Utilizing Artificial Intelligence for Early Detection of Breast Cancer

There is a well-known challenge in the medical field regarding the limitations of mammography in detecting all cancers, especially in women with dense breast tissue. This issue has raised concerns among both healthcare providers and patients.

A potential solution involving the use of artificial intelligence (AI) may be on the horizon.

Overview

• An AI tool called AISmartDensity can pinpoint women at high risk of undetected breast cancer after a negative mammogram.
• This AI tool outperforms doctors in identifying a smaller, high-risk group of women who may benefit from additional screenings such as MRI.
• Initial results from the ScreenTrustMRI trial suggest that using AI to guide supplementary MRI screenings significantly enhances the detection of aggressive breast cancers, especially in women with dense breast tissue.

When Breast Cancer Goes Unnoticed by Mammograms

Many women need additional breast cancer screening through magnetic resonance imaging (MRI) to complement mammograms and reduce undetected cases of cancer. However, the cost-effectiveness of MRI screening is hindered by the shortage of qualified staff and the high operational expenses of equipment, limiting its widespread adoption.

There is excitement surrounding a new tool that may help address this challenge...

An Artificial Intelligence (AI) tool has been developed to determine which women should undergo supplementary MRI scans following negative mammograms. This tool identifies women with the highest risk of hidden cancer post-mammography.

The ScreenTrustMRI trial utilized this tool, known as AISmartDensity, to assess each mammogram. Participants with a negative mammogram and a high AI score (top 6.9%) were invited to join the trial. Upon agreeing to participate, women were randomly assigned to receive additional MRI scans or not.

Uncovered Four Times More Cancers

Preliminary findings from a randomized trial set to conclude in August 2025 reveal that the AI tool has detected numerous cancers that would have otherwise gone unnoticed.

The main goal is to identify aggressive cancers that may rapidly progress or have already spread to lymph nodes. This new AI tool has demonstrated remarkable success, nearly quadrupling the detection rate compared to conventional methods, identifying around 64 cancers per 1,000 MRI exams versus the 16-17 cancers detected using traditional approaches.

Detection of More Aggressive Cancers

The utilization of an AI-based scoring system to select a specific group for supplementary MRI scans after negative mammograms has proven to be effective in detecting overlooked cancers, making the process more cost-effective compared to traditional mammography.

Researchers, led by Frederik Strand, MD, PhD, from Karolinska University Hospital in Stockholm, Sweden, reported a four-fold increase in cancer detection rate using the AI tool compared to standard breast density assessments. The majority of additional cancers identified were invasive, some with multiple foci, indicating timely discovery.

It is worth noting that multifocal cancer refers to the presence of multiple tumors in the same area of the breast, typically within the same quadrant.

Addressing Various Diagnostic Challenges

Dr. Strand and colleagues emphasize that while mammography has been beneficial in early cancer detection, many cancers found in screened women are interval cancers, detected prior to the next scheduled screening due to symptoms. The AI tool may facilitate earlier detection by identifying high-risk individuals for screening.

Moreover, MRI remains more accurate than mammography for women with dense breast tissue. The research team suggests that AI-based imaging analysis could streamline the diagnostic process leading to a cancer diagnosis through MRI.

The ScreenTrustMRI study enrolled 59,354 women evaluated with AISmartDensity in their mammograms. Among them, 3,821 qualified for MRI based on a top 6.9% AISmartDensity score, with 1,315 eventually randomized and 559 completing the MRI for analysis. The median age of participants was 56, with a small percentage having a history of breast cancer or a family history of the disease.

Our Conclusion

While this new AI tool does not address the issue of breast compression during mammograms potentially releasing cancer cells, it does present a promising alternative to solely relying on mammography for cancer detection. Thermography could also be a valuable consideration.

Thermography utilizes infrared technology to measure the temperature of the breast's surface, detecting heat patterns and changes in blood flow. Elevated temperatures in specific areas may indicate increased blood flow associated with cancerous tissue. By monitoring temperature patterns over time, doctors can identify potential developments suggestive of cancer.

Summary

The limitations of mammography in detecting cancers, particularly in women with dense breast tissue, have led to the use of additional MRI screening. The development of AISmartDensity, an AI tool, aims to identify individuals at high risk for further MRI screening following negative mammograms. This AI tool has demonstrated significant improvements over traditional methods, resulting in enhanced cancer detection rates. Early findings from the ScreenTrustMRI trial demonstrate that using AISmartDensity enhances the detection of invasive cancers, making supplementary MRI screening more cost-effective.