As all knows lung cancer is the deadliest cancer in the world and its resulting more deaths then the next three major cancer together. It is additionally incredibly challenging for people to find the infection ahead of time by examining the scans.
Scientists at MIT (Massachusetts Establishment of Innovation) have developed another AI tool model that can foresee the lung cancer risk as long as six years in advance through a one low-dose CT scan and named it ‘Sybil’.
This AI model is designed to use a single low dose chest scan to predict the risk of lung cancers occurring 1-6 years after a screening. Whereas, present lung cancer risk predicting models require a mixture of clinical risk factors, radiologic annotations and demographic information to predict lung cancer risks.
Peter Mikhael, MIT PhD student and co-first-author likened the process to “trying to find a needle in a haystack.”
Make sure that ‘Sybil’ working fine and predict accurate result of lung cancer risks Fintelmann and his team worked with a hundreds of scans from two hospitals and the National Lung Cancer Screening Trial. The study showed that Sybil earned C-indices scores ranging from 0.75 to 0.80 (Values over 0.8 indicate a strong model.) and was able to predict both long-term and short-term lung cancer risks.
This model was more accurate while screening lung cancer risks one year in advance and obtained 0.86 to 0.94 range on a ROC-AUC probability curve (considered excellent for AUC values with 1.00 being the highest possible score).
The information shared in this blog is for educational purposes only and is not a substitute for medical advice. Please consult your healthcare professional for any medical needs.
Subscribe to our newsletter or follow us on Twitter or Facebook today and never miss an update!