Transforming lung cancer screening with SYBIL
13 Hospitals, 9 Countries
Our Mission
Over 50% of patients in the United States who are diagnosed with lung cancer die within 1 year of their diagnosis, making it the deadliest cancer. Early detection dramatically increases the 5-year survival rate for lung cancer up to 95%, but only 21% of patients diagnosed with lung cancer receive an early-stage diagnosis. By the time symptoms appear, it is often too late.
Our researchers are developing state-of-the-art clinical AI models that can learn from imaging data to improve current models of disease progression, prevent over-treatment, and narrow down to the cure. All of our algorithms are publicly available to the research community.
12
Hospitals
8
Countries
85,000+
Low-dose CT scans validated
SYBIL has been validated on over 85,000 low-dose CT (LDCT) scans from patients all over the world. This validation process is key to ensuring that SYBIL maintains high performance across an array of local populations and that it can be safely used on real patients.
Collaborate With Us
What is SYBIL?
SYBIL is a deep learning model that can analyze a patient’s LDCT to accurately predict the patient’s risk of developing lung cancer in the next 6 years.
SYBIL assigns a personalized risk score to the LDCT scan, helping clinicians determine when a patient should return for their next screening.
Who is SYBIL for?
Lung cancer is the leading cause of cancer-related deaths worldwide, accounting for the highest mortality rates among both men and women. It accounts for about 1 in 5 of all cancer deaths in the U.S. due to late stage diagnosis, which limits treatment options.
While smoking is the leading cause of lung cancer, responsible for approximately 85% of all cases, lung cancer rates among nonsmokers are on the rise.
The team behind SYBIL
Combining leading expertise in AI and clinical practice, researchers at MIT and Mass General Brigham joined forces to build SYBIL, a state-of-the-art deep learning model that assesses a person’s risk of developing lung cancer up to 6 years in advance by analyzing their low-dose CT scan.
Through this breakthrough technology, the researchers hope to improve early detection rates of lung cancer without increasing the burden on healthcare systems.
Democratizing world-class technology
12 hospitals around the world have installed SYBIL for the purposes of screening and risk assessment.
Thanks to the support of Wellcome Trust, we have been able to expand the deployment of SYBIL to include hospitals serving under-resourced regions.
The scale and computational power needed for bold ideas to work is not possible without the support of grants, foundations, and donors. Your support would enable MIT Jameel Clinic to make a better, healthier future for all.
Give Now
Overall, the model performance was outstanding with a risk prediction area under the curve of 0.92 at 1 year in the NLST test set and area under the curves of 0.86 and 0.94 in the Massachusetts General Hospital and CGMH validation cohorts, respectively. The findings were also nearly as impressive out to 6 years.Journal of Clinical Oncology
VOLUME 41, NUMBER 12 / APRIL 2023
The authors have graciously offered to provide the model code to other investigators to validate the usefulness or limitations of the model. Our hope is that investigators worldwide will take them up on that.Journal of Clinical Oncology
VOLUME 41, NUMBER 12 / APRIL 2023
In Chile, there are no cancer screening policies as defined by the WHO, only recommendations and for some types of cancers, financial coverage for the population. Therefore, there are tremendous opportunities regarding cancer prevention and early detection, especially for lower income people.Dr. Arnaldo Marín, PhD
Faculty of Medicine, Department of Basic and Clinical Oncology
University of Chile
At IRST, we firmly believe that integrating deep learning technologies like SYBIL and MIRAI in clinical practice will radically improve patient care. The use of AI algorithms by physicians and medical physicists can capture this opportunity and lead to the future of precision medicine because the utilization of imaging data is now only partially exploiting its full potential.Dr. Giacomo Feliciani, PhD
Medical Physics Unit
IRCCS Istituto Romagnolo per lo Studio dei Tumori “Dino Amadori” IRST
Using deep learning for cancer diagnosis and treatment
What does the future of cancer screening look like? Jameel Clinic AI faculty lead Regina Barzilay discusses current challenges in lung cancer screening, the importance of developing a personalized screening protocol, and how history may hint at what the future holds.
