American Journal of Roentgenology (AJR)

AJR InBrief February 2019

The Role of Augmented Intelligence in Breast Imaging

A new review in the February 2019 issue of AJR examines the role of augmented intelligence (AI) in breast imaging, specifically how it can support and enhance current practices to provide more personalized care.

The review examined several areas where developing AI could address current challenges in breast imaging, such as automating breast density assessment, improving lesion analysis, developing an integrated approach to risk stratification, and predicting prognosis and treatment response.

In this Q&A, we talked with two of the authors, Andrea Arieno and Stamatia V. Destounis, about their review and what’s next for AI.

How did you and your co-authors come up with the idea for this study?

Investigation into AI in the field of radiology has been growing immensely. We wanted to review the topic as it applies to breast imaging to describe the current status of AI and investigate where AI is going by highlighting some of the areas where AI can contribute, such as improving radiologist and practice efficiencies and patient outcomes.

What should readers take away from your article?

AI in breast imaging has the potential to provide our field with some great new tools to enhance those we already have and improve clinical decision making. Understanding the concepts of augmented intelligence will be extremely important for physicians, as AI weaves into more areas of breast imaging.

We see AI techniques that are being investigated and applied to areas of breast density assessment, computer-assisted diagnosis and detection, patient risk assessment, and patient outcomes. Ultimately, AI techniques may be implemented in an effort to streamline patient care and support breast imagers in detection, diagnosis, and patient management.

What recommendations do you have for future research as a result of this article?

Important areas for continued investigation and development include data mining to meld imaging and clinical data to better improve patient care; pattern recognition software for a variety of imaging modalities; and, ultimately, algorithms that can assist the radiologist with clinical decision making.