Inclusive AI for radiology: Optimising ChatGPT-4 with advanced prompt engineering.

Journal: Clinical imaging
PMID:

Abstract

This letter responds to the article "Encouragement vs. liability: How prompt engineering influences ChatGPT-4's radiology exam performance," offering additional perspectives on optimising ChatGPT-4 for Radiology applications. While the study highlights the significance of prompt engineering, we suggest that addressing additional key challenges such as age-related diagnostic needs, socio-economic diversity, data security, and liability concerns is essential for responsible AI integration. Incorporating adaptive prompts, training the model on diverse datasets, and securely integrating it with electronic health records (EHRs) can enhance its reliability and inclusiveness. By balancing prompt design with privacy and accountability frameworks, ChatGPT-4 can become a more effective tool in radiology, aiding clinicians without compromising human oversight.

Authors

  • Juhi Yasmeen
    School of Advanced Sciences and Languages, VIT Bhopal University, Sehore, MP, India. Electronic address: juhiyasmeen@vitbhopal.ac.in.
  • Md Tauseef Qamar
    School of Advanced Sciences and Languages, VIT Bhopal University, Sehore, MP, India. Electronic address: tauseefqamar@vitbhopal.ac.in.
  • Subuhi Yasmeen
    Faculty of Dentistry, Jamia Millia Islamia, New Delhi, India.