Using a new artificial intelligence tool to screen for deadly cancers may slash disease rates, a study suggests.
A low-dose x-ray called a mammogram is the gold standard for detecting breast cancer, the most common cancer in American women that strikes over 300,000 every year.
While the method is largely accurate in about 87 percent of cases, roughly one in eight cases may be missed, especially in younger women and those with dense breasts.
However, a new study of more than 100,000 women suggests using AI to help guide mammograms could identify more cancers and reduce diagnoses in the years after.
Women in the study were randomly divided into two groups: one that underwent AI-assisted mammograms and another that had standard screening.
The AI was designed to analyze mammograms and determine if they needed to be read by one or two radiologists. Lower risk cases went into the one-doctor group while higher-risk ones were referred for interpretation by two doctors. It also helped spot highly suspicious areas on the scans that could be cancerous.
Two years after initial scans, the team found a 12 percent decrease in interval breast cancer diagnoses in the group that had AI-assisted scans. Interval cancers are those detected between routine screenings, which often means they developed recently or were missed during initial tests.
The findings suggest that the AI-assisted mammograms reduced the amount of cancers that were missed, leading to fewer diagnoses overall during the follow-up period.
Sarah Citron was diagnosed with breast cancer at 33 after noticing a lump in her armpit. Doctors originally blamed the lump on hormonal changes from having her IUD removed to try for another child
Your browser does not support iframes.
It’s important to note that the study was conducted in Sweden, where it is standard for two doctors to review a woman’s mammogram. The AI, the researchers said, was meant to mark which cases were low risk and only needed one doctor, reducing workload and helping doctors move faster.
The US usually uses just one radiologist to review a mammogram.
Dr Kristina Lång, co-study author and breast radiologist at Lund University in Sweden, said: ‘Widely rolling out AI-supported mammography in breast cancer screening programs could help reduce workload pressures amongst radiologists, as well as helping to detect more cancers at an early stage, including those with aggressive subtypes.
‘However, introducing AI in healthcare must be done cautiously, using tested AI tools and with continuous monitoring in place to ensure we have good data on how AI influences different regional and national screening programs and how that might vary over time.’
AI-assisted mammograms are not standard in the US or Sweden, but the experts suggest the study, along with future experiments, could make the case for the technology becoming more widespread.
A mammogram is the gold standard screening method for breast cancer, though the new study suggests AI may help it be more accurate (stock image)
The findings come as breast cancer surges in young American women. Recent research from the American Cancer Society (ACS) found that from 2004 to 2021, cases in patients ages 20 to 39 rose nearly three percent, a rate more than double the increase seen in women in their 70s.
ACS estimates 326,580 women will be diagnosed with breast cancer in the US this year and that 42,670 will die. US women also have roughly a one in eight chance of developing the cancer, and about one in 10 cases are in women under 45.
The new study, published Thursday in The Lancet, looked at 106,000 women in Sweden between ages 40 and 74, with an average participant age of 55. In the US, regular mammograms are recommended beginning at age 40.
About half of the women included were put into the AI-assisted group while the other half underwent standard mammograms.
The rate of interval cancers in the AI group was 1.5 per 1,000 women compared to 1.7 per 1,000 in the control group, a 12 percent reduction. The AI-supported group also had an 80.5 percent cancer detection sensitivity rate compared to 74 percent for the standard group, an 8.4 percent improvement.
Additionally, the AI group had 16 percent fewer invasive, 21 percent larger and 27 percent fewer aggressive sub-types of breast cancer.
The team noted that this technology may help determine when one versus two doctors need to evaluate a mammogram, saving time and resources.
Savannah Caldwell (pictured above) was diagnosed with stage four breast cancer at 25 after previously been told she was ‘too young’ to have cancer
Jessie Gommers, first study author and PhD student at Radboud University Medical Centre in the Netherlands, said: ‘Our study does not support replacing healthcare professionals with AI as the AI-supported mammography screening still requires at least one human radiologist to perform the screen reading, but with support from AI.
‘However, our results potentially justify using AI to ease the substantial pressure on radiologists’ workloads, enabling these experts to focus on other clinical tasks, which might shorten the waiting times for patients.’
There were several limitations to the research, including only looking at women in one country and using one AI system. There was also no information on race and ethnicity, which can impact cancer rates.
Dr Lång said: ‘Further studies on future screening rounds with this group of women and cost-effectiveness will help us understand the long-term benefits and risks of using AI-supported mammography screening.
‘If they continue to suggest favorable outcomes for AI-supported mammography screening compared with standard screening, there could be a strong case for using AI in widespread mammography screening, especially as we face staff shortages.’
