OAK BROOK, Ill. – A groundbreaking study published today in the journal Radiology reveals that an artificial intelligence (AI) algorithm can significantly enhance breast cancer screening by reducing interval cancers by up to one-third. The research, conducted by experts at Massachusetts General Hospital and Harvard Medical School, highlights the potential of AI to improve the performance of digital breast tomosynthesis (DBT), a cutting-edge 3D mammography technology.
Interval breast cancers, which are diagnosed between regular screening mammography exams, often have poorer outcomes due to their aggressive nature and rapid growth. DBT has been instrumental in enhancing the visualization of breast lesions, particularly in dense tissue, but long-term data on its effectiveness remains limited. The study, led by Dr. Manisha Bahl, underscores the importance of AI in bridging this gap.
AI’s Role in Breast Cancer Detection
Dr. Bahl and her team conducted a retrospective analysis of 1,376 cases, focusing on 224 interval cancers in women who underwent DBT screening. The AI algorithm, known as Lunit INSIGHT DBT v1.1.0.0, successfully localized 32.6% of cancers that were initially missed by radiologists.
“My team and I were surprised to find that nearly one-third of interval cancers were detected and correctly localized by the AI algorithm on screening mammograms that had been interpreted as negative by radiologists, highlighting AI’s potential as a valuable second reader,” Dr. Bahl stated.
This study is among the first to specifically examine AI’s assistance in detecting interval cancers on DBT exams, setting a precedent for future research and clinical applications.
Methodology and Findings
To ensure accuracy, the research team employed a lesion-specific analysis, crediting the AI only when it correctly identified and localized the exact site of the cancer. This approach contrasts with exam-level analysis, which can overestimate the algorithm’s sensitivity by giving credit for any positive exam, regardless of accuracy.
“Focusing on lesion-level accuracy provides a more accurate reflection of the AI algorithm’s clinical performance,” Dr. Bahl explained.
The study revealed that cancers detected by the AI tended to be larger and more likely to be lymph node-positive, suggesting that AI may preferentially detect more aggressive tumors or those already advanced at screening.
Implications for Clinical Practice
The AI algorithm demonstrated impressive accuracy, correctly localizing 84.4% of 334 true-positive cancers and accurately categorizing 85.9% of 333 true-negative cases. It also identified 73.2% of 333 false-positive cases as negative, indicating its potential to refine screening outcomes.
“Our study shows that an AI algorithm can retrospectively detect and correctly localize nearly one-third of interval breast cancers on screening DBT exams, suggesting its potential to reduce the interval cancer rate and improve screening outcomes,” Dr. Bahl noted.
These findings support the integration of AI into DBT screening workflows, but the real-world impact will depend on radiologist adoption and validation across diverse clinical settings.
Looking Forward
The announcement comes as healthcare systems globally are increasingly adopting AI technologies to enhance diagnostic accuracy and efficiency. The potential reduction in interval cancer rates represents a significant advancement in breast cancer screening, promising better patient outcomes and reduced morbidity and mortality.
As AI continues to evolve, its role in healthcare is expected to expand, offering new opportunities for early detection and personalized treatment strategies. The study by Dr. Bahl and her colleagues marks a pivotal step in this journey, paving the way for future innovations in cancer detection and management.
Meanwhile, ongoing research and collaboration between technologists and healthcare providers will be crucial in ensuring that AI’s integration into clinical practice is both effective and ethical, ultimately benefiting patients worldwide.
