ArticleBMC oral health2026
Evaluation of the performance of four different large language models (ChatGPT, DeepSeek, Copilot, and Gemini) in answering oral, and maxillofacial radiology questions: pilot study.
Article in BMC oral health, 2026. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 4 papers.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
4 citing papers in PubMed.
- Article
- Performance of large language models on undergraduate endodontic multiple-choice questions.BMC oral health · 2026Article
- A comparative analysis of large language models for providing oral cavity cancer information.Scientific reports · 2026Article
- A Narrative Review of Teledentistry in the AI Era: Scope, Opportunities, and Future Prospects in India.Cureus · 2026Review
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
6 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
objectivesThe aim of this study is to evaluate and compare the accuracy of large language models (LLMs) such as ChatGPT-4, DeepSeek-R1, Google Gemini and Microsoft Copilot in answering multiple-choice questions related to oral and maxillofacial radiology (OMFR). It is also aimed to analyze the accuracy rates of answers across different topics and question types. MATERIALS AND
methodsIn this study, 240 multiple-choice questions posed to dental students were used as a sample. While 200 of the questions were text-based and consisted of 10 different topics (20 questions each), 40 were image-based questions supported by JPEG format images. Topics included radiation physics, projection geometry, radiobiology, radiographic anatomy, imaging modalities, cysts and tumors, systemic diseases, paranasal sinus diseases, temporomandibular joint (TMJ) diseases and salivary gland diseases. The accuracy rate of the chatbots was determined by evaluating the answers of each large language model to the questions.
resultsFor text-based questions, the accuracy rates were as follows: ChatGPT 90.5%, DeepSeek 84.5%, Copilot 82.5%, and Gemini 81.0%. For image-based questions, ChatGPT achieved 90.0%, while DeepSeek, Copilot, and Gemini scored 0.0%, 32.5%, and 65.0%, respectively. ChatGPT demonstrated the highest overall accuracy, while DeepSeek had the lowest. Among all topics, the lowest success rate was observed in salivary gland diseases.
conclusionsThese results emphasize the supportive and complementary role of LLMs in OMFR education. However, it should be noted that these models may not be sufficient on their own in cases requiring specialized expertise. CLINICAL RELEVANCE: LLMs can provide significant benefits in dentistry, including support for diagnosis and treatment planning, improvement of clinical decision-making processes, patient communication and education, contribution to academic research, increased efficiency in routine tasks, and applications in tele-dentistry.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.