The research also noted that 22% of the increased cancer detection could be attributed to selection bias, as higher-risk patients were more likely to participate in the program. However, the remaining 21% improvement was credited to AI's role in increasing recall rates for additional imaging. Despite these promising results, the researchers emphasized the need for further investigation through randomized controlled trials to better quantify AI's benefits in mammography screening.
Key takeaways:
- A U.S. breast-screening program using AI-enhanced mammography scans showed a 21% increase in cancer detection among women who opted for these scans.
- DeepHealth, an AI firm owned by RadNet, reported that overall cancer detection rates were 43% higher for women in the program compared to those not enrolled.
- The study indicated that 22% of the increased detection rate was due to selection bias, with higher-risk patients more likely to enroll in the program.
- Researchers are planning randomized controlled trials to better quantify the benefits of AI in mammography screening.