OAK BROOK, Ill. – A groundbreaking study published today in Radiology, the journal of the Radiological Society of North America (RSNA), reveals that an artificial intelligence (AI) algorithm can significantly improve the detection of interval breast cancers, potentially reducing their occurrence by up to one-third. This advancement could enhance the effectiveness of digital breast tomosynthesis (DBT), a newer form of 3D mammography.
Interval breast cancers, which are symptomatic cancers diagnosed between regular screening mammograms, often have poorer outcomes due to their aggressive nature and rapid growth. The study, led by Dr. Manisha Bahl of Massachusetts General Hospital and Harvard Medical School, analyzed 1,376 cases and found that the AI algorithm, Lunit INSIGHT DBT v1.1.0.0, correctly localized 32.6% of previously undetected interval cancers.
AI’s Role in Breast Cancer Screening
DBT has been lauded for its ability to improve visualization of breast lesions, particularly in dense breast tissue. However, because it is a relatively new technology, long-term data on patient outcomes remain limited. Dr. Bahl explained, “Given the lack of long-term data on breast cancer-related mortality measured over 10 or more years following the initiation of DBT screening, the interval cancer rate was often used as a surrogate marker. Lowering this rate is assumed to reduce breast cancer-related morbidity and mortality.”
The study’s findings underscore the potential of AI as a valuable second reader in screening mammograms. Dr. Bahl noted, “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.”
Understanding the Study’s Methodology
The research team employed a lesion-specific analysis to avoid overestimating the AI’s sensitivity. This method credits the AI only when it correctly identifies and localizes the exact site of cancer. “In contrast, an exam-level analysis gives AI credit for any positive exam, even if its annotation is incorrect or unrelated to the actual cancer site, which may inflate the algorithm’s sensitivity,” Dr. Bahl explained. “Focusing on lesion-level accuracy provides a more accurate reflection of the AI algorithm’s clinical performance.”
“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 said.
Implications for Clinical Practice
The study’s results suggest that AI may preferentially detect more aggressive or rapidly growing tumors. Cancers detected by the algorithm tended to be larger and more likely to be lymph node-positive. “These findings suggest that AI may preferentially detect more aggressive or rapidly growing tumors, or that it identifies missed cancers that were already advanced at the time of screening,” Dr. Bahl stated.
Among 1,000 patients, the algorithm correctly localized 84.4% of 334 true-positive cancers, correctly categorized 85.9% of 333 true-negative cases, and identified 73.2% of 333 false-positive cases as negative. These statistics highlight the algorithm’s potential to enhance cancer detection and reduce false positives in clinical settings.
Future Directions and Challenges
The integration of AI into DBT screening workflows could significantly enhance cancer detection rates. However, its real-world impact will depend on radiologist adoption and validation across diverse clinical environments. Dr. Bahl emphasized the importance of further research and clinical trials to validate these findings and explore AI’s role in breast cancer screening.
The announcement comes as the healthcare industry increasingly explores AI’s potential to improve diagnostic accuracy and patient outcomes. As AI technology continues to evolve, its application in medical imaging and diagnostics is expected to expand, offering new tools for early cancer detection and improved patient care.
Ultimately, the study represents a promising step forward in the fight against breast cancer, with AI poised to play a crucial role in enhancing screening accuracy and reducing the burden of interval cancers.
