Recently researchers at the National Cancer Institute and Global Good have developed an algorithm that can apparently identify precancerous changes in a woman’s cervix, and is said to be 1.3 times better compared to standard tests. One of the tests used to detect cervical cancer involves swabbing the cervix with a solution, but it seems that this isn’t particularly effective as sometimes the results might be confused between an infection, inflammation, or precancer.
By training an AI to know what precancer is supposed to look like, it will be faster and more effective at detecting it compared to humans. Speaking to the Daily Mail, Jennifer Loukissas, chief of communications in the Division of Cancer Epidemiology and Genetics at NCI said, “The cervix is swabbed with a weakened vinegar solution. The solution turns HPV white and any cellular changes white while pink is normal. But white can be infection, inflammation or precancer. That’s why it’s so difficult to interpret with the human eye.”
She adds, “With a human expert, it’s open to interpretation and, even though the brain is good at detecting patterns, the neural networks are better.” Now it might be a while before we start seeing it used widely, but it is a step in the right direction.