AI-Powered Breast Cancer Screening: A New Era of Efficiency and Accuracy

A recent clinical trial has demonstrated the feasibility of using artificial intelligence (AI) to streamline breast cancer screening, reducing radiologist workload whereas maintaining, and in some cases improving, cancer detection rates. The study, conducted in Spain, offers promising evidence for the integration of AI into routine mammography practices, potentially leading to faster diagnoses and more efficient healthcare systems.

The Study: A Paired, Noninferiority Trial

Researchers at the Maimónides Biomedical Research Institute of Córdoba, Spain, conducted a prospective, paired, noninferiority clinical trial involving 31,301 women undergoing routine mammograms between March 2022 and January 2024. The study, published in Nature, compared two screening strategies:

• Standard Double-Blind Reading: Traditional mammogram review by two radiologists without AI assistance.
• AI-Supported Screening: Cases classified by an AI system as low risk were automatically considered normal. The remaining cases, flagged as potentially cancerous by the AI, were double-read by radiologists with AI support.

The trial received a favorable ruling from the Institutional Review Board (IRB) at Reina Sofía University Hospital of Córdoba Research Ethics Committee (IRB No 4932) in March 2021. The study protocol was preregistered at ClinicalTrials.gov (NCT04949776).

Key Findings: Workload Reduction and Improved Detection

The results revealed significant benefits from the AI-supported screening strategy:

• Workload Reduction: Radiologist workload was reduced by 63.6%.
• Cancer Detection Rate: The cancer detection rate increased by 15.2% (from 6.3 to 7.3 per 1,000 screenings), with a 95% confidence interval of 6.6% to 24.4% (P < 0.001).
• Recall Rate: The recall rate was 14.8% higher (95% confidence interval 9.0% to 20.6%).

Subanalyses showed similar workload reductions in both digital mammography (-62.1%) and digital breast tomosynthesis (-65.5%). While the cancer detection rate increased in digital mammography (by 1.6 per 1,000), it remained stable in digital breast tomosynthesis.

The AI System: Transpara 1.7

The study utilized Transpara (version 1.7) from ScreenPoint Medical, a commercially available AI system for breast cancer detection. The system has been previously investigated in over 30 peer-reviewed studies, demonstrating comparable stand-alone cancer detection performance to radiologists and improved accuracy when used as a decision support tool.

Transpara analyzes mammograms using deep convolutional neural networks, identifying and classifying suspicious regions with a risk score from 1 to 100. The system automatically classifies exams as low risk (scores 1-7), intermediate risk (8-9), or elevated risk (10).

Implications for Breast Cancer Screening

This research suggests that partially automated AI workflows can safely and effectively reduce the burden on radiologists in breast cancer screening programs. By automating the assessment of low-risk cases, AI allows radiologists to focus their expertise on more complex and potentially cancerous images. This could lead to earlier detection of breast cancer and improved patient outcomes.

Study Details and Safety

The study was conducted within the Córdoba Breast Cancer Screening Unit in Spain, as part of the Andalusian screening program. Participants were women aged 50 to 71. Exclusion criteria included symptomatic breast cancer, breast prostheses, and images incompatible with the AI system (e.g., due to breast implants or image transfer errors). The study protocol prioritized patient safety, with full anonymization of images and adherence to data protection regulations. No adverse events were reported.

Future Directions

While these findings are encouraging, further research is needed to validate these results in diverse populations and healthcare settings. Continued development and refinement of AI algorithms, coupled with careful implementation and monitoring, will be crucial to maximizing the benefits of AI in breast cancer screening.