystem (Fig. 4). Although direct empirical evidence on AI-integrated VS remains limited at present, these perspectives may help identify promising directions for future research and educational innovation.

Fig. 4. The core development directions for the incorporation of AI into dental VS systems.

The overarching developmental objective of VS in undergraduate dental education is to move beyond conventional technological frameworks and application-level constraints by establishing a next-generation AI–VS incorporation paradigm. In this framework, AI technologies function as the central enabler of end-to-end innovation across five interrelated dimensions, including intelligent virtual patients and AI tutors, enhancing virtual simulation fidelity, intelligent teaching content, intelligent supplementary of disciplinary background and personalized adaptive learning.

From the technological perspective, the incorporation of AI into dental VS may reshape core system functions, including intelligent virtual patients, AI tutors, haptic realism, and scene fidelity. Although the current second-generation VS systems have achieved remarkable results in clinical operation simulation and real-time assessment, their core framework is still constrained by traditional simulation teaching, making it difficult to meet the demand for personalized competency development in undergraduate stomatology education^35,36,39. At present, the application of cutting-edge technologies such as generative AI in VS systems is still in the exploration stage, with a lack of mature solutions and models. This also constitutes a key direction for the research and development of the third-generation AI–VS incorporation^19,30.

On one hand of technology, future development may promote the innovative use of intelligent virtual patients and AI tutors. Driven by generative AI and large language models (LLMs), future virtual patients may move beyond the static dialog limitations of traditional systems and simulate more diverse doctor–patient communication scenarios across different dental subspecialties^30,89,90. This enables learners to improve the accuracy of medical history collection and empathy in doctor–patient communication through repeated practice⁵¹. Meanwhile, AI tutors support multi-modal interactions such as voice questioning and operation analysis, providing real-time expert-level feedback on key links, including instrument selection and treatment plan design^22,89. This effectively addresses the shortcomings of traditional teaching, such as limited faculty resources and inadequate supervision coverage, and builds a bridge connecting simulation training with real clinical diagnosis and treatment^65,75.

On the other hand, the fidelity of simulators can be further enhanced. In terms of haptic feedback optimization, AI technology breaks away from the limitation of traditional systems that rely on simplified physical equations^84,91,92. By deep learning massive biomechanical data of real clinical operations, it constructs mechanical models to accurately replicate the differences in hardness, elasticity, and operational resistance of oral hard and soft tissues, thereby improving the clinical adaptability of practical skills^11,29,36,52. In terms of scene visualization, in-depth integration of AI and AR technology seamlessly overlays virtual treatment plans and anatomical landmark annotations onto real-time images or 3D-printed models^93–95. This significantly improves the clinical realism of the training environment and provides technical support for high-fidelity scene simulation⁵⁴.

From the application perspective, the incorporation of AI into dental VS should be adapted to the educational objectives, technological maturity, and practical training needs of different dental subspecialties. The findings of this review suggest that undergraduate dental VS has already achieved broad disciplinary coverage, but its technological maturity remains uneven across subspecialties^1,7. Building on this pattern, future AI incorporation should not follow a uniform pathway across dental disciplines¹⁹. Instead, differentiated upgrading strategies may be needed according to the educational objectives, technological readiness, and practical training demands of different subspecialties⁸⁹. Preliminary exploratory applications in endodontics also suggest that AI may first emerge in areas where simulation systems are already relatively mature and where structured diagnostic or procedural workflows are well established^11,23,30.

Future development may focus on implementing differentiated optimization and upgrading paths for dental subspecialties at different technological stages, promoting the iteration of VS teaching towards intelligence^19,89. For subspecialties still in the first-generation technology stage, the transformation should focus on generative AI and LLMs^30,90, combined with hardware such as lightweight VR devices and force feedback control components^15,91, to construct a virtual diagnosis and treatment system with dynamic disease progression, intelligent Q&A interaction, and diagnosis and treatment logic guidance. Simultaneously, practical training modules should be supplemented to make up for the shortcomings in skill development^15,16. For core operational subspecialties that have adopted second-generation technologies, the focus is on improving the quality and efficiency of operative training^13,28,36. Core data such as students’ operation force parameters, instrument movement paths, error types, and durations are accurately collected, and personalized competency profiles are constructed through deep learning algorithms to dynamically adapt to the training needs of students with different foundations^11,25,96.

Additionally, AI may help address the limitations of teaching content and efficiently construct high-quality teaching resources^19,89. On one hand, it integrates multi-modal data such as CBCT images and digital intraoral impressions to reconstruct high-fidelity operation carriers with complete diagnosis and treatment information⁹⁷. On the other hand, it focuses on the teaching gaps in cutting-edge disciplinary technologies, rare diseases, and complex multidisciplinary cases^98,99. Based on treatment guidelines and massive literature of each subspecialty, it constructs a complex case simulation library covering all dental subspecialties^68,77,99, and builds a multidisciplinary team VS platform with a dynamically updated professional knowledge base^30,89,90. This ensures the timeliness, authority, and clinical relevance of teaching content, guides undergraduates to establish systematic clinical thinking, and improves their ability to diagnose and treat complex cases^46,77.

From the assessment perspective, the incorporation of AI into dental VS may support more continuous, objective, and personalized competency-based evaluation. Most existing studies lack adequate longitudinal assessments¹⁰⁰. Additionally, significant differences exist in the evaluation tools adopted by different teaching programs, with a lack of unified standards and quantitative systems¹⁶. This limitation severely hinders cross-study benchmarking, result recognition, and systematic review of teaching effectiveness, failing to meet the demand for competency-oriented personalized education. Therefore, there is an urgent need to establish a systematic, multi-level, and dynamic evaluation framework based on AI, integrating real-time performance analysis, phased skill transfer assessment, and continuous user feedback to realize iterative optimization of teaching schemes and precise improvement of educational outcomes^1,11,25.

The core of AI empowerment lies in constructing a personalized adaptive evaluation mechanism, which addresses the subjectivity, lag, and homogenization of traditional evaluation methods²². Relying on multi-dimensional sensors in VS systems, AI captures real-time core data of learners, including instrument movement trajectories, force application characteristics, operation accuracy, duration, and error frequency^96,101,102. After feature extraction via deep learning algorithms⁹⁶, the system objectively quantifies skill deficiencies through visualization tools such as three-dimensional heat maps, trajectory playback, and error annotation¹¹. It then generates targeted feedback and stepwise training recommendations, overcoming the subjective and ambiguous limitations of experience-based evaluation^25,39. Based on full-cycle evaluation data, the system establishes a dynamically updated personalized competency profile for each learner, accurately identifying their proficiency levels in knowledge mastery, practical operation, critical thinking, and other dimensions^22,88. Subsequently, it automatically adjusts training content: strengthening basic operational training and detailed guidance for learners with weak foundations, and expanding complex case drills for those who meet competency standards²², thereby supporting active and repeatable self-directed learning^3,103. This personalized adaptive evaluation mechanism aligns preclinical skill training more closely with individual competency development needs, facilitating the achievement of competency-oriented personalized training goals for stomatology undergraduates^22,88.

Despite these potential benefits, the incorporation of AI into dental VS also raises several important concerns that must be addressed before large-scale implementation to ensure clinical safety and pedagogical integrity^19,24. Algorithmic bias remains a primary concern. If AI models are trained on non-representative datasets, they may produce skewed diagnostic suggestions, potentially compromising health equity^104,105. Privacy and data security are equally paramount, as the integration of student performance and clinical imaging into AI ecosystems requires robust encryption to prevent sensitive data leaks^89,106. Ethically, the “black-box” nature of some AI tools poses a challenge to the development of critical thinking, necessitating a “human-in-the-loop” approach where AI acts as a co-pilot rather than a substitute for faculty mentorship^107,108. Finally, the field suffers from a lack of large-scale, cross-institutional validation. Moving forward, AI tools must undergo rigorous clinical trials to prove that their simulated environments translate effectively into superior real-world operative competence^19,20.

The transition toward AI–VS incorporated dental education presents significant practical challenges that institutions must address to ensure a realistic and sustainable implementation^19,24. Financially, the primary barrier is the high capital expenditure required for advanced hardware^29,33. For instance, AI-integrated implant navigation systems necessitate substantial investment in optical tracking sensors and haptic feedback devices^14,84. While AI can eventually reduce costs by automating the creation of patient-specific models from existing CT libraries, the initial “entry price” remains a major hurdle for many dental schools^29,40. Technologically, institutional infrastructure must evolve toward cloud-native architectures capable of high-speed, real-time data synchronization^84,91. In applications like implant navigation, any latency between the physical tracking sensors and the AI evaluator can compromise the objective assessment of drilling precision, entry-point deviation, and depth control^14,55,56. Without robust network support and seamless integration of clinical data (e.g., DICOM and STL files), the fidelity of the simulation is significantly diminished^57,94. Pedagogically, faculty training represents the most critical pivot point^41,75. The educator’s role must transition from traditional clinical observation to big data-driven mentorship^20,89. This requires instructors to develop the ability to interpret complex AI-generated outputs, such as surgical “precision heatmaps” and trajectory logs, to provide targeted feedback^11,25. This shift necessitates a profound change in instructional methodology and extensive institutional support for retraining^15,67. Ultimately, a realistic transition requires a phased approach. The approach starts with modular AI tools before moving toward fully integrated ecosystems, while it simultaneously bridges the socio-technical gap in faculty expertise^19,20,89.

Several limitations of the current evidence base should be considered when interpreting the findings of this review. A substantial proportion of studies originated from a single country, and some were concentrated around a limited number of specific commercial or locally developed platforms^1,91. Although this pattern reflected the distribution of currently available eligible evidence rather than preferential selection, it may limit the generalizability of the findings to other educational settings^7,16. Therefore, the findings should be interpreted as reflecting broader trends and shared challenges in undergraduate dental VS rather than favoring any particular country or platform. More multi-center studies from diverse educational and technological contexts are needed to strengthen global applicability^7,16. Despite these limitations, the current evidence still provides a useful basis for considering future VS development and AI incorporation^8,19.

Methods

Study design

This scoping review was conducted in accordance with the PRISMA-ScR guidelines¹⁰⁹. A descriptive approach based on Arksey and O’Malley’s five-stage framework was employed¹¹⁰: (1) identifying the research question; (2) identifying relevant studies; (3) selecting studies; (4) charting the data; and (5) collating, summarizing, and reporting the results. In addition, to provide methodological context for interpreting the evidence base, a brief critical appraisal of the included studies was conducted.

Identifying the relevant studies

The search strategy was developed around four core concepts: VS-related technologies, educational context, undergraduate training, and dental discipline. The core search terms comprised the following keywords: (“Augmented Reality” OR “Virtual Reality” OR haptic OR simulation) AND (educat* OR curriculum OR course OR program OR “educational methodology” OR “teaching method*”) AND (undergraduate OR “pre-postgraduate”) AND (dentistry OR oral OR dental OR dentist* OR odontol* OR stomatol*).

A wide-open systematic search of the literature was conducted in PubMed, ERIC (Institute of Education Sciences [IES], Washington, DC), Scopus™ (Elsevier B.V., Amsterdam, The Netherlands), and Web of Science (Clarivate Analytics, Philadelphia, PA) for records published up to October 2025. General controlled vocabulary (MeSH terms) and keywords were used where appropriate, and the search syntax was adapted to the indexing system and search interface of each database. Although the search itself was not limited by language, only full-text articles published in English were included during the study selection process.

Eligibility criteria and study selection

The inclusion criteria were as follows: (1) original research evaluating the educational use, implementation, or effectiveness of VS in dental or stomatology education; (2) participants were undergraduate dental or stomatology students; and (3) eligible study designs were not restricted and included a range of study types, such as randomized controlled studies, non-randomized comparative studies, descriptive studies, and relevant qualitative studies.

Studies were excluded if they: (1) did not examine VS as a core educational intervention or application; (2) did not focus on undergraduate dental or stomatology education; (3) were not original research articles, such as reviews, editorials, commentaries, conference abstracts, or study protocols; or (4) were not published in English.

Three reviewers participated in the two-stage screening process. Before formal screening, the reviewers aligned the eligibility criteria and screening procedures through discussion to ensure consistency in study selection. In the first stage, titles and abstracts were independently screened by two reviewers according to the predefined inclusion and exclusion criteria. In the second stage, all full-text articles that passed the initial screening were independently reviewed by two of the three reviewers. The reviewers held regular meetings to clarify the screening criteria and collectively resolve disagreements throughout the selection process. The final set of included studies was subsequently confirmed by the author team. Additionally, the reference lists of selected papers were manually analyzed to identify any relevant manuscripts that may have been missed during the electronic database search or other grey literature sources (Fig. 1).

Critical appraisal of included studies

A brief critical appraisal of the included studies was conducted using the Mixed Methods Appraisal Tool (MMAT), version 2018¹¹¹, to provide methodological context for interpreting the evidence base. Given the scoping nature of this review and the heterogeneity of eligible study designs, the appraisal was intended to support interpretation of the findings rather than to determine study eligibility.

Charting the data

The authors extracted information from the included articles, including basic study information, subject theme, VS system, training item, assessment methods, outcomes, and challenges. The data were extracted by three reviewers, and any inconsistencies were resolved through discussion.

Author contributions

S.C. contributed to conception, design, drafted, and critically revised the manuscript. L.Q. and X.D. contributed to conception and drafted the manuscript. P.Y., X.L., S.H. and J.X. (Jiankun Xu) contributed to conception and figure design. J.X. (Jieyun Xu) and Z.C. contributed to conception and critically revised the manuscript. All authors gave final approval and agreed to be accountable for all aspects of the work.

Data availability

No datasets were generated or analyzed during the current study.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41746-026-02806-z.