A new study has revealed that the use of an artificial intelligence (AI) tool can enhance breast cancer detection rates by more than 10%. Conducted by the University of Aberdeen in collaboration with NHS Grampian, the research evaluated how AI could assist healthcare professionals in the routine screening of over 10,000 women. The findings indicate that not only does the AI software improve detection, but it also allows for quicker notification of results to patients.
The AI tool, named Mia, was developed by the medical technology firm Kheiron. It is designed to identify subtle and often overlooked areas of concern on mammogram scans that may be missed during manual reviews. The study, published in the Nature Cancer journal on October 3, 2023, found that the tool could increase detection rates by 10.4%. Additionally, it has the potential to lighten the workload for medical staff while expediting the process of notifying women about their screening results.
Yvonne Cook, a participant in the study from Aberdeen, expressed her gratitude for the role AI played in her diagnosis. In what she expected to be a routine mammogram in 2023, Cook opted into the AI research project after seeing a sign in the waiting area. “It didn’t occur to me for a minute to opt out,” she stated. Shortly after her appointment, she received a recall letter indicating that further imaging was needed.
During the follow-up visit, medical professionals informed Cook that the AI analysis had detected a small, Grade 2 tumour that was not visible to the human eye. She described the experience as overwhelming, stating, “I just felt incredibly lucky that I was part of the research programme and that it had been picked up at this early stage.”
Following her diagnosis, Cook began medication to inhibit the tumour’s growth and subsequently underwent surgery. She reflected on the potential consequences had the AI not identified the cancer early. “Had the AI not picked up the small tumour when it did, it could have been discovered at my next routine mammogram three years later,” she noted. This delay might have resulted in a more invasive surgical procedure and possibly the need for chemotherapy, significantly impacting her recovery and quality of life.
Prof. Gerald Lip, the clinical director for breast screening in northeastern Scotland, emphasized the significance of the study’s findings. He stated that the results demonstrate AI’s ability to “effectively support” cancer detection services while simultaneously alleviating some of the workload faced by healthcare professionals. “The bottom line here is without AI, doctors would not have caught these cancers as early,” he remarked.
As the research progresses, the study will expand into further trials exploring the application of AI in breast screening across various sites in the United Kingdom. Prof. Lip acknowledged that integrating AI into clinical practice presents operational challenges, but he believes the findings will contribute significantly to the ongoing dialogue surrounding the use of AI in healthcare.
