Abstract
This paper explores methods for extracting information from radiology reports that generalize across exam modalities to reduce requirements for annotated data. We demonstrate that multi-pass T5-based text-to-text generative models exhibit better generalization across exam modalities compared to approaches that employ BERT-based task-specific classification layers. We then develop methods that reduce the inference cost of the model, making large-scale corpus processing more feasible for clinical applications. Specifically, we introduce a generative technique that decomposes complex tasks into smaller subtask blocks, which improves a single-pass model when combined with multitask training. In addition, we leverage target-domain contexts during inference to enhance domain adaptation, enabling use of smaller models. Analyses offer insights into the benefits of different cost reduction strategies.- Anthology ID:
- 2023.clinicalnlp-1.38
- Volume:
- Proceedings of the 5th Clinical Natural Language Processing Workshop
- Month:
- July
- Year:
- 2023
- Address:
- Toronto, Canada
- Editors:
- Tristan Naumann, Asma Ben Abacha, Steven Bethard, Kirk Roberts, Anna Rumshisky
- Venue:
- ClinicalNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 344–357
- Language:
- URL:
- https://aclanthology.org/2023.clinicalnlp-1.38
- DOI:
- 10.18653/v1/2023.clinicalnlp-1.38
- Cite (ACL):
- Sitong Zhou, Meliha Yetisgen, and Mari Ostendorf. 2023. Building blocks for complex tasks: Robust generative event extraction for radiology reports under domain shifts. In Proceedings of the 5th Clinical Natural Language Processing Workshop, pages 344–357, Toronto, Canada. Association for Computational Linguistics.
- Cite (Informal):
- Building blocks for complex tasks: Robust generative event extraction for radiology reports under domain shifts (Zhou et al., ClinicalNLP 2023)
- PDF:
- https://preview.aclanthology.org/fix-volume-bibkeys/2023.clinicalnlp-1.38.pdf