Could AI provide better cervical cancer screening than traditional methods?

Researchers at Lehigh University’s Image Data Emulation & Analysis Laboratory spent 10 years developing an artificial intelligence (AI) screening tool for cervical cancer.

The data published in Pattern Recognition revealed that this tool may be better than traditional methods—such as pap smears and HPV tests—and is much more affordable. As part of their development, the Lehigh team programmed a computer to recognize abnormalities in cervix cells and tissues.

Using algorithms, the researchers trained the computers to familiarize and classify the dysplasia seen in cervigrams. Seven series of algorithms were made to grade dysplasia.

As part of the study, 1,112 patients’ cervigrams were used, including 345 patients with moderate or severe dysplasia, with higher chances of developing cancer, and 767 low grade, mild dysplasia, which the body can normally get rid of on its own.

Full article here: