Research published this month in the Lancet journal indicates that artificial intelligence (AI) detected cancer at a higher rate on breast cancer screenings compared to standard human-led readings.
“Retrospective studies have shown promising results using artificial intelligence (AI) to improve mammography screening accuracy and reduce screen-reading workload,” said the study authors. “However, to our knowledge, a randomized trial has not yet been conducted.”
So, they sought to see if AI could be used to help reduce the screening workload following mammography tests by recruiting women aged 40–80 years eligible for mammography screening at four Swedish screening sites.
Overall, data from more than 80,000 women was included in the study. Around half were assigned to the AI system and around half were assigned to the more traditional screening method.
“The AI system (Transpara version 1.7.0) provided an examination-based malignancy risk score on a 10-level scale that was used to triage screening examinations to single reading (score 1–9) or double reading (score 10), with AI risk scores (for all examinations) and computer-aided detection marks (for examinations with risk score 8–10) available to the radiologists doing the screen reading,” said the study.
AI-supported screening resulted in 244 screen-detected cancers, 861 recalls, and a total of 46,345 screen readings, while standard screening resulted in 203 screen-detected cancers, 817 recalls, and a total of 83,231 screen readings.
Cancer detection rates were 6.1 per 1,000 screened for the intervention group and 5.1 per 1,000 screened for the control group. Recall rates were 2.2% in the intervention group, compared to 2% in the control group and the false positive rate was 1.5% in both groups.
“The screen-reading workload was reduced by 44.3% using AI,” researchers concluded.
According to Dr. Robert O’Connor of the National Cancer Registry in Ireland, the research is “game changing,” and that is demonstrated that “AI software tool use upfront in breast cancer screening analysis could help rank & prioritise [sic] mammograms for human analysis.”
By the end of 2020, there were 7.8 million women in the world who had been diagnosed with breast cancer during the previous five-year period, making it the most prevalent cancer on the globe, said the World Health Organization. That year, 685,000 women died from the disease – a condition in which breast tissue cells grow out of control – worldwide.
In the U.S., approximately about 240,000 cases of breast cancer are diagnosed in women and about 2,100 in men annually, per the U.S. Centers for Disease Control and Prevention. Around 42,000 women and 500 men in the nation die from the disease every year.
The United States Preventive Services Task Force (USPSTF), an organization made up of doctors and disease experts, recommends that women age 50 to 74 years old and are at average risk for breast cancer get a mammogram every two years. This task force also recommends that women age 40 to 49 years old talk to their doctor or other health care provider about when to start and how often to get mammograms.
Another study published in June also indicated that AI algorithms performed better than the Breast Cancer Surveillance Consortium (BCSC) clinical risk model for predicting breast cancer risk “at 0 to 5” years. When AI and the risk model were combined, it further improved prediction, said that study.
Per the “interpretation” section of the study published this month, “AI-supported mammography screening resulted in a similar cancer detection rate compared with standard double reading, with a substantially lower screen-reading workload.”
Results also indicated that the use of AI in mammography screening is safe, and that research into the subject is expected to continue.
Related
Related
