Artificial Intelligence Could Transform Breast Cancer Detection and Recurrence Prediction
News, Oncoscience
June 30, 2026“Breast cancer remains one of the leading causes of death among women worldwide, early detection is critical for improving survival.”
BUFFALO, NY — June 30, 2026 — A new review was published in Volume 13 of Oncoscience on May 19, 2026, titled “Enhancing breast cancer detection with AI for early diagnosis and recurrence prediction.”
The review was led by first author Sidney Andre and corresponding authors Swarda Bandiwadekar and Mrunalini Pattarkine from the Harrisburg University of Science and Technology, PA.
Breast cancer remains one of the most common cancers affecting women worldwide, and early detection plays a critical role in improving survival. Although mammography, magnetic resonance imaging (MRI), ultrasound, and biopsy have dramatically advanced breast cancer diagnosis over recent decades, each method has important limitations. False-positive and false-negative results, variability in image interpretation, operator dependence, and high costs can delay diagnosis or expose patients to unnecessary procedures. As artificial intelligence (AI) rapidly evolves, researchers are increasingly exploring how it can enhance existing screening technologies and improve patient care.
In this review, the authors examine evidence from studies published between 2006 and 2025 to evaluate how AI is being integrated into conventional breast cancer screening methods. The review compares traditional diagnostic approaches with AI-assisted technologies, highlighting their potential to improve early detection, increase diagnostic accuracy, reduce radiologist workload, and predict cancer recurrence more effectively.
One of the strongest areas of progress involves mammography. The review summarizes evidence showing that AI-assisted mammography detected 29% more breast cancers than conventional mammography interpretation without increasing false-positive findings. In addition, AI reduced radiologist reading time by approximately 40%, suggesting that intelligent image analysis may improve both efficiency and diagnostic performance.
The authors also discuss advances in three-dimensional digital breast tomosynthesis. Compared with conventional two-dimensional mammography, AI-assisted three-dimensional imaging detected an additional 1.6 cancers per 1,000 screening examinations while reducing patient recall rates by approximately 2.2%, helping minimize unnecessary follow-up testing.
Magnetic resonance imaging represents another promising application. According to studies reviewed in the article, AI models were able to identify imaging features associated with future breast cancer development up to one year before diagnosis and correctly localized future cancer sites in 57% of cases. Rather than simply interpreting existing abnormalities, these systems may eventually help identify women at elevated risk before cancer becomes clinically apparent.
Artificial intelligence also demonstrated important benefits in breast ultrasound, particularly for less experienced radiologists. By assisting with lesion classification and image interpretation, AI improved diagnostic performance and helped reduce variability between readers. Similarly, AI-assisted pathology tools analyzed biopsy specimens more efficiently while improving prediction of breast cancer recurrence risk through integration of tissue imaging with clinical information.
Beyond improving diagnostic accuracy, the review highlights AI’s potential to support precision medicine. Rather than replacing clinicians, AI systems function as decision-support tools that identify subtle imaging patterns, quantify complex tissue features, estimate recurrence risk, and prioritize suspicious findings for further evaluation. These capabilities may allow physicians to make more informed clinical decisions while reducing unnecessary procedures and improving patient outcomes.
At the same time, the authors emphasize that important challenges remain before widespread clinical implementation. Many published AI models have been developed using data from single institutions or relatively homogeneous populations, making external validation essential. The review also discusses the need for larger prospective studies, careful evaluation of cost-effectiveness, regulatory oversight, and strategies to ensure equitable access across diverse healthcare settings.
“Integrating AI into breast cancer redefines diagnostic excellence, shifting breast cancer management toward a more proactive, precise, and patient-centered approach.”
According to the authors, future research should focus on validating AI models across larger and more diverse patient populations while addressing ethical, regulatory, and implementation challenges. They note that successful clinical adoption will require AI systems that are accurate, transparent, equitable, and capable of complementing—not replacing—the expertise of healthcare professionals.
Overall, this review highlights the growing role of artificial intelligence in breast cancer care. Evidence from multiple published studies suggests that AI has the potential to improve early detection, enhance diagnostic consistency, support recurrence prediction, and streamline clinical workflows. As these technologies continue to mature, they may help advance more personalized and efficient approaches to breast cancer screening and management.
DOI: https://doi.org/10.18632/oncoscience.660
Correspondence to: Swarda Bandiwadekar – Sbandiwadekar@my.harrisburgu.edu, Mrunalini Pattarkine – lpattarkine@harrisburgu.edu
Keywords: cancer, AI in breast cancer screening modalities, early detection, artificial intelligence (AI), recurrence prediction, breast imaging
Click here to sign up for free Altmetric alerts about this article.
About Oncoscience:
Oncoscience is a peer-reviewed, open-access, traditional journal covering the rapidly growing field of cancer research, especially emergent topics not currently covered by other journals. This journal has a special mission: freeing oncology from publication costs. It is free to readers and authors.
Oncoscience is indexed and archived by PubMed, PubMed Central, Scopus, META (Chan Zuckerberg Initiative) (2018-2022), and Dimensions (Digital Science).
To learn more about Oncoscience, visit Oncoscience.us and connect with us on social media:
For media inquiries, please contact media@impactjournals.com.
