In honour of Breast Cancer Awareness Month, we are exploring the important role of AI in enhancing breast cancer detection and diagnosis. We will go deep into the modern world of artificial intelligence to explore its full capacity and impact on healthcare; we are also going to take a glance at how AI is making such great progress in regard to diagnosis-especially in the area of breast cancer diagnosis-where it all comes down to mammogram reading.
But, more importantly, what is AI? Artificial Intelligence basically refers to the replication of human intelligence in machines, by enabling them to perform tasks that ordinarily require human cognition, such as problem-solving and decision-making.
The level of transformation possible with AI in healthcare can be very high, right from refining diagnostic accuracy to personalisation of treatments; the use cases are endless. Imagine data-driven insights shaping medical decisions for more effective treatments and a smoother healthcare process. One such area where AI has started making a profound difference is in the field of breast cancer detection.
The surest way to treat any health-related issue in the breast is early detection, and towards this direction, AI is contributing much by helping the doctors analyse the screening mammograms, thereby increasing the speed and accuracy of the detection of any probable problem.
Mammograms represent that kind of screening program where the majority of patients do not present symptoms like lumps or any other kind of warning signs. People in the UK are invited for screening between the ages of 47 and 50, and this mammogram happens once every three years. AI algorithms are now analysing x-ray images of breast tissue. These algorithms are trained on very subtle patterns in the image that may indicate the presence of breast cancer, such as masses or micro calcifications. They can revolutionise detection. Currently, traditional mammogram interpretation by radiologists is performed. Low-risk mammograms are reviewed by only one radiologist, while higher-risk cases are reviewed by two.
That is where AI becomes crucial: it serves as an assistant for the radiologists by flagging potential concerns so they are more likely to catch anomalies.
In a randomised clinical trial conducted in Sweden, over 80,000 eligible women were considered for screening. They were assigned to either an intervention group in which AI would read their mammogram images, or to a control group in which two radiologists would review the images. Categorisation of mammograms by AI was done either as low or high risk for cancer. High-risk ones would be reviewed by a radiologist. The results were nothing short of astonishing. AI-assisted screening, on the other hand, yielded 6.1 cancers per thousand compared to the traditional human review at 5.1 per thousand. Most impressively, AI reduced radiologists’ work by 44%, saving roughly five months in going through some 40,000 images.
What they would really wish to do afterward is investigate which types of cancers were detected with and without AI assistance. The primary endpoint of the study is the interval cancer rate-a measure of the cancers developing between scheduled screenings that usually have a worse prognosis compared to those detected at screening. This will be done after 100,000 women in the study have had at least a two-year follow-up period. Results will show whether AI-supported screening can further raise the quality and accuracy of screenings. Because it’s beyond just detection, AI can learn and improve continuously, therefore becoming more precise over time. The applications will span mammogram readings, from risk assessment to patient stratification and personalised treatments, extending fully to remote diagnostics via Telehealth.
We are en-route to the future in health care whereby AI continuously undergoes improvements in trying to increase the quality of patient care. However, it is important to note that AI is indeed a potent tool but not a replacement for human expertise. This actually means progress abounds as this artificial intelligence technology coupled with health professionals can cause innovation in driving healthier outcomes.