Sybil: A Validated Deep Learning Model to Predict Future Lung Cancer Risk From a Single Low-Dose Chest Computed Tomography

This study used lung cancer screening trial low dose CT scans to develop a model (Sybil) that predicts the likelihood of development of a lung cancer within the next year. The model was then tested using images from independent data sets totaling more than 27,000 patients. Areas under the receiver-operator curves were in the range of 0.85 to 0.95, indicating very good predictive value. The model may help individualize subsequent screening frequency or the need for diagnostic CT.