Researchers from the University of California, Berkeley (UCB) and the University of California, San Francisco (UCSF) have released Pillar (Pillar-0), an open-source AI model.
The model analyzes CT and MR medical images and recognizes conditions with accuracy, according to validation studies performed by the researchers. Pillar-0 interprets 3D volumes directly and can recognize hundreds of conditions from a single CT or MRI exam.
The team validated Pillar-0 on chest CT, abdomen CT, brain CT, and breast MRI scans from UCSF. The model achieved an area under the curve (AUC) of 0.87 across over 350 findings on this data.
The researchers also fine-tuned Pillar-0 and improved the lung cancer prediction tool, Sybil-1, by 7% in an external validation study at Massachusetts General Hospital in Boston.
The team is releasing the complete Pillar-0 codebase, trained models, evaluation, and data pipelines to the public to accelerate research and clinical adoption. It also plans to expand capabilities across additional imaging modalities to full-grounded report generation.
