AI can now predict breast cancer risk up to five years before diagnosis, by analyzing slight asymmetries in breast tissue.
Researchers at Duke University have introduced AsymMirai, a simplified deep learning algorithm designed to predict breast cancer up to five years before official diagnosis. Unlike previous complex models, this system focuses on “asymmetries” – the small differences between the left and right breasts that show up on mammograms.
By focusing on these specific structural changes, artificial intelligence achieves high accuracy while remaining much easier for radiologists to interpret. This development points to an important direction in medical technology: using AI not for content creation, but to detect vital patterns that the human eye might miss.
The study, which analyzed over 210,000 mammograms, confirms that breast asymmetry is a powerful, but underestimated, biomarker of long-term health risk. Lead researcher Jon Donnelly says the simplicity and reliability of AsymMirai could revolutionize public health strategies, tailoring screening to individual risk.
By offering a five-year window for intervention, this technology brings a more preventative approach to cancer, catching it at its earliest and most treatable stages, and shifting the focus from reactive treatment to long-term prevention.
