Performance of a commercially available Generative Pre-trained Transformer (GPT) in describing radiolucent lesions in panoramic radiographs and establishing differential diagnoses.

OBJECTIVES: To evaluate the performance of a commercially available Generative Pre-trained Transformer (GPT) in describing and establishing differential diagnoses for radiolucent lesions in panoramic radiographs. MATERIALS AND METHODS: Twenty-eight panoramic radiographs, each containing a single radiolucent lesion, were evaluated in consensus by three examiners and a commercially available ChatGPT-3.5 model. They provided descriptions regarding internal structure (radiodensity, loculation), periphery (margin type, cortication), shape, location (bone, side, region, teeth/structures), and effects on adjacent structures (effect, adjacent structure). Diagnostic impressions related to origin, behavior, and nature were also provided. The GPT program was additionally prompted to provide differential diagnoses. Keywords used by the GPT program were compared to those used by the examiners and scored as 0 (incorrect), 0.5 (partially correct), or 1 (correct). Mean score values and standard deviation were calculated for each description. Performance in establishing differential diagnoses was assessed using Rank-1, -2, and - 3. RESULTS: Descriptions of margination, affected bone, and origin received the highest scores: 0.93, 0.93, and 0.87, respectively. Shape, region, teeth/structures, effect, affected region, and nature received considerably lower scores ranging from 0.22 to 0.50. Rank-1, -2, and - 3 demonstrated accuracy in 25%, 57.14%, and 67.85% of cases, respectively. CONCLUSION: The performance of the GPT program in describing and providing differential diagnoses for radiolucent lesions in panoramic radiographs is variable and at this stage limited in its use for clinical application. CLINICAL RELEVANCE: Understanding the potential role of GPT systems as an auxiliary tool in image interpretation is imperative to validate their clinical applicability.

ChatGPT: a tool for scientific writing or a threat to integrity?

The use of ChatGPT as a tool for writing and knowledge integration raises concerns about the potential for its use to replace critical thinking and academic writing skills. While ChatGPT can assist in generating text and suggesting appropriate language, it should not replace the human responsibility for creating innovative knowledge through experiential learning. The accuracy and quality of information provided by ChatGPT also require caution, as previous studies have reported inaccuracies in references used by chatbots. ChatGPT acknowledges certain limitations, including the potential for generating erroneous or biased content, and it is essential to exercise caution in interpreting its responses and recognize the indispensable role of human experience in the processes of information retrieval and knowledge creation. Furthermore, the challenge of distinguishing between papers written by humans or AI highlights the need for thorough review processes to prevent the spread of articles that could lead to the loss of confidence in the accuracy and integrity of scientific research. Overall, while the use of ChatGPT can be helpful, it is crucial to raise awareness of the potential issues associated with the use of ChatGPT, as well as to discuss boundaries so that AI can be used without compromising the quality of scientific articles and the integrity of evidence-based knowledge.