AI could be a game changer in breast cancer screening, Swedish study raises hopes

Technology has taken another significant leap in healthcare. A large-scale study in Sweden has shown that using artificial intelligence (AI) in breast cancer screening is not only safe but also significantly more effective than traditional methods.

Published in The Lancet in 2023, the findings have drawn global attention from doctors and researchers. Here’s a closer look at what the study revealed and why it could be a game-changer for the future.

AI vs Standard Screening

The study, conducted under Sweden’s national breast cancer screening program, found that AI-assisted mammography detected 20% more cancer cases compared to standard screening.

The trial ran from April 2021 to December 2022, involving over 105,900 women, randomly divided into two groups—one screened using AI and the other with standard double reading by radiologists.

Reduced Workload for Radiologists

AI also significantly reduced the workload for doctors. The study found that radiologists’ screening time dropped by 44% when using AI. Dr. Kristina Lång from Lund University described it as the first randomized controlled trial of its kind.

AI helped detect aggressive cancers at an early stage, improving the chances of timely treatment.

Impact on Interval Cancers

“Interval cancers” are those diagnosed between two scheduled screenings—meaning the previous test was negative, but the disease appears before the next check-up.

• AI group: 1.55 interval cancers per 1,000 women

• Standard group: 1.76 interval cancers per 1,000 women

This shows a 12% reduction in interval cancers with AI, indicating that it caught cases during screening itself. Importantly, both groups had a similar rate of false positives.

Challenges for India

While Sweden’s results are promising, implementing AI-assisted screening in India faces several hurdles:

• Infrastructure gaps: India lacks a nationwide, population-based screening program like Sweden’s.

• Variability in quality: Imaging equipment and quality vary widely between large urban hospitals and rural clinics. Standardization is needed for AI to work effectively.

• Data limitations: Sweden has a robust cancer registry for tracking patients. India currently lacks a similar follow-up system, making it hard to monitor missed cases.

AI-assisted mammography has the potential to transform breast cancer screening, catching more cases while reducing radiologists’ workload. However, adapting it to each country’s healthcare system remains a significant challenge.