Analyzing the Performance of ChatGPT in Cardiology and Vascular Pathologies
This demonstrates ChatGPT's potential as a tool for medical education in cardiology, though it is an incremental application of an existing model to a new domain.
The study evaluated ChatGPT's performance on cardiology and vascular pathology multiple-choice questions, finding it scored 92.10% accuracy (175/190 correct), outperforming two medical students who scored 85.78% and 82.63%.
The article aims to analyze the performance of ChatGPT, a large language model developed by OpenAI, in the context of cardiology and vascular pathologies. The study evaluated the accuracy of ChatGPT in answering challenging multiple-choice questions (QCM) using a dataset of 190 questions from the Siamois-QCM platform. The goal was to assess ChatGPT potential as a valuable tool in medical education compared to two well-ranked students of medicine. The results showed that ChatGPT outperformed the students, scoring 175 out of 190 correct answers with a percentage of 92.10\%, while the two students achieved scores of 163 and 159 with percentages of 85.78\% and 82.63\%, respectively. These results showcase how ChatGPT has the potential to be highly effective in the fields of cardiology and vascular pathologies by providing accurate answers to relevant questions.