An artificial intelligence algorithm for dynamic contrast-enhanced breast MRI offered a 93.9 percent AUC for breast cancer detection, and a 92.3 percent sensitivity in BI-RADS 3 cases, according to new research presented at the Society for Breast Imaging (SBI) conference.
Currently being studied for use with dynamic contrast-enhanced magnetic resonance imaging (DCE MRI), an emerging artificial intelligence (AI) algorithm may offer significant benefit in breast cancer detection.
In a multicenter study presented at the Society for Breast Imaging (SBI) conference, researchers assessed the AI algorithm in 8,409 DCE MRI exams. The cohort included 1,305 cases of breast cancer, 319 breasts with benign lesions and 14,988 normal breasts, according to the study.
The study authors found that the AI algorithm provided a pooled 93.9 percent area under the receiver operating characteristic curve (AUC), For BI-RADS 3 cases, the researchers noted the AL algorithm provided an 87 percent AUC, 92.3 percent sensitivity and 81.6 percent specificity.1
In a multicenter study involving over 8,400 patients who had dynamic contrast-enhanced magnetic resonance imaging (DCE MRI) exams, researchers found that an emerging artificial intelligence (AI) algorithm demonstrated a 99.7 percent negative predictive value (NPV) for cases involving normal breasts. (Images courtesy of the Society for Breast Imaging.)
The AI algorithm also had a negative predictive value (NPV) of 99.7 percent for breasts classified as normal breasts, according to the study authors.1
“A first evaluation of an AI system for abbreviated DCE-MRI shows potential for decision support in detecting breast cancer and particularly for accurate identification of very likely normal exams with a consistent performance among different levels of background parenchymal enhancement (BPE),” noted lead study author Koen A.J. Eppenhof, Ph.D., clinical and AI innovation lead with ScreenPoint Medical, and colleagues.
For the 1709 cases in the cohort with documented BPE, the study authors noted an 85.4 percent AUC with cases involving moderate BPE and an 86.6 percent AUC for MRI scans showing marked BPE.1
(Editor’s note: For related content, see “Can Abbreviated Breast MRI Have an Impact in Assessing Post-Neoadjuvant Chemotherapy Response?,” “Abbreviated Breast MRI and Dense Breasts: What Emerging Research Reveals” and “Can AI Automate BPE Assessment of Dense Breasts on MRI?”)
The researchers also found that the AI algorithm performed well across four different MRI vendors with an average AUC of 94.5 percent.1
“AI tools to support radiologists reading breast MRI could improve workflow by improving specificity and reducing reading time,” added Eppenhof and colleagues.
Reference
1. Eppenhof KAJ, Rodriguez-Ruiz A, Boxtel JV, et al. Multi-site validation of a novel AI system for cancer detection in breast MRI. Presented at the Society for Breast Imaging (SBI) conference, April 24-27, 2025, Colorado Springs, Colo. Available at: https://www.sbi-online.org/events/2025-sbi-breast-imaging-symposium . Accessed April 25, 2025.
Considering Breast- and Lesion-Level Assessments with Mammography AI: What New Research Reveals
June 27th 2025While there was a decline of AUC for mammography AI software from breast-level assessments to lesion-level evaluation, the authors of a new study, involving 1,200 women, found that AI offered over a seven percent higher AUC for lesion-level interpretation in comparison to unassisted expert readers.
SNMMI: Botox May Facilitate Relief from Dry Mouth Side Effect of PSMA-Targeted Radiopharmaceuticals
June 25th 2025For patients being treated with radiopharmaceutical agents for metastatic prostate cancer, the combination of botulinum toxin and an anti-nausea patch led to a 30 percent reduction in PSMA uptake in the salivary glands, according to preliminary research findings presented at the SNMMI conference.