Artificial Intelligence-Based Chatbots’ Ability to Interpret Mammography Images: A Comparison of Chat-GPT 4o and Claude 3.5
Authors: Betül Nalan Karahan, Emre Emekli, Mahmut Altuğ Altın
Why is this study important?
The early and accurate interpretation of mammography images is vital for the timely diagnosis of breast cancer, the most common cancer among women worldwide. However, routine screening programs place a significant workload on radiologists. As artificial intelligence (AI)-based chatbots rapidly evolve, their ability to interpret medical images—particularly mammograms—holds potential for supporting healthcare systems. This study is important because it critically assesses whether current chatbot technologies like ChatGPT-4o and Claude 3.5 can accurately classify BI-RADS categories and breast parenchymal types, which are essential for clinical decision-making. By highlighting their current limitations and variability, the research underscores the need for more robust training and development of these tools before they can be safely integrated into clinical workflows. It also opens the door for future advancements and sets a benchmark for evaluating AI tools in radiologic diagnostics.
Read the Full Article (PDF)
Citation
Karahan, B. N., Emekli, E., & Altın, M. A. (2025). Artificial Intelligence-Based Chatbots’ Ability to Interpret Mammography Images: A Comparison of Chat-GPT 4o and Claude 3.5. 31(1):28–34.European Journal of Therapeutics. doi:10.58600/eurjther2599
@article{karahan2025chatbots,
author = {Karahan, Betül Nalan and Emekli, Emre and Altın, Mahmut Altuğ},
title = {Artificial Intelligence-Based Chatbots’ Ability to Interpret Mammography Images: A Comparison of Chat-GPT 4o and Claude 3.5},
journal = {European Journal of Therapeutics},
year = {2025},
volume = {31},
number = {1},
pages = {28--34},
doi = {10.58600/eurjther2599},
url = {https://doi.org/10.58600/eurjther2599}
}