In a recent interview, Manisha Bahl, M.D., discussed new research showing significant associations between AI detection and the histologic grade and lymph node status of breast cancer.
New research presented at the European Society of Breast Imaging (EUSOBI) Annual Meeting suggests that artificial intelligence (AI) for digital breast tomosynthesis (DBT) provides greater than 90 percent detection of breast cancer and may also be significantly associated with histologic grade and lymph node status.
For the retrospective study, researchers examined the use of DBT-based AI (Genius AI® Detection 2.0, Hologic) in a cohort of 599 women (mean age of 66) with a total of 602 screening-detected breast cancer. According to the study, invasive carcinomas accounted for 80.1 percent of the screening-detected breast cancer with the remaining 19.9 percent being ductal carcinoma in situ (DCIS).
The researchers found positive AI scores in 93 percent of the cases. In a recent interview, lead study author Manisha Bahl, M.D., pointed out the correlation between higher AI scores and higher histologic grades. There was also a strong association between higher AI scores and lymph node involvement, according to Dr. Bahl.
“In our cohort, approximately 10 percent of women had positive lymph nodes at (the) time of diagnosis. That is, the breast cancer had metastasized to the ipsilateral axillary lymph nodes, and we found that lymph node positive tumors had significantly higher AI scores than lymph node negative tumors. The average AI score for lymph node positive tumors was 72 out of 100 which was significantly higher than the average AI score for lymph node negative tumors, which was 60,” observed Dr. Bahl, an associate professor of radiology at Harvard Medical School and director of the breast imaging fellowship program at Massachusetts General Hospital.
Noting little in the way of studies examining the use of AI scores as potential biomarkers of breast tumor biology, Dr. Bahl said the current study findings suggest expanded utility of AI beyond adjunctive detection in breast cancer screening pending future research.
“ … Genius AI Detection 2.0 has a high sensitivity for the detection of breast cancer. It can help us detect breast cancers and potentially reduce the false negative rate of screening mammography, which is an important metric. Now, in addition, our current study shows that AI could also serve as an imaging biomarker and provide insights into underlying tumor biology and tumor aggressiveness. Further research is needed to validate our findings and explore the clinical implications of these findings.”
(Editor’s note: For related content, see “What a DBT Screening Study Reveals About False Positives with AI and Radiologist Assessments,” “Study Shows Higher Recall PPV and Lower False-Positive Recall Rate with Combination of DBT and Synthesized Mammography” and "Digital Breast Tomosynthesis Study Assesses Impact of Architectural Distortion on Malignancy Rates")
For more insights from Dr. Bahl, watch the video below.
Stay at the forefront of radiology with the Diagnostic Imaging newsletter, delivering the latest news, clinical insights, and imaging advancements for today’s radiologists.