Pro Health Fitness


Cancer Detection Enhanced: AI Plus Mammograms Show Promise

AI improves mammogram evaluations

United States: According to a recent study, artificial intelligence (AI) can enhance physicians’ evaluations of mammograms by precisely identifying even the smallest breast tumors with fewer unsettling false positive results. Artificial Intelligence is leading towards the growth of healthcare as well and even findings are promising.

Study Findings

Researchers found that compared to physicians evaluating mammograms independently, AI-assisted mammography discovered significantly more breast tumors with a lower false-positive rate. Their findings were published in the journal Radiology on June 4.

Researchers discovered that when AI assisted medical professionals in interpreting breast imaging, about 21% fewer women were required to return for a follow-up mammography.

Impact on Follow-up Exams

According to researcher Andreas Lauritzen, a post-doctoral student at the University of Copenhagen in Denmark, “we believe AI has the potential to improve screening performance.”

Research Methodology

Before artificial intelligence (AI) was developed for this use, about 61,000 Danish women between the ages of 50 and 69 who had mammograms performed between October 2020 and November 2021 had their images reviewed by human radiologists for the study.

The team then hired an AI program to conduct the first analysis of about 58,000 women’s breast screening data between November 2021 and October 2022. Human radiologists then reviewed mammograms that the AI had determined to be normal a second time to validate the program’s findings.

The results indicate that AI-assisted screening found more breast tumors (0.8% versus 0.7%) and had a lower false-positive rate (1.6% versus 2.4%).

Improved Screening Performance

The probabilities that a woman actually had breast cancer following a good result—the positive predictive value of AI-assisted screening—were also greater, 34% versus 23%, the researchers discovered.

AI also aided in the discovery of 45% against 37% more cancers measuring one centimeter or less.

According to researchers, AI support also helped radiologists read less than 33% less, which is crucial for increasing the accessibility of mammograms.

“Population-based screening with mammography lowers the death rate from breast cancer, but it puts a significant burden on radiologists who have to read a lot of mammograms, most of which don’t warrant a patient recall,” Lauritzen stated in a journal news release.

Future Directions

More research, according to the experts, is necessary to assess women’s long-term results and improve the accuracy of the AI for specific women.

According to Lauritzen, “Radiologists usually have access to the women’s prior screening mammograms, but the AI system does not.” “We would like to work on that in the future.”

Leave a Reply

Your email address will not be published. Required fields are marked *