AI decision support improves radiologists' breast cancer detection accuracy when interpreting screening mammograms, mainly by enhancing focus on suspicious regions.
Key Details
- 1Peer-reviewed study published in Radiology journal (RSNA).
- 212 radiologists read 150 screening mammograms (75 with cancer, 75 without).
- 3AI support increased detection accuracy without affecting average reading time or sensitivity/specificity.
- 4Eye tracking showed radiologists spent more time on actual lesions when aided by AI.
- 5AI region scores influenced radiologist attention—higher scores led to more careful examination.
- 6Ongoing research is exploring the best timing and usage of AI decision support.
Why It Matters
The findings highlight AI's practical benefit in improving radiologists' performance during mammography screening, which can contribute to earlier breast cancer detection. Understanding how AI shapes radiologist workflow and perception is crucial for optimizing its integration into clinical practice.
Source
EurekAlert
Related News
•EurekAlert
AI-Powered OCT Enables Rapid 'Optical Biopsy' for Early Endometrial Cancer Detection
A team at Washington University has developed a catheter-based 3D OCT system with AI to quickly and noninvasively detect early endometrial cancers.
•EurekAlert
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
•EurekAlert
KAIST, MIT, Microsoft Develop Efficient AI Image Upsampling for Robotics
KAIST, MIT, and Microsoft have created 'Upsample Anything,' a training-free AI method to restore high-resolution visual data from compressed images with up to 16x improved GPU memory efficiency.