While early detection of cancer allows for better care, technology is evolving towards better diagnostics. In the United States, a new method has emerged that is more precise than the cancer biomarkers currently available on the market.
A better way to detect cancer early
Let us first remember that there are many cancers and that they are not all at the same level. For example, ovarian cancer is judged rare, deadly and underfundedaccording to some specialists. Audra Moran, director of the Ovarian Cancer Research Alliance (Ocra) recalled in a BBC article on December 20, 2024 that the best solution for cancer not to kill is to detect it on as soon as possible. Ideally, this detection should take place five years before the first symptoms.
Daniel Heller is a biomedical engineer at Memorial Sloan Kettering Cancer Center in New York (United States). The person concerned is at the origin of work allowing a new means of detecting more or less serious diseases, including cancer. More precisely, it is detect the beginnings of cancer using blood tests and this, with the help of artificial intelligence.
Scientists have placed nanotubes in blood samplescapable of generating different lights depending on the molecules present there. After collecting the data from the nanotubes, it is up to them to read and then interpret them. The researchers then trained an AI using around a hundred blood samples. The objective? Provide a valuable help in understanding the databy identifying which samples come from sick people.
Marketing in the near future
For Daniel Heller and his team, whose work was the subject of a first publication in 2022, their method offers a superior accuracy to already available cancer biomarkers. However, scientists are continuing their research to improve their prototype. This includes using more efficient sensors or even including more blood samples in AI training.
« Several serum biomarker tests for ovarian cancer are already in use. Unfortunately, these standalone biomarker measurements have proven ineffective for early detection. Currently, no screening strategy can identify ovarian cancer at an early enough stage to reduce mortality.« we can read in the publication.
A final version of this AI could be on the market within three to five years. Furthermore, the researchers hope that their tool will make it possible – effectively – to sort all diseases and not just detect ovarian cancer. In the near future, patients should have the possibility of knowing quickly if they have cancer and what cancer it is.
