AI-enhanced breast MRI with a 75% reduced gadolinium dose maintained diagnostic sensitivity comparable to full-dose protocols.
Key Details
- 1Single-center study with 20 female participants having biopsy-proven malignancies or high-risk indications.
- 2Low-dose (0.025 mmol/kg) MRI scans with AI were compared to standard-dose (0.1 mmol/kg) with/without AI.
- 3Diagnostic sensitivity with low-dose + AI was 83.3% (p=1 vs. standard), accuracy was 74.5% vs. 69.5% (standard) and 65.5% (standard+AI).
- 498 lesions detected; 12 were biopsy-proven malignancies; high NPVs (93-94%) across all protocols.
- 5AI-assisted low-dose MRI had the highest PPV (50%) and substantial interreader agreement (kappa 0.65-0.74).
- 6Study was small and lacked long-term outcome data; funded by Bracco.
Why It Matters
Gadolinium retention and cost are growing concerns in breast imaging. This study shows that AI-powered image enhancement could allow safer, lower-dose protocols without compromising cancer detection, marking a significant step for both patient safety and clinical feasibility.
Source
AuntMinnie
Related News
•AuntMinnie
Study: Computer Vision Models Best LLMs in Chest CT Breast Abnormality Detection
Computer vision models (CVMs) surpass large language models (LLMs) in accurately labeling incidental breast abnormalities on chest CT scans.
•Radiology Business
Private Equity Backs AIRS Medical to Expand MRI AI Globally
TA Associates is investing in AIRS Medical to accelerate its global expansion of AI-powered MRI efficiency solutions.
•AuntMinnie
Deep Learning Models Rival Radiologists for Pancreatic Cancer Detection on CT
Deep-learning models achieved comparable or superior accuracy to experienced radiologists in detecting pancreatic cancer on CT scans, especially for small tumors.