AI Revolutionizes Breast Cancer Prognosis: Precision Beyond Pathologists, Transforming Treatment Strategies
A study by Northwestern Medicine unveils the potential of artificial intelligence (AI) to revolutionize breast cancer treatment strategies. Unlike traditional methods, an AI tool outperformed expert pathologists in predicting patient outcomes by evaluating both cancerous and non-cancerous tissue elements. This innovation could spare patients unnecessary chemotherapy, addressing a critical need to minimize side effects. With breast cancer diagnoses expected to affect 300,000 U.S. women in 2023, the study’s findings offer a transformative approach to individualized treatment plans and improved prognostic accuracy.
- AI surpasses expert pathologists in predicting breast cancer patient outcomes.
- Identification of long-term survivors among high or intermediate-risk patients allows for reduced chemotherapy intensity or duration.
- Non-cancerous cell patterns are crucial in predicting outcomes, challenging traditional grading methods.
- The AI model assesses 26 properties in breast tissue, generating an overall prognostic score and individual scores for cancer, immune, and stromal cells.
- Adoption of the AI model can empower patients with more accurate risk estimates, facilitating informed decision-making.
- Potential to assess therapeutic response and adjust treatment intensity based on microscopic tissue changes over time.
- A collaboration with the American Cancer Society produced a diverse dataset, improving model generalization beyond academic medical centers.
- Prospective evaluation and validation for clinical use are underway, coinciding with Northwestern Medicine’s transition to digital diagnosis.
These patterns are challenging for a pathologist to evaluate as they can be difficult for the human eye to categorize reliably. The AI model measures these patterns and presents information to the pathologist in a way that makes the AI decision-making process clear to the pathologist.
— Lee Cooper, corresponding study author, associate professor of pathology at Northwestern University Feinberg School of Medicine