Prime Highlights
- Artificial intelligence can assess the risk of breast cancer recurrence after DCIS treatment by analyzing routine mammograms, helping guide follow-up care and treatment decisions.
- Patients with AI risk scores of 73.5% or higher were significantly more likely to experience recurrence, showing the tool can match or outperform current clinical models.
Key Facts
- Researchers studied over 1,700 women treated for DCIS between 2012 and 2017, tracking recurrence over at least one year of follow-up.
- The AI system works with standard imaging, providing a non-invasive method to support treatment planning without requiring extra procedures.
Background
Artificial intelligence can help predict the risk of breast cancer recurrence after treatment for ductal carcinoma in situ by analysing mammograms taken before surgery, according to a new study published in the American Journal of Roentgenology.
Researchers assessed a commercially available AI system that measures recurrence risk using pre-operative imaging. They examined medical records of more than 1,700 women who underwent surgery for DCIS between 2012 and 2017 and had at least one year of follow-up.
The study found 28 cases of a second breast cancer after treatment. Seven patients had cancer come back in the same breast after a mastectomy, while 25 developed cancer in the other breast. Patients with an AI risk score of 73.5% or higher were much more likely to have a recurrence at both five and ten years.
The findings show that the AI tool can match or beat current clinical risk models used to predict recurrence. Researchers said the system provides a non-invasive method to support treatment planning and long-term monitoring.
DCIS has high survival rates, but if recurrence is detected late, the disease can become more aggressive. Doctors can tailor check-ups, imaging, and treatment by identifying high-risk patients early.
Lead author Jung Min Chang said AI-generated scores from routine mammograms could help doctors make better decisions on follow-up care and personalise treatment strategies.
The team found that the tool works with standard imaging, so clinics can use it without extra procedures. They said testing it in larger and more diverse patient groups will help confirm its use in routine care.
Experts say AI-based risk prediction could improve outcomes by ensuring closer monitoring for patients most likely to develop recurrent disease.
