Curia: A Multi-Modal Foundation Model for Radiology
This addresses the impracticality of narrow, single-task models in radiology by enabling broad generalization across modalities and low-data settings, though it is incremental as an application of foundation models to this domain.
The authors tackled the lack of broad generalization in AI for radiology by introducing Curia, a multi-modal foundation model trained on 150,000 real-world exams, which accurately identifies organs, detects conditions, and predicts outcomes, matching or surpassing radiologists and recent models on a 19-task benchmark.
AI-assisted radiological interpretation is based on predominantly narrow, single-task models. This approach is impractical for covering the vast spectrum of imaging modalities, diseases, and radiological findings. Foundation models (FMs) hold the promise of broad generalization across modalities and in low-data settings. However, this potential has remained largely unrealized in radiology. We introduce Curia, a foundation model trained on the entire cross-sectional imaging output of a major hospital over several years, which to our knowledge is the largest such corpus of real-world data-encompassing 150,000 exams (130 TB). On a newly curated 19-task external validation benchmark, Curia accurately identifies organs, detects conditions like brain hemorrhages and myocardial infarctions, and predicts outcomes in tumor staging. Curia meets or surpasses the performance of radiologists and recent foundation models, and exhibits clinically significant emergent properties in cross-modality, and low-data regimes. To accelerate progress, we release our base model's weights at https://huggingface.co/raidium/curia.