Vision Foundation Models for Computed Tomography
This work addresses the need for adaptable and reliable AI tools in radiology, though it is incremental as it applies existing foundation model concepts to medical imaging.
The authors tackled the challenge of developing a versatile AI model for radiology by creating CT-FM, a large-scale 3D pre-trained model for computed tomography, which achieved superior performance across tasks like segmentation and image retrieval compared to state-of-the-art models.
Foundation models (FMs) have shown transformative potential in radiology by performing diverse, complex tasks across imaging modalities. Here, we developed CT-FM, a large-scale 3D image-based pre-trained model designed explicitly for various radiological tasks. CT-FM was pre-trained using 148,000 computed tomography (CT) scans from the Imaging Data Commons through label-agnostic contrastive learning. We evaluated CT-FM across four categories of tasks, namely, whole-body and tumor segmentation, head CT triage, medical image retrieval, and semantic understanding, showing superior performance against state-of-the-art models. Beyond quantitative success, CT-FM demonstrated the ability to cluster regions anatomically and identify similar anatomical and structural concepts across scans. Furthermore, it remained robust across test-retest settings and indicated reasonable salient regions attached to its embeddings. This study demonstrates the value of large-scale medical imaging foundation models and by open-sourcing the model weights, code, and data, aims to support more adaptable, reliable, and interpretable AI solutions in radiology.