AI can find women at risk of developing breast cancer during a mammogram screening

Today, a large number of women are diagnosed at a late stage and can also develop breast cancer between screening sessions. The AI-based risk model can be used to identify women who need additional screenings as a complement to regular screening so that breast cancer can be detected early. The current study confirms previous studies that showed that an artificial intelligence-based risk model found a group of women who were approximately seven times more likely to develop breast cancer compared to the general population.

Individual examination

– About six percent of women had such a high risk. Today these women are screened in the same way as low-risk women. “We believe that a tailored screening could be more suitable for these women,” says Mikael Eriksson.

But clinical use was not the aim of the current study, but rather to check whether the method, which has been previously evaluated in Sweden and the USA, also works in different mammography programs across Europe.

“First, you develop the model and test it on a slightly limited population, then you work further to show generalizability to other populations, and then you start to get to a clear point where we think the model works,” says Mikael Eriksson.

The next step in the research is to conduct a clinical study in Europe where women are tested during mammography screening and given different treatments depending on the risk value the AI ​​model gives them. This method has already been clinically evaluated in the USA for a few years.

– We are now studying the possibility of introducing this method in Europe as well, says Mikael Eriksson.

The research is funded by the Swedish Research Council and the Breast Cancer Society. Mikael Eriksson has obtained a patent for an image-based risk model for breast cancer licensed to iCAD, Nashua, NH.

Publishing

“European validation of an AI-based image-derived short-term risk model for individualized breast cancer screening – a nested case-control study” Mikael Eriksson, Marta Roman, Axel Graingholt, Xavier Castells, Andrea Nitrossi, Pierpaolo Baiacini, Silvia Hewang Kobrunner. Paolo G. Rossi. Lancet Regional Health – Europeonline December 6, 2023.