@inproceedings{nagar-etal-2025-umedsum,
    title = "u{M}ed{S}um: A Unified Framework for Clinical Abstractive Summarization",
    author = "Nagar, Aishik  and
      Liu, Yutong  and
      Liu, Andy T.  and
      Schlegel, Viktor  and
      Dwivedi, Vijay Prakash  and
      Kaliya-Perumal, Arun-Kumar  and
      Kalanchiam, Guna Pratheep  and
      Tang, Yili  and
      Tan, Robby T.",
    editor = "Che, Wanxiang  and
      Nabende, Joyce  and
      Shutova, Ekaterina  and
      Pilehvar, Mohammad Taher",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingestion-acl-25/2025.acl-long.134/",
    pages = "2653--2672",
    ISBN = "979-8-89176-251-0",
    abstract = "Clinical abstractive summarization struggles to balance faithfulness and informativeness, sacrificing key information or introducing confabulations. Techniques like in-context learning and fine-tuning have improved overall summary quality orthogonally, without considering the above issue. Conversely, methods aimed at improving faithfulness and informativeness, such as model reasoning and self improvement, have not been systematically evaluated in the clinical domain. We address this gap by first performing a comprehensive benchmark and study of six advanced abstractive summarization methods across three datasets using five reference-based and reference-free metrics, with the latter specifically assessing faithfulness and informativeness. Based on its findings we then develop uMedSum, a modular hybrid framework introducing novel approaches for sequential confabulation removal and key information addition. Our work outperforms previous GPT-4-based state-of-the-art (SOTA) methods in both quantitative metrics and expert evaluations, achieving an 11.8{\%} average improvement in dedicated faithfulness metrics over the previous SOTA. Doctors prefer uMedSum{'}s summaries 6 times more than previous SOTA in difficult cases containing confabulations or missing information. These results highlight uMedSum{'}s effectiveness and generalizability across various datasets and metrics, marking a significant advancement in clinical summarization. uMedSum toolkit is made available on GitHub."
}