The role of large language models in dental diagnosis, decision-making, and communication: A systematic review

The integration of LLMs in the healthcare sector offers significant potential for improving medical practice and patient care. They assist in medical diagnosis by analyzing patient symptoms and other relevant data. This systematic review proposes to assess the existing literature on the use of LLMs. The comprehensive review was conducted in PubMed, Web of Science, Google Scholar and Scopus databases using keywords related to LLMs for articles published in English language at all times. Data was extracted to determine the applications, evaluation criteria and outcomes of LLMs in dentistry. Of 1142 records, 38 studies fulfilled the inclusion criteria. Among all studies, ChatGPT-3.5 was the most frequently used LLM. Most of the studies addressed patient questions during or after dental treatment. Most studies asked open-ended questions, and the Likert scale was the most used evaluation scale. This systematic review has shown that LLMs can increase efficiency, improve patient care and diagnostic support in dentistry. However, due to the risk of incorrect and incomplete information, LLMs should be used as clinical decision support tools. To increase their effectiveness and reliability, they should be trained on rigorously validated and continuously updated data sets.