Can artificial intelligence detect more cancers from mammograms

Can artificial intelligence detect more cancers from mammograms
Can artificial intelligence detect more cancers from mammograms

According to a new study, an artificial intelligence (AI) analysis of mammograms discovered more tumors than two breast radiologists working together, without raising false positives and nearly decreasing the radiologists’ effort.

The preliminary findings of the first randomized trial of AI in a national breast cancer screening program, published in the journal Lancet Oncology, suggested that AI-supported screening detected 20% more cancers than routine double reading of mammograms by two breast radiologists.

Must Read: Jeffrey Celavie AI

To ensure excellent sensitivity, European guidelines advocate double reading of screening mammograms. The United States does not have the same criterion, but it is having a lack of breast radiologists, as do many other countries.

Breast cancer is the most frequent cancer in women in the United States, accounting for approximately 30% of all new cancer cases in women each year.

The study found that using AI to diagnose breast cancer is safe, and it supports the technology’s potential to improve screening efficacy and reduce radiologists’ burden. The AI cut the mammography reading workload by 44%.

“Taken together, the evidence suggests that the use of AI could potentially benefit mammography screening by reducing screen reading workload and the number of interval cancers, but randomized trials are needed to assess the efficacy of AI-supported screening,” the study said.

Previous research investigated the use of AI to detect breast cancer in mammography, but they were retrospective, implying that the screens had already been reviewed by radiologists.

According to the study’s co-author, it is too early to begin integrating AI in hospitals right now.

Must Read: The role of Artificial Intelligence in the Metaverse

“These promising interim safety results should be used to inform new trials and program-based evaluations to address many countries’ severe radiologist shortage.” “However, they are not sufficient on their own to confirm that AI is ready for implementation in mammography screening,” Kristina Lng, an associate professor of radiology diagnostics at Sweden’s Lund University, said in a statement.

The study looked at mammograms from over 80,000 women in Sweden between April 2021 and July 2022. Half of the women had their mammograms read by an AI algorithm before being examined by a radiologist, whereas the other half had their mammograms read by two radiologists without using AI.

The AI-screened patients had 244 malignancies detected, compared to 203 tumors in the non-AI group. In both groups, the false positive rate was the same. Overall, AI-assisted screening resulted in a cancer detection rate of six per 1,000 tested women, compared to five per 1,000 for standard double reading without AI, resulting in the discovery of one more cancer for every 1,000 women examined. 

The final findings will examine if AI can minimize the number of interval cancers – instances discovered between screenings — and whether the use of AI in screening is warranted. However, those results are not expected for several years.

“The greatest potential of AI right now is that it could allow radiologists to be less burdened by an excessive amount of reading,” Lng added. “While our AI-supported screening system requires at least one radiologist in charge of detection, it has the potential to eliminate the need for double reading of the majority of mammograms, relieving workload pressures and allowing radiologists to focus on more advanced diagnostics while shortening patient wait times.”

Must Read: China Announces Measurements Regarding Services Used by Generative AI

Leave a Comment