Researchers Develop AI Model that can Predict Lung Cancer Years in Advance

Photo from Alex Ouyang

Researchers have developed a new AI deep learning model at MIT to predict lung cancer risks up to six years in advance through a single CT scan. 

Lung cancer is the deadliest type of cancer in the world, and was the cause of almost 1.8 million deaths in 2020. “It’s the biggest cancer killer because it’s relatively common and relatively hard to treat,” says Florian Fintelmann, MGCC thoracic interventional radiologist.

In a paper published in the Journal of Clinical Oncology, Jameel Clinic, MGCC, and CGMH, research claimed that “Sybil”, the AI-based lung cancer detection model, obtained C-indices of 0.75, 0.81, and 0.80 over the course of six years from various sets of lung scans. The Concordance Index (C-index) is a commonly used metric to evaluate how good a prediction model is. Models achieving a C-index score over 0.7 are considered good, and over 0.8 is considered strong. 

Despite the model's success, Sybil was a huge challenge. Peter Mikhael, an MIT PhD student in electrical engineering and computer science, as well as an affiliate of Jameel Clinic and the MIT Computer Science and Artificial Intelligence Laboratory, described the process as “trying to find a needle in a haystack”.