Abstract
Introduction
Artificial Intelligence (AI) and Machine Learning (ML) are increasingly being incorporated into dentistry to enhance diagnostics, treatment planning, and clinical outcomes. However, their translation into routine practice is contingent upon regulatory approval to ensure safety and effectiveness. In the United States, the Food and Drug Administration (FDA) typically evaluates moderate-risk AI/ML-based dental devices under the 510(k) Premarket Notification pathway. Understanding the approval trends and characteristics of these devices is crucial for assessing the clinical integration of AI in dentistry.
Materials and methods
A manual search of the FDA’s 510(k) Premarket Notification Database was conducted in December 2024 using the terms “AI/ML devices” and “dentistry.” Devices were screened for relevance, and data were extracted regarding applicant name, device type, predicate device, year of approval, AI algorithm employed, clinical indication, country of origin, and status of real-world deployment.
Results
Fifty-two AI/ML dental devices were identified. The year 2022 saw the highest number of approvals (n = 8, 15.38%). Nearly half (n = 25, 48%) of the devices were indicated for oral radiology, followed by applications in implantology and regenerative procedures (n = 18, 32%). Overjet, Inc. received four approvals for diagnostic tools targeting caries, calculus, and charting. Ewoosoft Co., Ltd. accounted for six approvals, primarily versions of its Ez3D-i imaging platform. Most devices originated from the United States (n = 24, 46.15%), followed by South Korea (n = 5, 9.61%) and Canada (n = 3, 5.76%). Notably, 60% of devices did not disclose the type of AI/ML algorithm used (n = 31), and 50% lacked public documentation on clinical deployment (n = 26).
Conclusion
Most FDA-cleared AI/ML dental devices in the 510(k) database are imaging-based diagnostics; greater transparency, real-world validation, and attention to equity are needed for safe adoption.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12903-025-07302-6.
Keywords: Artificial intelligence, Machine learning, Dentistry, FDA cleared devices, 510(k) database, Regulatory approval
Introduction
Artificial intelligence (AI) is increasingly being explored for its vast potential in the field of dentistry, offering new opportunities to improve diagnostics, treatment planning, and overall patient care [1]. Until recently, many concepts related to potential applications of AI were mere ideas. However, those applications are now demonstrating high prospect as workable solutions for rapid and accurate disease diagnosis, treatment planning, and patient management [2]. Ideas for the practical application of AI in dentistry are being put forth exponentially, however, their integration into practice necessitates rigorous regulatory evaluation to ensure safety, efficacy, and reliability.
In the United States, the Food and Drug Administration (FDA) oversees this process [3]. The oversight of FDA towards medical/dental devices is a risk-based stratification (Class I-III) [4]. Devices that pose the highest risks to patients have the most rigorous requirements to demonstrate safety and effectiveness before they can be legally marketed in the US. Such devices are known as Class III devices and for them a Premarket Approval (PMA) is generally required. On the other hand, Class II devices are moderate risk, which require a 510(k) Premarket Notification, where manufacturers must demonstrate that their device is substantially equivalent to an already legally marketed device (predicate). In contrast, Class I devices pose a minimal risk and are generally exempted from premarket review.
The FDA maintains multiple publicly accessible repositories to store information about the devices it approves. These include the 510(k) Premarket Notification Database for Class II devices, the PMA Database for Class III devices, and the De Novo Classification Database for novel low-to-moderate risk devices. Each repository provides detailed information about the approved or cleared devices, including their intended use, safety evaluations, and regulatory pathways.
While PMA devices often represent high-risk innovations, the majority of AI/ML applications in dentistry are currently classified as Class II (510 (k)) due to their moderate risk profile [5]. In the context of dentistry, these devices range from AI-assisted diagnostic tools that enhance the detection of pathologies, such as caries or peri-implant diseases, to machine learning algorithms that optimize treatment planning and predict patient outcomes. Therefore, this study deliberately focuses on 510(k) database rather than all other FDA resources [6].
Analyzing FDA-cleared AI/ML dental devices is significant as it offers valuable insight into the current state of technological integration of AI within dentistry [7]. The analysis will help identify the most approved types of AI/ML applications and highlight trends in their design and functionality. Moreover, evaluating the regulatory pathways these devices follow, particularly the 510(k) process, provides a clearer picture of the safety and efficacy standards that AI/ML technologies must meet. This can inform developers about the key considerations for regulatory compliance, potentially streamlining the approval process for future innovations. Lastly, the clinical impact and utility of these devices can be assessed by examining their intended use, reported outcomes and any limitations identified during the approval process.
Therefore, this study aims to analyze FDA-cleared class II dental devices listed in the 510(k) database, focusing on AI/ML technologies, to identify the current trends and innovation in the field. Furthermore, the analysis aims to map the geographical and clinical distribution of these technologies, particularly in underserved regions or specialties.
Materials and methods
Search strategy
The publicly available 510(k) Premarket Notification Database of FDA was searched by a single author (NN) on 15-December-2024 (UTC + 5, Islamabad, Pakistan) to retrieve results for AI/ML devices in dentistry (https://510k.innolitics.com).
The keywords “AI/ML devices” and “dentistry” were used to retrieve relevant results. The obtained results were manually screened to remove any irrelevant records by two authors (NN and SA). Any disagreement between the two was sought through discussion with the third author (FU) who then also rechecked the extracted information.
As the search strategy relied on keyword-based retrieval, there is a possibility that some relevant information not explicitly labeled with these terms may have been missed.
Data extraction
The following characteristics were recorded on a customized proforma by two authors (NN and SA):
Applicant
Device name
Predicate device
K-number
Product code
Year of review
AI/ML description (type of algorithm used)
Indication for use
Clinical application/utility
Geographic distribution
Real world deployment
To determine whether a device was categorized as “deployed” or “unknown,” supplementary sources such as company websites, press releases, product pages, professional announcements, and industry reports were systematically reviewed (Supplementary file).
Unit of analysis
The unit of analysis was each unique 510(k) submission (K-number). Iterative updates to the same platform were considered separate entities, as each was assigned a distinct K-number by the FDA.
Results
The initial search yielded a total of 61 records. After removing irrelevant records, 52 FDA cleared dental AI/ML devices were analyzed in the current study (Fig. 1).
Fig. 1.
Simple selection flow diagram representing search strategy
The individual characteristics of the included devices are presented in Supplementary Table 1.
The year wise trend of approval for AI/ML devices in dentistry is depicted in Fig. 2, highlighting fluctuations in approval rates over the years. The number of FDA-cleared devices varied significantly over the years, with no consistent linear trend. The highest number of approvals occurred in 2022 with 8 cleared devices (15.38%), followed closely by 2023 and 2018, each with 7 clearances (13.46%). Conversely, 2015 saw no approvals, making it the lowest point in the dataset.
Fig. 2.
FDA-cleared AI/ML devices in dentistry over the years with dotted line serving as a trendline
The applicants with the most cleared AI/ML dental devices over the years are presented in Fig. 3. Ewoosoft Co., Ltd led the field with 6 clearances (11.53%), all related to its Ez3D-i maxillofacial imaging software, indicating a sustained regulatory presence through iterative enhancements and reapprovals of a core diagnostic platform. Overjet, Inc. followed with 4 distinct device clearances (7.69%), including Overjet Caries Assist, Overjet Calculus Assist, Overjet Dental Assist, and Overjet Charting Assist, each designed to support various AI-enabled diagnostic and documentation tasks in dental practice. Additionally, a group of companies including MATERIALISE DENTAL NV, 3D Industrial Imaging Co., Ltd., Nobel Biocare AB, and VideaHealth, Inc., each achieved 3 approvals (5.76%), followed by Software Nemotec S.L., Denti.AI Technology, Inc., and Straumann USA, LLC, among others, with 2 approvals each (3.84%), reflecting their roles as emerging contributors to AI-driven dental innovation.
Fig. 3.
Applicants with most cleared FDA AI/ML devices in dentistry. The counts are presented based on distinct FDA 510(k) numbers, with devices grouped by applicant; thus, multiple products cleared under separate K-numbers but submitted by the same applicant are represented within a single category
The distribution of FDA-cleared devices across different geographic locations as shown in Fig. 4 highlighted the dominance of the United States (US), accounting for the highest number of approvals (n = 24, 46.15%). Among other regions, the Republic of Korea (South Korea) followed US with 5 cleared devices (9.61%) which was then followed by Canada (n = 3, 5.76%), Israel and China with 2 devices each (3.84%). Several other countries, including Spain, Sweden, Belgium, Taiwan, France, and Finland had a lower contribution to the overall count with only 1 clearance each (1.9%).
Fig. 4.
Geographical trend in the FDA clearance of AI/ML dental devices, with country classified according to the applicant’s location as listed in the 510(k) database
The trend in specialization and adoption of AI/ML devices in dentistry over time, categorized by specialty is presented in Fig. 5. The timeline from 2011 to 2024 showed progressive integration of AI in different dental fields, with a noticeable increase in approvals in recent years. The analysis revealed that Oral Radiology had the highest number of FDA-approvals (n = 25, 48%), followed by 32% in the field of Implantology and tissue regeneration (n = 18). Other specialties, such as Orthodontics (n = 4, 7%), Periodontology (n = 3, 5%), Oral and Maxillofacial surgery (n = 2, 4%), and oral diagnostic imaging (n = 2, 4%) had fewer approvals indicating that the adoption of AI in these fields is still developing. A small crosstab (year x specialty) is presented in Supplementary Table 2.
Fig. 5.
Specialty-wise distribution of AI/ML devices in dentistry over time. *3DIEMME RealGUIDE: Supports both implant and oral & maxillofacial surgical planning. **Planmeca Romexis: Works as both diagnostic imaging and implant planning software
Among the 52 FDA-cleared devices analyzed, majority did not specify the type of AI/ML algorithm used in their 510(k) submissions (n = 31, 60%). The remaining devices employed machine learning and deep learning algorithms (computer vision techniques) such as deep neural networks, convolutional neural networks (CNN), and the U-Net architecture. Additionally, some devices utilized advanced image processing algorithms, including supervised machine learning (decision trees, support vector machines), while others integrated a combination of these approaches.
Moreover, approximately half of the devices (n = 26, 50%) had no specific deployment information available. This means that there is no publicly available data confirming their real-world clinical use at the time of writing of this paper. In contrast, the remaining devices were identified as deployed or available for clinical use, with some being actively used in treatment planning, diagnostics, or AI-driven decision support.
Discussion
The integration of AI/ML in dentistry is rapidly expanding, yet its adoption hinges on regulatory approval to ensure safety and efficacy [8]. In the US, FDA plays a critical role in this process, with most AI-driven dental devices classified as moderate-risk (Class II) and cleared through the 510(k) Premarket Notification pathway [4].
Our analysis of the 510(k) database revealed increasing adoption of AI/ML in dentistry, with a significant number of cleared devices focusing on diagnostic imaging and treatment planning. The predominance of diagnostic tools such as AI-assisted radiographic interpretation for caries detection and periodontal disease assessment highlights the growing reliance on computational models to augment clinical decision-making. These technologies aim to reduce diagnostic errors, enhance consistency across practitioners, and improve early disease detection, ultimately contributing to better patient outcomes [9].
Beyond diagnostics, an emerging trend was noted in AI/ML applications in treatment planning, assisting in orthodontic simulations, prosthodontic restorations, and implantology. Such innovations enable personalized treatment approaches, leveraging large datasets to predict patient-specific responses and optimize procedural outcomes. However, despite these advancements, the overall presence of AI/ML in the 510(k) database remained relatively modest compared to other medical fields, suggesting that dental AI is still in its early stages of regulatory adoption [10].
The year wise trend of approval for AI/ML devices in dentistry showed interesting patterns in recent years, with 2022 showing a peak, followed by a decline in 2023 and 2024. These fluctuations could be influenced by factors such as regulatory changes, technological advancements, market demand, or external disruptions like the COVID-19 pandemic, which may have impacted approval processes in 2020 and 2021.
The predominance of Ewoosoft Co., Ltd and Overjet, Inc in FDA device approvals highlights two distinct trajectories within AI-driven dental technology. Overjet’s portfolio comprises multiple distinct devices, each tailored to specific clinical tasks such as caries detection, calculus identification, diagnostic decision support, and automated charting, reflecting a diversified and application-specific innovation strategy. In contrast, Ewoosoft’s repeated approvals primarily pertain to its maxillofacial imaging software, suggesting a pattern of iterative updates or modifications to an existing platform that necessitated separate regulatory submissions over time. This could be attributed to evolving software features, adaptation to new hardware, or periodic reclassification under updated FDA guidelines. Together, these examples illustrate differing pathways to regulatory success, one through diversification and the other through sustained enhancement of a core product.
An important aspect of this study was assessing the geographical distribution of devices as recorded in the FDA 510(k) database. Our findings revealed a concentration of clearances among applicants based in the US. This pattern may reflect disparities in how AI-driven dental innovations are brought to the US market, with resource-rich companies and institutions more frequently pursuing FDA clearance. It is important to note, however, that these data reflect only the FDA regulatory pathway and may not capture devices approved through regional or local regulatory bodies outside the US [11]. The financial and procedural requirements associated with FDA clearance may also contribute to the observed concentration of applicants from higher-resource settings.
From a clinical perspective, the applications of AI/ML were heavily skewed toward imaging-based diagnostics. While this is expected given the visual nature of dental assessments, other areas such as AI-assisted prosthetic design, endodontic decision-making process, and predictive analytics for oral disease progression remain under-represented. This highlights potential areas for further innovation, where AI could expand its role to include comprehensive patient management solutions.
While some manufacturers provided generic details on deep learning, convolutional neural networks (CNNs), or machine learning-based classifiers, a significant portion of devices in the FDA 510(k) submissions lacked transparency in their algorithmic approach. The omission of this information limits the ability to assess the specific computational methodologies utilized and raises concerns regarding standardization in reporting AI methodologies, highlighting the need for greater clarity in regulatory submissions.
The absence of real-world deployment data for nearly half of the devices analyzed could indicate either a lack of transparency in reporting or that these devices remain in the preclinical validation phase, undergoing software validation and benchtop testing. While such validation is essential, it does not fully capture real-world complexities. Moreover, although FDA clearance indicates regulatory approval for clinical use, this does not necessarily translate into active deployment in dental practices. Some devices may require additional clinical validation, while others encounter barriers related to clinician trust, cost, and integration into existing workflows [12, 13]. Concerns about interoperability with practice management systems, training requirements for end-users, and the economic feasibility of implementation may delay or limit adoption. Furthermore, the absence of deployment data underscores the importance of post-market evidence, as ongoing surveillance is needed to verify long-term performance, patient safety, and sustained clinical benefit.
In this study, Oral Radiology and Oral Diagnostic Imaging were considered as distinct categories. Oral Radiology referred to devices focused on acquisition and interpretation of conventional radiographic modalities (periapical, bitewing, panoramic, and CBCT). In contrast, Oral Diagnostic Imaging was reserved for software tools that primarily supported image enhancement, processing, or diagnostic decision assistance, where FDA submissions did not explicitly classify the device under radiology. This distinction was maintained to reflect the categorizations provided in the original 510(k) submissions and to capture potential differences in intended clinical use.
While AI/ML offers significant advancements in dentistry, its regulatory approval remains complex. The 510(k) pathway, though facilitating market entry, may not be well-suited for AI/ML models that continuously evolve through iterative learning [6]. Unlike static medical devices, AI models require frequent updates and refinements based on newly acquired data, necessitating additional regulatory submissions to maintain compliance [14]. This process, although crucial for patient safety, can delay the adoption of improvements thereby slowing the rate of integration of AI-driven dental technologies into clinical practice.
Moreover, the lack of explainability in AI decision-making remains a critical concern, yet it is not a formal requirement for FDA Class II devices cleared through the 510(k) pathway [15]. Many AI models function as “black boxes,” where the reasoning behind specific predictions is not easily interpretable, which may hinder clinician trust and adoption [16]. Despite this, current FDA regulations primarily focus on performance validation rather than model transparency. Given the increasing reliance on AI in clinical decision-making, there is a strong need to incorporate explainability requirements into the 510(k) framework to ensure both regulatory oversight and clinical utility.
Recent policy efforts reflect growing recognition of these challenges. The FDA has outlined a Predetermined Change Control Plan (PCCP) framework to manage modifications in AI/ML-enabled devices, alongside the multi-stakeholder development of Good Machine Learning Practice (GMLP) principles to guide trustworthy model design and validation [17]. Moreover, the FDA has emphasized the importance of real-world evidence in assessing device performance post-clearance, signaling a shift toward adaptive regulatory oversight that balances innovation with patient safety. For dental AI, aligning with these evolving policies will be critical for fostering clinician confidence, ensuring equitable deployment, and accelerating responsible adoption.
The outcomes of this study are substantial but remain explicitly confined to the FDA 510(k) database. First, it provides an overview of FDA-cleared AI/ML dental devices, highlighting their contributions to improving diagnostic precision, patient outcomes, and clinical workflows. Moreover, it identifies emerging trends, geographical disparities, and gaps in the market, offering valuable data for researchers, developers, and policymakers. Finally, this analysis contributes to the broader dental and regulatory communities by promoting an evidence-based understanding of AI/ML integration in dentistry. These insights can inform future research, guide innovation, and support the ethical and effective adoption of AI technologies, ultimately enhancing patient care and health equity globally.
However, there are some limitations of this study. The scope was explicitly restricted to Class II dental devices cleared through the FDA 510(k) pathway, and therefore devices regulated under other FDA mechanisms or classified differently were not included. Likewise, regulatory approvals from other jurisdictions, such as the European CE mark or regional authorities, were outside the scope of this analysis. Finally, this study provides a time-limited snapshot of FDA-cleared devices as of the stated access date; subsequent software updates, refinements, or new submissions are not captured and may alter the performance, availability, or clinical integration of these technologies.
Conclusion
This study provides a detailed analysis of FDA-cleared AI/ML dental devices, offering valuable insights into their regulatory approval, clinical applications, and market distribution. The findings highlight the growing role of AI in diagnostics and treatment planning, yet also reveal gaps in transparency, real-world validation, and geographical distribution. It is important to note that this work represents an overview of FDA-cleared devices in the 510(k) database at the time of analysis, reflecting both progress achieved and challenges that remain.
Supplementary Information
Acknowledgements
None.
Abbreviations
- AI
Artificial Intelligence
- ML
Machine Learning
- FDA
Food and Drug Administration
- CNN
Convolutional Neural Network
- PMA
Premarket Approval
- PCCP
Predetermined Change Control Plan
- GMLP
Good Machine Learning Practice
Authors’ contributions
NN contributed to conceptualization, methodology, data curation, analysis, interpretation, writing and reviewing the main manuscript file; SA contributed to data curation, analysis, interpretation and reviewing the final draft; FU contributed to conceptualization, reviewing the final draft and supervision of the project.
Funding
None.
Data availability
The datasets generated or analyzed during this study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Mayta-Tovalino F, Munive-Degregori A, Luza S, Cárdenas-Mariño FC, Guerrero ME, Barja-Ore J. Applications and perspectives of artificial intelligence, machine learning and dentronics in dentistry: a literature review. J Int Soc Prev Community Dent. 2023;13(1):1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ding H, Wu J, Zhao W, Matinlinna JP, Burrow MF, Tsoi JK. Artificial intelligence in dentistry—a review. Frontiers in Dental Medicine. 2023;4:1085251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Available from. https://www.fda.gov. Accessed 25 01 2025.
- 4.Available from: https://www.fda.gov/medical-devices/510k-clearances/medical-device-safety-and-510k-clearance-process#:~:text=The%20510(k)%20clearance%20process%20involves%20a%20comprehensive%20review%20of,is%2C%20a%20predicate%20device). Accessed 28 01 2025.
- 5.Muralidharan V, Adewale BA, Huang CJ, Nta MT, Ademiju PO, Pathmarajah P, et al. A scoping review of reporting gaps in FDA-approved AI medical devices. NPJ Digit Med. 2024;7(1):273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Available from: https://510k.innolitics.com/yshrestha/510-k-Search-1109907e6a274ec5bd5d8388a570bf18. Accessed 30 01 2025.
- 7.Joshi G, Jain A, Araveeti SR, Adhikari S, Garg H, Bhandari M. Fda-approved artificial intelligence and machine learning (AI/ML)-enabled medical devices: an updated landscape. Electronics. 2024;13(3):498. [Google Scholar]
- 8.Dubin JR, Ibad H, Cil A, Murray M. The FDA and ensuring safety and effectiveness of devices, biologics, and technology. J Am Acad Orthop Surg. 2022;30(14):658–67. [DOI] [PubMed] [Google Scholar]
- 9.Kabir T, Lee CT, Chen L, Jiang X, Shams S. A comprehensive artificial intelligence framework for dental diagnosis and charting. BMC Oral Health. 2022;22(1):480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ailia MJ, Thakur N, Abdul-Ghafar J, Jung CK, Yim K, Chong Y. Current trend of artificial intelligence patents in digital pathology: A systematic evaluation of the patent landscape. Cancers (Basel). 2022;14(10):2400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Fleming TR, Demets DL, McShane LM. Discussion: the role, position, and function of the FDA-the past, present, and future. Biostatistics. 2017;18(3):417–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mennella C, Maniscalco U, De Pietro G, Esposito M. Ethical and regulatory challenges of AI technologies in healthcare: a narrative review. Heliyon. 2024;10(4):e26297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lal A, Nooruddin A, Umer F. Concerns regarding deployment of AI-based applications in dentistry – a review. BDJ Open. 2025;11:27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Aquino YSJ, Rogers WA, Jacobson SLS, Richards B, Houssami N, Woode ME, et al. Defining change: exploring expert views about the regulatory challenges in adaptive artificial intelligence for healthcare. Health Policy Technol. 2024;13(3):100892. [Google Scholar]
- 15.Samaranayake L, Tuygunov N, Schwendicke F, Osathanon T, Khurshid Z, Boymuradov SA, et al. The transformative role of artificial intelligence in dentistry: a comprehensive overview. Part 1: Fundamentals of AI, and its contemporary applications in dentistry. Int Dent J. 2025. 10.1016/j.identj.2025.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Edith Chan AW. The shape of artificial intelligence: just a black box? Chem Biol Drug Des. 2020;96(3):882–5. [DOI] [PubMed] [Google Scholar]
- 17.Umer F, Khan M. A call to action: concerns related to artificial intelligence. Oral Surg Oral Med Oral Pathol Oral Radiol. 2021;132(2):255. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated or analyzed during this study are available from the corresponding author upon reasonable request.