BibTex

@inproceedings{liu-etal-2023-exploring-boundaries,
    title = "Exploring the Boundaries of {GPT}-4 in Radiology",
    author = "Liu, Qianchu  and
      Hyland, Stephanie  and
      Bannur, Shruthi  and
      Bouzid, Kenza  and
      Castro, Daniel  and
      Wetscherek, Maria  and
      Tinn, Robert  and
      Sharma, Harshita  and
      P{\'e}rez-Garc{\'\i}a, Fernando  and
      Schwaighofer, Anton  and
      Rajpurkar, Pranav  and
      Khanna, Sameer  and
      Poon, Hoifung  and
      Usuyama, Naoto  and
      Thieme, Anja  and
      Nori, Aditya  and
      Lungren, Matthew  and
      Oktay, Ozan  and
      Alvarez-Valle, Javier",
    editor = "Bouamor, Houda  and
      Pino, Juan  and
      Bali, Kalika",
    booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
    month = dec,
    year = "2023",
    address = "Singapore",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2023.emnlp-main.891",
    doi = "10.18653/v1/2023.emnlp-main.891",
    pages = "14414--14445",
    abstract = "The recent success of general-domain large language models (LLMs) has significantly changed the natural language processing paradigm towards a unified foundation model across domains and applications. In this paper, we focus on assessing the performance of GPT-4, the most capable LLM so far, on the text-based applications for radiology reports, comparing against state-of-the-art (SOTA) radiology-specific models. Exploring various prompting strategies, we evaluated GPT-4 on a diverse range of common radiology tasks and we found GPT-4 either outperforms or is on par with current SOTA radiology models. With zero-shot prompting, GPT-4 already obtains substantial gains ($\approx$ 10{\%} absolute improvement) over radiology models in temporal sentence similarity classification (accuracy) and natural language inference ($F_1$). For tasks that require learning dataset-specific style or schema (e.g. findings summarisation), GPT-4 improves with example-based prompting and matches supervised SOTA. Our extensive error analysis with a board-certified radiologist shows GPT-4 has a sufficient level of radiology knowledge with only occasional errors in complex context that require nuanced domain knowledge. For findings summarisation, GPT-4 outputs are found to be overall comparable with existing manually-written impressions.",
}