Cancer prevention: AI detects lung cancer long before it becomes visible
Researchers in Boston are on the verge of a major advance in lung cancer screening: artificial intelligence can find early signs of the disease years before they are recognisable on a CT scan.
Time is a decisive factor when it comes to the treatment of cancer. For this reason, regular, low-dose CT scans are recommended for people at risk for early detection of lung cancer. Radiologists can then detect the development of cancer using the CT images. These screenings can already reduce the risk of dying from lung cancer by up to 24%. But even with regular check-ups, the trained eye cannot recognise everything. This is where the artificial intelligence "Sybil" comes into play.
AI recognises cancer long before it is visible on CT scans
The new AI tool offers the chance to detect lung cancer much earlier and thus gain life-saving time for treatment. With a probability of 86 to 94 per cent, "Sybil" can predict whether a person will develop lung cancer in the next year. Researchers from the Mass General Cancer Centre and the Massachusetts Institute of Technology in Cambridge found this in their study ""Sybil": A Validated Deep Learning Model to Predict Future Lung Cancer Risk From a Single Low-Dose Chest Computed Tomography".
Other AI tools are already being used in radiology, mostly to support doctors in the diagnosis and treatment of cancer. However, "Sybil" is still unique in its ability to predict a person's future cancer risk.
Sybil has not yet been approved for use outside of clinical trials, but when it is, the AI could play a decisive role in increasing the early detection rate of lung cancer and thus possibly also increasing the survival rate of those affected.
How does "Sybil" work?
"Sybil" looks for clues as to where cancer is likely to occur and recognises early signs of lung cancer that are not yet visible to the human eye on a CT scan. To predict the risk of cancer, the AI relies on a single CT scan. It analyses the three-dimensional image and looks not only for signs of abnormal growth in the lungs, but also for other patterns or disorders that the researchers themselves do not yet fully understand.
Based on what it sees, "Sybil" then predicts whether a person will develop lung cancer in the next one to six years. In some cases, "Sybil" has already been able to detect signs of cancer that would only be recognised by the human eye years later on a CT scan. The AI tool can therefore support radiologists in making important treatment decisions, but cannot replace them entirely.
However, the AI is far from perfect
"Sybil" can make a decisive difference in the future, but not yet for everyone. And this is where the problem that often goes hand in hand with artificial intelligence becomes apparent: the information with which "Sybil" has been trained
Much of the data that comes from medical institutions or clinical studies does not represent the diversity in the population. As a result, AI tools are not developed in such a way that they are also reliable for people of colour, for example. The data used to develop the AI tool did not yet include enough black or Hispanic people to ensure broad applicability.
Work is currently underway to ensure that diversity is taken into account during the approval process for medical products so that the results benefit all people equally.
Saving lives through artificial intelligence
Lung cancer is one of the most common types of cancer. According to the Federal Statistical Office, 3014 people died from lung cancer in Switzerland in 2019 and 4500 cases were diagnosed. In Germany, there were 47,560 deaths attributable to lung cancer in the same year. Earlier diagnosis can literally save lives. But early detection of lung cancer is complicated. When symptoms such as persistent coughing or breathing difficulties appear, the cancer is usually already advanced and often unstoppable.
Caption photo:pexels/mart production
Science editor and biologist. I love animals and am fascinated by plants, their abilities and everything you can do with them. That's why my favourite place is always outside - somewhere in nature, preferably in my wild garden.