2023-08-04 14:24:00
Around 2.5 million women have a screening mammogram each year in France. This examination, which makes it possible to diagnose breast cancer, is always reviewed by two radiologists for greater safety.
A Swedish study, conducted by the University of Lund, showed that the use of artificial intelligence (AI) increases the detection of breast tumors compared to standard screening, while making it possible to use a single radiologist.
Breast cancer: 20% more tumors detected with AI
To conduct the study, the researchers randomly divided 80,033 women into two groups: half underwent an AI-assisted screening, while the other half had a standard double screening without the aid of the new technology.
“In our trial, we used AI to identify high-risk breast cancer screening exams, which were double-read by radiologists. The remaining exams were categorized as low-risk and were read by only one radiologist. In screen reading, radiologists used AI as a detection aid, in which it highlighted suspicious findings on images.”says Kristina Lång, researcher and associate professor in diagnostic radiology at Lund University and consultant at Skåne University Hospital, in a communiqué.
According to the data collected, the use of artificial intelligence led to the detection of 20% more cancers compared to the standard examination, without increasing false positives. A false positive occurs when a woman is called back for further testing, but no cancer is found.
Mammography: AI also reduces the workload of radiologists
Besides improving tumor detection, the use of AI in breast cancer screening has also improved working conditions for radiologists. According to work published in the journal The Lancet Oncology in early August, AI image reading reduced their workload by 44%. This means that radiologists can spend more time caring for patients and analyzing results in depth rather than reading screening images.
An important point for researchers since Sweden, like France, is facing a shortage of specialists. For them, AI might thus save precious time for practicing doctors and fight once morest excessively long delays (appointments, results) in medicine.
Artificial intelligence: promising results requiring further testing
Although the results of this study are promising, the researchers stress that more trials are needed to confirm these findings in other conditions. For example, it would be interesting to study the effectiveness of AI with other radiologists or other AI algorithms.
“Screening is complex. The balance between the advantages and disadvantages must always be considered. Just because a screening method finds more cancers does not necessarily mean that it is a better method. “What is important is finding a method that can identify clinically significant cancers at an early stage. However, this must be weighed once morest the harm of false positives and overdiagnosis of indolent cancers. The results of our first analysis show that AI-assisted screening is safe since the cancer detection rate did not decrease despite a substantial reduction in screen reading workload.The planned analysis of interval cancers will show whether AI-assisted screening also leads to a more accurate and efficient screening program”concludes the head of the study.
1691162093
#artificial #intelligence #improves #mammography #screening
