MOSAIC: A Multilingual, Taxonomy-Agnostic, and Computationally Efficient Approach for Radiological Report Classification in Low-Resource Settings

Alice Schiavone, Marco Fraccaro, Lea Marie Pehrson, Silvia Ingala, Rasmus Bonnevie, Michael Bachmann Nielsen, Vincent Beliveau, Melanie Ganz, Desmond Elliott


Abstract
Radiology reports contain rich clinical information that can be used to train imaging models without relying on costly manual annotation. However, existing approaches face critical limitations: rule-based methods struggle with linguistic variability, supervised models require large annotated datasets, and recent LLM-based systems depend on closed-source or resource-intensive models that are unsuitable for clinical use. Moreover, current solutions are largely restricted to English and single-modality, single-taxonomy datasets. We introduce MOSAIC, a multilingual, taxonomy-agnostic, and computationally efficient approach for radiological report classification. Built on a compact open-access language model (MedGemma-4B), MOSAIC supports both zero-/few-shot prompting and lightweight fine-tuning, enabling deployment on consumer-grade GPUs. We evaluate MOSAIC across seven datasets in English, Spanish, French, and Danish, spanning multiple imaging modalities and label taxonomies. The model achieves a mean macro F1 score of 88 across five chest X-ray datasets, approaching or exceeding expert-level performance, while requiring only 24 GB of GPU memory. With data augmentation, as few as 80 annotated samples are sufficient to reach a weighted F1 score of 82 on Danish reports, enabling large-scale cohort classification with minimal human effort. Code and models are open-source, offering a practical alternative to large or proprietary LLMs in clinical settings.
Anthology ID:
2026.clinicalnlp-1.35
Volume:
Proceedings of the 8th Workshop on Clinical Natural Language Processing (Clinical NLP) @ LREC 2026
Month:
May
Year:
2026
Address:
Palma, Mallorca (Spain)
Editors:
Asma Ben Abacha, Steven Bethard, Danielle Bitterman, Tristan Naumann, Kirk Roberts
Venues:
ClinicalNLP | WS
SIG:
Publisher:
ELRA Language Resources Association (ELRA)
Note:
Pages:
313–323
Language:
External URL:
https://lrec.elra.info/lrec2026-ws-clinicalnlp-35
DOI:
10.63317/3qvne99cm779
Bibkey:
Cite (ACL):
Alice Schiavone, Marco Fraccaro, Lea Marie Pehrson, Silvia Ingala, Rasmus Bonnevie, Michael Bachmann Nielsen, Vincent Beliveau, Melanie Ganz, and Desmond Elliott. 2026. MOSAIC: A Multilingual, Taxonomy-Agnostic, and Computationally Efficient Approach for Radiological Report Classification in Low-Resource Settings. In Proceedings of the 8th Workshop on Clinical Natural Language Processing (Clinical NLP) @ LREC 2026, pages 313–323, Palma, Mallorca (Spain). ELRA Language Resources Association (ELRA).
Cite (Informal):
MOSAIC: A Multilingual, Taxonomy-Agnostic, and Computationally Efficient Approach for Radiological Report Classification in Low-Resource Settings (Schiavone et al., ClinicalNLP 2026)
Copy Citation: