An artificial intelligence (AI) based deep learning model can predict lung cancer risk using only a single low dose computed tomography (CT) scan, according to research carried out in Korea.
The deep learning model, which is known as Sybil, was originally developed using National Lung Screening Trial (NLST) data by investigators from the Massachusetts Institute of Technology and Harvard Medical School.
The current study was presented at the 2025 American Thoracic Society International Conference this week.
The study included 21,087 individuals aged 50-80 years who arranged their own lung cancer screening between 2009 and 2021, and the participants were followed up until 2024.
The results showed Sybil was able to predict whether people in the study would go on to develop lung cancer within six years to a good level of accuracy.
In the overall group, the model could correctly predict whether or not a person would get lung cancer within one year and six years with 86% and 74% accuracy, respectively.
【MORE】