A groundbreaking study published in The Lancet Digital Health has unveiled the promising impact of artificial intelligence (AI) in breast cancer screening, demonstrating a significant increase in cancer detection rates while simultaneously reducing the workload for radiologists.
The MASAI Trial: A Game Changer in Mammography
The Mammography Screening with Artificial Intelligence (MASAI) trial, conducted within Sweden’s national screening program, is the first randomized controlled study to evaluate AI-assisted mammography on a large scale. The trial enrolled over 105,000 women across four screening sites, randomly assigning them to either AI-supported screening or standard double reading by radiologists. The AI system, Transpara version 1.7.0, was used to triage examinations and highlight suspicious findings.
Key Findings: Higher Cancer Detection with AI
The study revealed that AI-supported screening detected 338 cancers among 53,043 participants, compared to 262 cancers among 52,872 participants in the standard screening group. This marked a 29% increase in cancer detection (6.4 per 1,000 screened vs. 5.0 per 1,000). Notably, AI-assisted screening detected more invasive cancers (270 vs. 217) and a higher proportion of high-risk, non-luminal A cancers, including triple-negative and HER2-positive cases. These cancers typically have poorer prognoses and a higher likelihood of becoming interval cancers.
In addition to increased cancer detection, the study found no significant rise in false positives (1.5% in the AI group vs. 1.4% in the control group) or recall rates (2.1% vs. 1.9%). Importantly, AI-supported screening resulted in a 44.2% reduction in screen-reading workload, freeing up valuable time for radiologists to focus on more complex cases. At the same time, the AI showed a significant increase in PPV of mammography recall of 19%.
Clinical Implications: A Step Toward Earlier Detection
The findings suggest that AI can contribute to the earlier detection of clinically significant breast cancer without an associated increase in overdiagnosis of low-risk lesions. The increase in detection of small, lymph-node negative invasive cancers implies a potential for downstaging, leading to improved treatment outcomes and reduced mortality rates.
The study’s authors, from Lund University Sweden, emphasized that AI’s role in screening is not to replace radiologists but to enhance their efficiency and accuracy. “The large reduction in screen-reading workload made possible by the AI-supported screen-reading procedure would free up time for breast radiologists to spend on more complex patient-centred tasks. Whether the time saving is cost-effective is related to the cost of the AI system. Since breast cancer treatment and related costs escalate with increasing stage, downstaging through earlier detection by use of AI would suggest lower morbidity and treatment costs. To address the cost- effectiveness of AI-supported screening, health economic analyses based on the MASAI trial are under way.“
The Road Ahead: AI’s Future in Breast Cancer Screening
Despite its promising results, the MASAI trial is still ongoing, with the primary outcome measure—interval cancer rate—set for assessment in 2025. This will provide further insights into whether AI-supported screening translates into lower rates of missed cancers between screening rounds.
Additionally, the economic viability of AI implementation remains an open question. While AI reduces workload and increases detection, its cost-effectiveness in national screening programs will need further analysis.
Conclusion
The MASAI trial presents compelling evidence that AI-supported mammography can improve cancer detection rates and reduces screen-reading workload without increasing false positive. As AI technology continues to evolve, its integration into breast cancer screening programs could lead to earlier diagnoses, improved patient outcomes, and a more efficient healthcare system.
With further studies and regulatory considerations, AI may soon become a standard component of mammography screening worldwide, offering a transformative approach to the early detection of breast cancer.
