Does AI Improve Risk Prediction in Women with Breast Cancer?

Comparison of Mammography AI Algorithms with a Clinical Risk Model for 5-year Breast Cancer Risk Prediction: An Observational Study

Vignesh A Arasu, Laurel A Habel, Ninah S Achacoso, Diana S M Buist, Jason B Cord, Laura J Esserman, Nola M Hylton, M Maria Glymour, John Kornak, Lawrence H Kushi, Donald A Lewis, Vincent X Liu, Caitlin M Lydon, Diana L Miglioretti, Daniel A Navarro, Albert Pu, Li Shen, Weiva Sieh, Hyo-Chun Yoon, Catherine Lee

Study design and methodology

This observational study compared the performance of mammography artificial intelligence (AI) algorithms with a clinical risk model for predicting breast cancer risk over a 5-year period. The study included women who had a negative screening mammogram in 2016 and followed them until 2021. The AI algorithms evaluated five deep-learning models, while the clinical risk model used was the Breast Cancer Surveillance Consortium (BCSC) risk model.

Performance of mammography AI algorithms vs clinical risk model

The study found that the AI algorithms performed better than the BCSC model in predicting breast cancer risk at 0 to 5 years. The AI algorithms had higher time-dependent area under the receiver operating characteristic curve (AUC) scores than the BCSC model, ranging from 0.63 to 0.67. When the AI algorithms were combined with the BCSC model, the prediction of breast cancer risk was further improved, with time-dependent AUC scores ranging from 0.66 to 0.68.

Enhanced breast cancer risk prediction with AI algorithms

These findings suggest that mammography AI algorithms can enhance breast cancer risk prediction beyond traditional clinical risk factors. The AI algorithms demonstrated better discrimination in identifying women at higher risk of developing breast cancer within a 5-year timeframe. Combining AI and clinical risk models improved the performance of risk prediction.

Implications and future research on AI algorithms for breast cancer risk assessment

The study highlights the potential of AI algorithms to provide more accurate risk assessment for breast cancer, assisting in personalized screening and prevention strategies. However, further research is needed to evaluate the performance of other AI algorithms and to assess risk prediction beyond 5 years. Implementation of AI models in clinical practice should also consider potential costs and technical considerations.

Link to Article: Comparison of Mammography AI Algorithms with a Clinical Risk Model for 5-year Breast Cancer Risk Prediction: An Observational Study - PubMed (nih.gov)

This article was summarized by an AI tool that uses natural language processing. The tool is not perfect and may make mistakes or produce inaccurate or irrelevant information, but is reviewed by the post’s author prior to publishing. If you want to learn more about the article, please refer to the original source that is cited at the end of the article.