YouTube™ Contents on Artificial Intelligence in Endodontics: Quality and Reliability

1. Introduction

The digital age has transformed the production, dissemination, and consumption of information, with YouTube™ emerging as one of the most widely used platforms for online video sharing.^1,3 It enables users to upload, view, and share diverse content, providing rapid access to information worldwide.^2,⁴ However, the absence of strict regulation, formal verification, and peer-review processes raises concerns regarding the reliability and quality of the information presented.⁵ Despite this, the platform’s global reach—characterized by over 500 hours of video uploaded per minute and more than one billion hours viewed daily—has made it a major source of health-related information.⁶

In recent years, artificial intelligence (AI) has shown significant potential in dentistry, particularly in diagnosis, treatment planning, and clinical decision-making.⁷ These developments have gained increasing attention in endodontics and are becoming more visible on digital platforms such as YouTube™. Although AI applications in endodontics have primarily focused on radiological diagnosis, their use in clinical treatment processes and outcome prediction is also expanding.⁸ Artificial intelligence models have many promising application areas, such as the detection of periapical pathologies, identification of root fractures, and determination of working length.⁹ The internet, particularly social media platforms such as YouTube™, hosts a wide range of content, including education, entertainment, and advertising. Increasing use of these platforms by dental and medical professionals to inform patients and the public has led to a rise in both educational and promotional content.¹⁰ This facilitates the dissemination of emerging topics such as artificial intelligence, which are not yet widely incorporated into formal education.¹¹ However, the absence of scientific evaluation allows low-quality or outdated information to be easily uploaded, viewed, and shared.¹²

Previous studies have evaluated the quality, reliability, and educational value of health- and dentistry-related videos on YouTube™.¹³⁻¹⁶ These studies have shown that the accuracy and quality of information vary considerably across different dental topics. In addition, some studies have assessed YouTube™ content related to artificial intelligence in dentistry.³,¹¹ However, to date, no study has specifically examined the content characteristics and information quality of YouTube™ videos addressing artificial intelligence in endodontics. The main research question was whether YouTube™ videos specifically addressing artificial intelligence in endodontics provide reliable and educationally useful information, and whether these characteristics differ according to the source of upload. Therefore, this study aimed to evaluate the content characteristics, educational quality, and reliability of YouTube™ videos related to artificial intelligence in endodontics.

2. Materials and Methods

2.1. Ethical Considerations

Because this study analyzed only publicly available YouTube™ videos and did not involve human participants, patient data, or animals, ethics committee approval was not required. The video selection process was illustrated using a PRISMA-style flow diagram to ensure a structured and transparent selection procedure (Figure 1).

2.2. Search Strategy

YouTube™ was searched between 10 and 12 March 2026. To reduce the effect of day-to-day fluctuations in search results, searches were performed on each day between 10:00 and 12:00 local time in Türkiye. The platform was accessed from Türkiye via Google Chrome. All

searches were conducted in incognito mode and without logging into a Google or YouTube™ account in order to minimize personalization bias.

The search terms were “artificial intelligence in endodontics,” “machine learning in endodontics,” “deep learning in endodontics,” and “neural network in endodontics,” and the default “sort by relevance” filter was applied. These terms were intentionally selected to identify videos explicitly labeled as being related to artificial intelligence in endodontics. Accordingly, the search strategy was designed to capture directly indexed content rather than all potentially relevant videos on broader endodontic or dental artificial intelligence applications.

2.3. Eligibility Criteria

Videos were included if they were: (1) in English, (2) publicly accessible on YouTube™, and (3) directly related to artificial intelligence in endodontics. Videos were excluded if they were duplicate records, non-English, irrelevant to the search topic, YouTube™ Shorts, advertisements, reuploaded content, or videos with insufficient educational or scientific content, as well as those with poor audio and/or visual quality. Videos were considered to have insufficient educational or scientific content if they mentioned artificial intelligence only superficially, lacked explanatory information relevant to endodontics, or did not provide scientifically interpretable information about AI-related endodontic applications.

2.4. Screening and Selection Process

A total of 101 records were initially identified across the four predefined YouTube™ search queries: 48 using “artificial intelligence in endodontics,” 23 using “machine learning in endodontics,” 16 using “deep learning in endodontics,” and 14 using “neural network in endodontics.” The results from all searches were pooled prior to screening, and 18 duplicate videos were removed. The remaining 83 videos were screened by title and description. At this stage, 44 videos were excluded because they were not related to endodontics, not related to artificial intelligence, non-English videos, YouTube™ Shorts, advertisements, or irrelevant content. Subsequently, 39 full-length videos were assessed for eligibility. Of these, 26 videos were excluded

because they had insufficient educational or scientific content, were reuploaded videos, had inadequate audio or video quality, or were not directly focused on artificial intelligence in endodontics. Finally, 13 videos met the eligibility criteria and were included in the qualitative synthesis. Video screening and scoring were performed independently by two evaluators with expertise in endodontics. Disagreements were resolved through discussion and consensus. The selection process is presented in the PRISMA-style flow diagram in Figure 1.

2.5. Data Collection

The demographic information (URL and title), duration (sec), number of views, number of likes, number of comments, and time since upload (days) of each video were evaluated. Duplicate videos were identified and removed by comparing their URLs prior to the final analysis. The source of the videos was initially classified as healthcare professionals, commercial companies, or others. Because no meaningful standalone analysis could be performed for commercial companies due to the small number of videos, non-healthcare sources were combined under the category “other” for statistical comparisons. The “other” category included non-healthcare individual content creators, general educational channels, and non-specialized sources that did not represent formal healthcare professionals.

The reliability of the content in the videos was evaluated using a modified 5-item DISCERN (mDISCERN) instrument. The original DISCERN instrument was adapted into a simplified 5-item version, focusing on key aspects of reliability such as clarity of aims, use of reliable sources, balance of information, and acknowledgment of uncertainty. Each item was evaluated using a dichotomous scoring system (yes = 1, no = 0), resulting in a total score ranging from 0 to 5. This modification was applied to improve feasibility and consistency when assessing short and heterogeneous video content, in line with previous studies evaluating online health information (Table 1). The quality of the video information was evaluated using the modified Global Quality Scale (mGQS), with scores ranging from 1 to 5 (Table 2). Inter-rater agreement between the two evaluators was assessed using kappa statistics. The kappa values were 0.866 for mDISCERN and 0.743 for mGQS.

Although these tools were originally designed for patient-oriented content, they have been widely used in studies evaluating online health information and were therefore adopted to enable comparison with existing literature.

2.6. Statistical Analyses

All statistical analyses were performed at a 95% confidence level using the IBM SPSS Statistics version 21.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics were presented as mean, standard deviation, median, minimum, and maximum. The Shapiro-Wilk test was used to test the normality of the distribution of continuous quantitative data. Independent Samples T-test was used to detect the difference between parametric data, and the Mann-Whitney U test was used to detect the difference between nonparametric data. The association between mDISCERN and mGQS scores was assessed using Spearman’s rank

correlation analysis, as both variables were ordinal in nature. Correlation coefficient (rho) and two-tailed p values were calculated.

3. Results

Inter-rater agreement was substantial for both evaluation tools, with kappa values of 0.866 for mDISCERN and 0.743 for mGQS. Of the included videos, 69.2% were uploaded by healthcare professionals and 30.8% by other sources.

Across all included videos, the mean duration was 1917.1 seconds, the mean number of views was 687.9, the mean number of likes was 17.1, the mean number of comments was 0.08, and the mean time since upload was 967.2 days.

mDISCERN scores were significantly higher in videos uploaded by healthcare professionals than in those uploaded by other sources (p=0.034). No significant differences were found between uploader groups in terms of mGQS scores, views, likes, comments, or time since upload (p>0.05). Videos uploaded by healthcare professionals were also significantly longer in duration (p=0.004) (Table 3).

A strong, positive, and statistically significant correlation was found between mDISCERN and mGQS scores (Spearman’s rho = 0.756, p = 0.003).

4. Discussion

In this study, the content characteristics and information quality of YouTube™ videos related to artificial intelligence in endodontics were evaluated. Although previous studies have examined YouTube™ content on artificial intelligence and dentistry,³,¹¹ to the best of our knowledge, no study has specifically assessed videos addressing artificial intelligence in endodontics. Therefore, the present study contributes to filling this gap in the literature.

The relatively small number of eligible videos may suggest that artificial intelligence in endodontics remains underrepresented on YouTube™. However, this finding should be interpreted with caution, as relevant content may also be indexed under broader or application-specific terms. Accordingly, the results likely reflect the visibility of explicitly labelled content rather than the total volume of relevant material.

Most of the included videos were uploaded by healthcare professionals (69.2%), supporting previous reports that professional sources contribute substantially to dental YouTube™ content.^17-20 This finding suggests that healthcare professionals may play a key role in disseminating information on emerging topics such as artificial intelligence in endodontics.

Additionally, it was determined that the mDISCERN scores of the videos uploaded by healthcare professionals were significantly higher compared to other uploaders. This finding is consistent with previous studies indicating that professionally produced YouTube™ content tends to demonstrate higher reliability and educational value.^21,22 These findings suggest that greater involvement of healthcare professionals may contribute to improving the reliability of educational content on emerging topics such as artificial intelligence in endodontics.

Although videos uploaded by healthcare professionals showed higher mDISCERN scores, no significant difference was observed between uploader groups in terms of mGQS. This discrepancy may reflect that mDISCERN and mGQS assess related but distinct constructs, with mDISCERN focusing on reliability and mGQS on overall educational quality. In addition, the small sample size may have limited the statistical power to detect differences in mGQS. While the positive correlation between mDISCERN and mGQS suggests that more reliable videos may also have higher overall quality, this relationship did not translate into significant differences between uploader groups.

It should also be noted that both mGQS and mDISCERN-based instruments were originally developed to assess patient-oriented health information. However, the topic of artificial intelligence in endodontics is relatively technical and may be more directly relevant to clinicians, dental students, researchers, and educators. Therefore, the present study primarily reflects the general educational quality and reliability of the videos rather than strictly patient-oriented usefulness. This potential

mismatch between the evaluation tools and the subject matter should be considered when interpreting the findings. Accordingly, the findings should be interpreted within the context of professional and general educational value.

On the other hand, no significant difference was found among the uploaders in terms of mGQS score, number of views, number of likes, number of comments, and time since upload. This lack of statistically significant differences may reflect limited statistical power rather than true equivalence between groups. Accordingly, video popularity or user interaction may not necessarily reflect content quality or scientific accuracy. In other words, high viewing and interaction rates do not always indicate high information quality; similarly, low interaction levels do not mean that the content quality is low. Similar inconsistencies have been reported in previous studies, which may be attributed to methodological differences such as sample size, topic, and evaluation criteria.^19,23

It should also be noted that the “other” category included a limited number of videos, which may have reduced the statistical power of comparisons between uploader groups. Therefore, the observed differences between healthcare professionals and other uploaders should be interpreted with caution. The small size of the “other” group may limit the reliability and generalizability of these findings.

Videos uploaded by healthcare professionals were also significantly longer in duration, which may reflect a more detailed presentation of artificial intelligence applications in endodontic diagnosis and treatment planning.

This study has several limitations. YouTube™ is a dynamic platform, and video characteristics may change over time; therefore, the findings reflect only the data at the time of collection. In addition, the analysis was limited to publicly available videos identified through specific keywords and did not include other digital platforms. The use of standardized evaluation tools may not fully capture the technical accuracy or clinical applicability of rapidly evolving artificial intelligence applications, and the simplified mDISCERN instrument may reduce the multidimensional assessment of reliability. Furthermore, the keyword-based search strategy may not represent all relevant content, and the relatively small number of videos reflects the emerging nature of this topic. Despite these limitations, the study provides an initial overview of the reliability and educational quality of YouTube™ content related to artificial intelligence in endodontics.

5. Conclusion

Within the limits of the present keyword-based search strategy, YouTube™ content explicitly addressing artificial intelligence in endodontics appears to be limited in number. Among the available videos, those uploaded by healthcare professionals tended to show higher reliability scores; however, this observation should be interpreted cautiously due to the limited sample size. Nevertheless, broader search strategies and future studies with expanded screening methods are needed to better characterize the overall presence and educational quality of AI-related endodontic content on YouTube™. These findings suggest the importance of critically evaluating YouTube™ content and may indicate the potential benefit of greater involvement of healthcare professionals in producing accurate and reliable educational materials on emerging topics such as artificial intelligence in endodontics.

References

1. Meghana K. Artificial intelligence and sentiment analysis in YouTube™ comments: a comprehensive overview. In: Proceedings of the 2nd International Conference on Intelligent Data Communication Technologies and Internet of Things (IDCIoT); 2024. IEEE; 2024.
2. Uzel İ, Ghabchi B, Akalın A, Eden E. YouTube™ as an information source in paediatric dentistry education: reliability and quality analysis. PLoS One. 2023;18(3):e0283300.
3. Cesur Aydın K, Güneç HG. Quality of information on YouTube™ about artificial intelligence in dental radiology. J Dent Educ. 2020;84(10):1166-1172.
4. Habes M, Tahat K, Tahat D, Attar RW, Mansoori A, Ketbi N. The theory of planned behavior regarding artificial intelligence in recommendations and selection of YouTube™ news content. In: Proceedings of the International Conference on Multimedia Computing, Networking and Applications (MCNA); 2023. IEEE; 2023.
5. Dias da Silva MA, Pereira AC, Walmsley AD. Who is providing dental education content via YouTube™? Br Dent J. 2019;226(6):437-440.
6. Osman W, Mohamed F, Elhassan M, Shoufan A. Is YouTube™ a reliable source of health-related information? A systematic review. BMC Med Educ. 2022;22(1):382.
7. Mallineni SK, Sethi M, Punugoti D, et al. Artificial intelligence in dentistry: a descriptive review. Bioengineering (Basel). 2024;11(12):1267.
8. Setzer F, Li J, Khan A. The use of artificial intelligence in endodontics. J Dent Res. 2024;103(9):853-862.
9. Lai G, Dunlap C, Gluskin A, Nehme WB, Azim AA. Artificial intelligence in endodontics. J Calif Dent Assoc. 2023;51(1):2199933.
10. Knösel M, Jung K, Bleckmann A. YouTube™, dentistry, and dental education. J Dent Educ. 2011;75(12):1558-1568.
11. Naik S, Al-Kheraif AA, Vellappally S. Artificial intelligence in dentistry: assessing the informational quality of YouTube™ videos. PLoS One. 2025;20(1):e0316635.
12. McLean S, Cook N, Rovira-Wilde A, Patel S, Kanagasingam S. Evaluating YouTube™ as a patient information source for the risks of root canal treatment. J Endod. 2023;49(2):155-161.
13. Akbaş M, Demir N, Bozoklu B, Yaman D. Assessing the quality and reliability of YouTube™ content on dental cleaning over time. Inform Health Soc Care. 2026;51(1):1-11.
14. AlSahafi RA, Alhazmi HA, Alkhalifah I, et al. Evaluation of the quality and educational value of YouTube™ videos on class IV resin composite restorations. Dent J (Basel). 2025;13(7):298.
15. Cerci Akçay H, Kargu EC, Şeker N. Assessing the reliability and educational value of YouTube™ videos on computer-controlled local anesthesia in dentistry. PLoS One. 2025;20(8):e0329291.
16. Hashim A, Shivappa AB. Evaluation of popularity, reliability, and quality of dental and oral microbiology videos on YouTube™ as sources of dental education. Int J Dent. 2026;2026:1509354.
17. Özkan EÇ. Are YouTube™ videos a reliable source to learn about the use of elastics (rubber bands) in orthodontic treatment? Curr Res Dent Sci. 2026;36(1):44-48.
18. Kuka GI, Gursoy H. Content quality and reliability of YouTube™ videos as a source of information about oral hygiene practices in adults. PeerJ. 2024;12:e18183.
19. Öz E, Kırzıoğlu Z. An analysis of YouTube™ videos on oral hygiene practices during the coronavirus disease 2019 pandemic. J Glob Oral Health. 2021;4:84-93.
20. Hwang S-H, Kim M-Y. Quality assessment of YouTube™ videos on dental caries and nutrition. J Korean Soc Dent Hyg. 2024;24(6):651-658.
21. Fidan M, Yağcı Ö. Evaluation of information reliability and quality of YouTube™ videos about teeth whitening. Curr Res Dent Sci. 2025;35(2):104-110.
22. Sarıbaş EE, Kandemir M. Are videos uploaded by dental professionals on lip repositioning surgery of higher quality? A YouTube™ video analysis. PLoS One. 2025;20(9):e0327194.
23. Özcan N, Orhan AE. Evaluation of YouTube™ videos about post and core restorations. HRU Int J Dent Oral Res. 2024;4(3):82-89.