Breast imaging AI tools are seeing only modest clinical benefits according to a recent radiologist survey.
Key Details
- 1A survey of 215 Society of Breast Imaging members was conducted.
- 247% had implemented breast cancer detection AI; 41.8% had not, 11.2% planned to.
- 3Only 34.7% of AI users reported reductions in recall rates, versus 59.3% of non-users expecting such impact.
- 4Similarly, only 9.1% of AI users reported biopsy rate reductions, compared with 36.4% of non-users expecting reductions.
- 5AI users and non-users had similar perceptions regarding patient outcomes (p = 0.87).
- 6Implementation cost was the top barrier (53.3%), followed by concerns about accuracy/reliability and workflow disruption.
Why It Matters
This study suggests that the clinical impact of breast AI is lagging behind expectations, highlighting the need for improved solutions and more robust evidence for AI adoption in practice. Understanding radiologists' real-world experiences will help guide future development and deployment of imaging AI tools.
Source
AuntMinnie
Related News
•Radiology Business
Multi-Vendor AI Adoption Delivers Efficiency Gains in Private Radiology
A large-scale European study shows that adopting multiple AI tools in private radiology practices leads to significant efficiency improvements.
•Radiology Business
Sweden Deploys AI as Standalone Second Reader for Mammography
Three Stockholm hospitals will adopt AI to fully replace radiologist second reads in breast cancer screening.
•AuntMinnie
AI Model Predicts Immunotherapy-Induced Pneumonitis From CT Scans
MD Anderson researchers developed the CIPHER AI model to predict pneumonitis in lung cancer patients using baseline CT scans, outperforming existing models.