:

Theodore Taehoon Kim1, Michael Makutonin1, Reza Sirous1

  • 1From the Department of Radiology, George Washington University School of Medicine and Health Sciences, 2300 I St NW, Washington, DC 20052 (T.T.K., R.J.); Yale School of Medicine, New Haven, Conn (M.M.); and University of California San Francisco, San Francisco, Calif (R.S.).

概括

大型语言模型 (LLM) 在放射学中提供了潜力,但面临着幻觉和偏见等挑战. 通过快速工程和微调优化LLMs对于安全有效的医疗应用至关重要.