Enhancing Mammography Efficiency: AI-Human Collaboration Could Reduce Screening Costs by 30%

Recent research suggests that combining artificial intelligence (AI) with human radiologists in mammography screenings can significantly cut costs—by up to 30%—without compromising patient safety. The study focuses on a 'delegation' approach, where AI is used to triage low-risk mammograms and identify high-risk cases requiring detailed human review. This strategy not only streamlines workflow but also addresses the growing demand for early breast cancer detection amid a shortage of radiologists.

The research, published in Nature Communications, was led by experts from the University of Illinois at Urbana-Champaign, the University of Texas at Dallas, and NYU Langone Health. Using a decision model to compare three strategies—current expert-alone, full automation by AI, and delegated AI assistance—the team found that the delegation method offered the greatest cost savings with no loss in diagnostic accuracy.

While fully automated systems may seem attractive, the study cautions that AI currently performs best at classifying straightforward, low-risk cases. For complex or ambiguous images, human judgment remains superior. Therefore, the delegation model leverages AI’s efficiency for routine cases, freeing up radiologists to focus on more challenging diagnoses.

The implications of this research are far-reaching, potentially transforming how hospitals approach screening workflows. In the U.S., where nearly 40 million mammograms are performed annually, reducing false positives—which can lead to unnecessary anxiety, follow-up procedures, and biopsies—could greatly benefit patients and healthcare providers alike.

Moreover, the approach may be especially advantageous in low-resource settings or regions with limited radiologist availability, such as developing countries. However, issues like legal liability for AI systems present hurdles, as healthcare organizations must consider accountability when deploying AI-assisted strategies.

The model's adaptability extends beyond breast cancer to other medical diagnostics like pathology and dermatology, where efficiency and accuracy are critically important. With AI capable of operating around the clock without fatigue, its integration into healthcare workflows promises to advance the efficiency and accessibility of medical services.

Ultimately, the study emphasizes a balanced approach: AI should serve as a supportive tool that enhances human expertise rather than replacing it, optimizing healthcare outcomes through strategic collaboration.