A new study published in Radiology paints a promising picture of the role of artificial intelligence (AI) in breast cancer screening. Researchers from the University of Copenhagen and the Capital Region of Denmark’s program collaborated to evaluate the impact of AI implementation. Their findings reveal a significant boost in the number of detected breast cancer cases, along with a welcome reduction in radiologists’ workload.
The study evaluates the deployment of AI in the region over nearly three years, demonstrating that the synergy between expert breast radiologists and the AI system significantly increases the detection of breast cancer cases while reducing the number of false positives.
Every two years, a breast cancer screening is provided to all Danish women aged 50 to 69. Mammograms are provided to those who have already undergone breast cancer surgery up until the age of 79. Overall, tens of thousands of X-rays are now being examined by artificial intelligence every year.
Screening an entire population using mammography results in fewer breast cancer deaths, but also places greater demands on radiologists. In the study, we conclude that artificial intelligence is a great help, and when combined with the assessments of experienced specialists, there are more cases of breast cancer caught in their early stages and fewer false positive results than before AI was introduced.
Andreas Lauritzen, Postdoc and Study First Author, Department of Computer Science, University of Copenhagen
Before artificial intelligence (AI) was developed, radiologists found 70 cases of breast cancer for every 10,000 screenings. AI has made it possible for radiologists to identify 82 women out of 10,000 who have breast cancer, an increase of 12 cases for every 10,000 screenings.
More Small Tumors Detected
In Denmark’s Capital Region, 119,000 women between the ages of 50 and 69 who had mammography screenings were split into two groups, which the researchers compared. Without using AI, the first group was screened in October 2020 and November 2021. Using AI to analyze the photos, the second group was screened between November 2021 and October 2022.
Not only did radiologists and AI identify a higher number of breast cancer cases, but a higher percentage of those cases involved small invasive tumors measuring 1 cm or less. The percentage of small tumors found with AI assistance was 44.93%, up from 36.60% before AI was introduced; this represents a significant benefit for these women’s care.
Both the number of tumors found and their size are crucial factors that really have an impact on these women’s likelihood of being cured.
Dr. Ilse Vejborg, Chief Physician and Head, Capital Mammography Screening Programme, University of Copenhagen
Dr. Vejborg is responsible for the implementation of AI in the screening program.
Help for Busy Radiologists
The workload for breast radiologists who perform assessments has decreased by 33.4% in the Capital Region of Denmark as a result of the use of artificial intelligence. When screening the entire population, radiologists used to double-read thousands of X-rays, the majority of which proved to be normal.
There is a shortage of breast radiology specialists in the Capital Region of Denmark and the rest of the country – as well as around the world. At the same time, the share of the population that is older has gotten larger, meaning that more people need to be screened. As such, our goal was to reduce the workload without compromising quality, which we have succeeded in doing.
Dr. Ilse Vejborg, Chief Physician and Head, Capital Mammography Screening Programme, University of Copenhagen
Two radiologists independently double-read the remaining 30% of mammograms so that the AI system can automatically focus on any suspicious lesions on the X-rays. As a result, at least one board-certified breast radiotherapist evaluates each examination, and breast radiologists always make the final decision.
When two breast radiologists cannot agree on an examination, the radiologists convene for a consensus meeting or consult with a third, more experienced breast radiologist to help reach a decision. Consequently, the final decision-making process is always carried out by a skilled breast radiologist. This is applicable in the event that the two radiologists or the AI cannot agree on their assessment.
Journal Reference:
Vejborg, I., et al. (2024) Early Indicators of the Impact of Using AI in Mammography Screening for Breast Cancer. Radiology. doi.org/10.1148/radiol.232479
Source: