AI Tool Predicts Breast Cancer Up To 5 Years Before It Develops

AI-assisted mammography may change breast cancer detection.

I’m Paul Ciuryek, MD, and this is MedSpresso: An independent, subscriber-supported newsletter summarizing the most important and impactful news, studies, and research from the fast-changing world of healthcare.

Early cancer detection is paramount to early and effective treatment.

The advent and fast-moving world of artificial intelligence (AI) will undoubtedly play a major role in oncology diagnosis and treatment.

For breast cancer patients, AI is already proving beneficial.

Researchers at Duke University developed an AI model—AsymMirai—to enhance early detection, reduce unnecessary testing, and save healthcare costs.

By using local bilateral dissimilarities and analyzing differences between left and right breast tissue, AsymMirai can predict breast cancer risk up to five years in advance.

“ASYMMIRAK” KEY FEATURES:

• Improved Prediction Accuracy: AsymMirai demonstrated high accuracy in predicting breast cancer risk, comparable to the state-of-the-art Mirai model, but with greater transparency. It was tested on over 210,000 mammograms from more than 81,000 patients, showing robust performance in predicting 1- to 5-year cancer risk.

• Easy Interpretability: The model's focus on tissue asymmetry makes its predictions understandable to clinicians, allowing them to verify and trust the AI's assessments. This transparency addresses a common criticism of AI in healthcare, where decision-making processes are often opaque.

HEALTHCARE IMPLICATIONS:

A transparent and accurate risk assessment tool that can detect breast cancer up to five years sooner should become widely adopted. Its use in clinical practice could lead to more efficient screening processes and better allocation of medical resources.

The AsymMirai model represents a step forward in making AI tools more trustworthy and accessible to the average patient. Additional research and validation in diverse populations will be essential to realizing its full benefits in the healthcare setting. This model, and those still to come, offer hope for the hundreds of thousands of patients who are diagnosed with breast cancer every year in the United States.

Sources: Medscape | Breastcancer.org


Like this post?

👉 Tap the 🧡 below to let us know!
👉 Hit ♻️ and share it with your network!
👉 Try a Premium Subscription free for 14 days!

Thanks for reading!