A real-world study found no significant improvement in breast cancer detection or recall rates with an FDA-approved AI mammography tool in a single-reader setting.
Key Details
- 1Study reviewed 24,520 screening mammograms post-AI and 21,630 pre-AI, all using digital breast tomosynthesis (DBT).
- 2Recall rate was 12.3% before and 12.8% after AI implementation (p=0.13).
- 3Cancer detection rate per 1,000 women was 6.7 pre-AI and 7 post-AI (p=0.69), indicating no significant difference.
- 4AI categorized 61.3% as low risk, 34.9% as intermediate, and 3.8% as elevated risk, with cancer detection rates in those groups at 0.5, 8.4, and 98 per 1,000, respectively (p<0.001).
- 5AI missed 15 out of 171 diagnosed cancers, mostly among low- and intermediate-risk cases not flagged by AI.
- 6Authors suggest AI may aid in patient risk stratification, but primary screening metrics did not improve.
Why It Matters
While prior European studies in double-reading environments showed AI could improve detection or reduce workload, this large U.S. study indicates no significant breast cancer screening benefit in single-reader practice. The findings highlight the need for more nuanced evaluation of AI tools in real-world clinical workflows and may guide future U.S. AI adoption and regulatory considerations.
Source
AuntMinnie
Related News
•Radiology Business
Automation Bias: How AI Can Compromise Radiologist Accuracy
AI decision support can induce automation bias, leading radiologists to accept incorrect interpretations and potentially reduce their diagnostic accuracy.
•AuntMinnie
Study Reveals Major Impact of Incorrect AI Suggestions in Mammography Reads
Incorrect AI suggestions, especially false-negatives, significantly reduce reader sensitivity and alter visual search in mammography interpretation.
•Radiology Business
Framework Assesses Real-World Financial Impact of Radiology AI Adoption
A new analysis presents a financial calculator for objectively assessing the return on investment (ROI) of implementing radiology AI solutions.