In recent advancements in medical technology, a novel tool named Sybil is aiding physicians in assessing lung cancer risk well before it manifests on imaging scans. This innovative approach has the potential to enhance current screening protocols, making early detection more accessible.
Standard lung cancer screening typically involves low-dose CT scans; however, only around 20% of those eligible for screenings actually participate. Prior to 2022, the guidelines stipulated that CT scans be reserved for individuals aged 50 to 80 who had a significant smoking history and either currently smoked or had quit within the last 15 years.
“In fact, 50% of people diagnosed with lung cancer in the US each year do not meet the criteria for screening,” noted Dr. Lecia Sequist, a medical oncologist at Mass General Brigham Cancer Institute.
In response to this disparity, physicians at Mass General Brigham collaborated with engineers from MIT to create Sybil, an AI tool capable of analyzing a single CT scan to provide a risk score that predicts the probability of developing lung cancer over the next six years.
Recently, in 2023, researchers reported that Sybil attained an accuracy rate ranging from 86% to 94% in differentiating between high-risk and low-risk patients within a single year.
“What Sybil is achieving is something that radiologists cannot do, which is forecasting what might happen one or two years from now—possible developments that may not be visible on the current scan,” Sequist explained.
Sybil employs advanced pattern recognition techniques, trained on a vast dataset of thousands of scans, to identify subtle biological signals that elude the human eye.
“If a patient receives a high-risk score from Sybil after their CT scan today, what should that mean for their management? Should we consider modifying their treatment now to reduce their risk and lower the chances of lung cancer developing in the future? While we don’t yet have definitive answers, that is our hope,” Sequist expressed.
