ARTIFICIAL INTELLIGENCE IN THE MEDICAL CHAIR: A SCENARIO-BASED CURRENT EVALUATION ON SCIENTIFIC ACCURACY, SCOPE, AND READABILITY

This study aims to evaluate the potential use of artificial intelligence (AI) tools in clinical consultations, based on a scenario involving a pregnancy diagnosed with Spina Bifida (SB) during the intrauterine period. Twenty frequently asked questions from parents were submitted in Turkish to three different AI models (ChatGPT, Grok, and Gemini). The responses generated by each model were evaluated by two perinatology experts using the JAMA criteria, Global Quality Score, Modified DISCERN, Likert Scale, and the Ateşman Readability Formula. These evaluations focused on scientific accuracy, reliability, overall quality, readability, and interpretability.The dataset provided in this Excel file includes the 20 questions, the AI-generated responses, readability analysis results, and expert scoring based on the aforementioned scales.