Artificial intelligence performance in answering multiple-choice oral pathology questions: a comparative analysis DOI Creative Commons
Baki Yılmaz, Büşra Yılmaz, Furkan Özbey

и другие.

BMC Oral Health, Год журнала: 2025, Номер 25(1)

Опубликована: Апрель 15, 2025

Artificial intelligence (AI) has rapidly advanced in healthcare and dental education, significantly impacting diagnostic processes, treatment planning, academic training. The aim of this study is to evaluate the performance differences between different large language models (LLMs) by analyzing their accuracy rates answers multiple choice oral pathology questions. This evaluates eight LLMs (Gemini 1.5, Gemini 2, ChatGPT 4o, 4, o1, Copilot, Claude 3.5, Deepseek) answering multiple-choice questions from Turkish Dental Specialization Examination (DUS). A total 100 2012 2021 were analyzed. Questions classified as "case-based" or "knowledge-based". responses "correct" "incorrect" based on official answer keys. To prevent learning biases, no follow-up feedback provided after LLMs' responses. Significant observed among (p < 0.001). o1 achieved highest (96 correct, 4 incorrect), followed (84 correct), 2 Deepseek (82 correct each). Copilot had lowest (61 correct). Case-based showed notable variations = 0.034), where excelled. For knowledge-based questions, demonstrated Post-hoc analysis revealed that performed better than most other across both case-based 0.0031). variable proficiency with showing higher accuracy. shows promise a supplementary educational tool, though further validation required.

Язык: Английский

Correcting Multiple Spaces in Adult Patients With Precise Tooth Movement Control Using In‐House Clear Aligners DOI Creative Commons
Nguyen Thị Thanh Van,

Trieu Thi Phuong,

Vu Thi Van Anh

и другие.

Clinical Case Reports, Год журнала: 2025, Номер 13(4)

Опубликована: Апрель 1, 2025

ABSTRACT Diastema or spacing in the anterior teeth is one frequent complaint of patients for undergoing orthodontic treatment. This case report presents treatment an adult patient with multiple spacings due to a Bolton discrepancy that was treated mostly by tipping using clear aligners. After 16 months treatment, patient's profile improved (Upper incisor inclination U1‐FH: decreased from 127° 117.86°; lower IMPA: 103° 96°; Upper lip E‐line: 1.81 0.25 mm, Lower 1.74 1.25 mm). Clear aligners can effectively address issues who have high aesthetic demands.

Язык: Английский

Процитировано

0
Artificial intelligence performance in answering multiple-choice oral pathology questions: a comparative analysis DOI Creative Commons
Baki Yılmaz, Büşra Yılmaz, Furkan Özbey

и другие.

BMC Oral Health, Год журнала: 2025, Номер 25(1)

Опубликована: Апрель 15, 2025

Artificial intelligence (AI) has rapidly advanced in healthcare and dental education, significantly impacting diagnostic processes, treatment planning, academic training. The aim of this study is to evaluate the performance differences between different large language models (LLMs) by analyzing their accuracy rates answers multiple choice oral pathology questions. This evaluates eight LLMs (Gemini 1.5, Gemini 2, ChatGPT 4o, 4, o1, Copilot, Claude 3.5, Deepseek) answering multiple-choice questions from Turkish Dental Specialization Examination (DUS). A total 100 2012 2021 were analyzed. Questions classified as "case-based" or "knowledge-based". responses "correct" "incorrect" based on official answer keys. To prevent learning biases, no follow-up feedback provided after LLMs' responses. Significant observed among (p < 0.001). o1 achieved highest (96 correct, 4 incorrect), followed (84 correct), 2 Deepseek (82 correct each). Copilot had lowest (61 correct). Case-based showed notable variations = 0.034), where excelled. For knowledge-based questions, demonstrated Post-hoc analysis revealed that performed better than most other across both case-based 0.0031). variable proficiency with showing higher accuracy. shows promise a supplementary educational tool, though further validation required.

Язык: Английский

Процитировано

0