This study aimed to evaluate the effectiveness of AI-based computer-aided detection/diagnosis (AI-CAD) scores in predicting invasive upgrade in ductal carcinoma in situ (DCIS) diagnosed on percutaneous biopsy. The authors included 440 DCIS cases from 420 women diagnosed via percutaneous biopsy between 2015 and 2019.
The results of the study showed that 26.6% of DCIS cases were upgraded to invasive cancer. Significant independent predictors of invasive upgrade included combined mammographic features, BI-RADS assessments of 4c or 5, and higher AI-CAD scores ≥ 50%. In cases where DCIS was detected on mammography, higher AI-CAD scores and combined features were also strong predictors.
The study concluded that the AI-CAD score is an independent predictor of invasive upgrade for DCIS. A higher AI-CAD score, especially above 75%, may serve as an objective imaging biomarker for predicting invasive cancer in DCIS diagnosed with percutaneous biopsy.
Key points:
- Predicting ductal carcinoma in situ upgrade is important, yet there is a lack of conclusive non-invasive biomarkers.
- AI-CAD scores—raw numbers, ≥ 50%, and ≥ 75%—predicted ductal carcinoma in situ upgrade independently.
- Quantitative AI-CAD results may help predict ductal carcinoma in situ upgrade and guide patient management.
Authors: Jiyoung Yoon, Juyeon Yang, Hye Sun Lee, Min Jung Kim, Vivian Youngjean Park, Miribi Rho & Jung Hyun Yoon
