Artificial intelligence gives hope in detecting breast cancer

Doctors and scientists from Sweden's Lund University found that mammography screening supported by artificial intelligence (AI) almost halved the workload of radiologists during a randomized study of Sweden's national breast cancer screening program

7980 views 2 comment(s)
Illustration, Photo: Shutterstock
Illustration, Photo: Shutterstock
Disclaimer: The translations are mostly done through AI translator and might not be 100% accurate.

Based on the first research results, artificial intelligence could help in better diagnosing breast cancer, especially in countries where there are not enough radiologists.

Doctors and scientists from Sweden's Lund University found that mammography screening supported by artificial intelligence (AI) almost halved the workload of radiologists during a randomized study of Sweden's national breast cancer screening program.

"Mammography screening supported by artificial intelligence resulted in a similar cancer detection rate compared to standard double reading, with a significantly lower workload of screen reading, indicating that the use of artificial intelligence in mammography screening is safe," the scientists point out.

"First results from a cohort of more than 80.000 women in Sweden reveal that AI-assisted screening detected 20 percent more cancers compared to a routine double-read mammogram by two specialized radiologists," the team added.

However, they pointed out that the research results will not be final for several years.

The scientists, whose findings were published in the journal "Lancet Oncology", said that it remains to be determined how the use of artificial intelligence affects the detection of cancer in the period between regular screenings before deciding whether the use of artificial intelligence in mammography examination is justified.

"We still need to understand the implications for patient health outcomes, especially whether combining the expertise of radiologists with artificial intelligence can help detect cancers in the three-year intervals between regular screenings that are often missed by traditional screenings, as well as the cost-effectiveness of the technology," said Kristina Lang from Lund University.

The team believes that the help of artificial intelligence may be needed in countries that do not have enough radiologists.

"There is a shortage of specialist breast radiologists in many countries, including a shortage of around 41 radiologists (eight per cent) in the UK in 2020 and around 50 in Sweden, and it takes more than a decade to train a radiologist capable of interpreting mammograms," they said.

Bonus video: