Wang Jinge, Yu Thomas C, Kolodney Michael S, Perrotta Peter L, Hu Gangqing
Department of Microbiology, Immunology and Cell Biology, West Virginia University, Morgantown, WV, 26506, USA.
Sona Dermatology, Bethesda, MD, 20817, USA.
Ann Biomed Eng. 2025 Jan;53(1):5-8. doi: 10.1007/s10439-024-03656-0. Epub 2024 Nov 27.
Color vision deficiency (CVD) affects a significant portion of the population, yet its impact is often overlooked in medical education, especially in visually demanding specialties like dermatology, pathology, and radiology. In this study, we investigated the potential of ChatGPT to comprehend CVD-simulated images in image-based diagnostic tasks. Notably, the model successfully adapted its diagnostic reasoning to match CVD-modified color perception while preserving high prediction accuracy. These findings highlight the potential of using ChatGPT to foster more inclusive learning environments for individuals with CVD in visually intensive medical specialties.
色觉缺陷（CVD）影响着相当一部分人口，然而其影响在医学教育中常常被忽视，尤其是在皮肤科、病理学和放射学等对视觉要求较高的专业领域。在本研究中，我们调查了ChatGPT在基于图像的诊断任务中理解CVD模拟图像的潜力。值得注意的是，该模型成功地调整了其诊断推理，以匹配CVD改变后的颜色感知，同时保持了较高的预测准确性。这些发现凸显了使用ChatGPT为视觉要求高的医学专业中患有CVD的个体营造更具包容性的学习环境的潜力。
