Clinicians are increasingly using AI in mammography to improve breast cancer detection and risk assessment. A recent study evaluated how a commercial AI system predicts second breast cancers after DCIS treatment. Because DCIS recurrence remains a significant concern, these advanced tools provide essential prognostic insights for doctors. Consequently, AI scores could help identify patients at higher risk of developing ipsilateral or contralateral cancers after surgery.
The Role of AI in Mammography for Post-DCIS Risk
This retrospective study analyzed 1,740 patients who underwent surgery for DCIS between 2012 and 2017. Researchers found that AI scores from preoperative mammograms significantly correlated with future cancer risks. Specifically, an AI score of 73.5% or higher indicated a much greater risk of ipsilateral recurrence after breast-conserving surgery. Therefore, this technology acts as a powerful predictor for long-term clinical outcomes. Additionally, the AI system processed images efficiently without requiring invasive procedures or extra biopsies. This approach allows doctors to refine follow-up strategies for high-risk individuals in Indian clinical settings. Moreover, the study focused on five different centers to ensure data reliability. Thus, the findings offer a robust look at AI capabilities.
How AI Scores Compare to Clinical Models
Clinicians often rely on the Van Nuys Prognostic Index or MSKCC nomograms for risk assessment. However, this study demonstrated that AI scores offer superior or independent predictive value. Furthermore, the AI system showed a stronger association with recurrence than some traditional pathologic models. Thus, incorporating AI into the workflow enhances the accuracy of preoperative evaluations significantly. Moreover, these findings suggest that AI can complement existing clinical data to provide a holistic risk profile. Consequently, oncologists may find these tools useful for personalized patient management and treatment planning. This advancement marks a shift toward precision medicine in radiology. Eventually, such tools could reduce the burden of secondary surgeries for many patients.
Frequently Asked Questions
Q1: Can AI predict cancer in the contralateral breast?
Yes, the study found that AI scores were associated with the development of second breast cancers in both the treated and the opposite breast.
Q2: Is the AI score more accurate than traditional nomograms?
The research indicated that AI scores provided significant independent associations with recurrence, often outperforming or successfully complementing existing clinical risk models like the MSKCC nomogram.
References
- Yoon JH et al. Commercially Available Artificial Intelligence Score on Preoperative Mammography for Prediction of Future Breast Cancer After DCIS Treatment. AJR Am J Roentgenol. 2026 Feb 11. doi: 10.2214/AJR.25.34364. PMID: 41670539.
- NITI Aayog. National Strategy for Artificial Intelligence. 2018.
- World Health Organization. Breast Cancer in India: Trends and Challenges. 2023.