Google's AI system for screening breast cancer will soon be included in commercial mammograms. The business stated on Nov. 28 that it has acquired its AI technology from iCAD, a healthcare technology company that delivers breast cancer screening services to universal healthcare facilities throughout the world.

While iCAD's cancer screening services already feature AI-based tactics, it will now integrate Google's technology, which Google has already been testing with academics at Northwestern University. It's a watershed moment for us, according to Greg Corrado, co-founder of Google Brain and chief scientist on Google's AI healthcare team. We're making the transition from scientific work to being able to use their algorithms in the actual world.

In previous research published in Nature in 2020, Google's mammography algorithm outperformed doctors in terms of reporting lesser false alarms while evaluating the pictures. The study included mammograms from over 91,000 women in the United States and the United Kingdom.

Google's approach reduced the false alarms by 6% in the United States, where most women aged 50 to 74 are advised to be examined every two years, and by 1.2% in the United Kingdom, where women aged 50 to 70 are encouraged to be checked every three years. The machine learning system significantly reduced false positive results by 9% in the United States and roughly 3% in the United Kingdom.

Utilizing Google Algorithms in Detecting Cancer

As reported by The Verge, this advantage is now provided economically for the very first moment to the 7,500 mammography sites worldwide that employ iCAD's services, including university health systems. While Corrado did not elaborate on how Google's algorithm differs from others being explored by other academics and firms in the sector, he did say the system uses information from a wide range of photos, including images other than breast tissue, to enhance the computer learning process. As part of the cooperation arrangement, iCAD plus Google will keep working to improve and perfect the technology.

The algorithm is not intended to replace radiologists soon. However, Stacey Stevens, president and Chief executive of iCAD, believes that in Europe it might help ease the pressure on radiologists because several countries (including the United Kingdom) demand two interpretations of a mammogram picture. According to her, iCAD is negotiating with health authorities to obtain the necessary clearance so that the company's AI-based interpretation can someday be one of them. Stevens anticipates that the first product incorporating Google's algorithm will be available in the United States in early 2024, as stated by Google in a press release.

Stevens further predicts that the AI-based approach will make mammography available to more individuals all over the world, especially in low-resource locations that cannot sustain the infrastructure necessary to host mammography picture storage gear. More and more information from mammograms fed into the program, as with any machine-learning system, the stronger it grows at recognizing the slightest variations that separate normal tissue from possibly malignant tissue.

Google Teams with iCAD in designing technology detecting breast cancer through '3D Mammography.'
(Photo : Michael Hanschke / Getty Images)
Medical personnel use a mammogram to examine a woman’s breast for breast cancer. Google Teams with iCAD in designing technology detecting breast cancer through '3D Mammography.' Google Health AI and Cloud technology to be integrated into iCAD’s portfolio with potential to elevate performance and expand access to millions of women worldwide.

ALSO READ: Smart Bra: The Future of Breast Cancer Early Detection Developed

Advanced '3D Mammography' Technology

Women who have mammograms utilizing AI-based technology will have their data sent directly into the algorithm, lacking any identifying information. At the moment, most women undergoing mammograms are likely unaware that an AI system may be working in the background, supplementing the radiologist, because no regulatory organizations have approved a totally AI-based interpretation of mammograms.

However, if additional AI algorithms, such as Google's, enter the market, this may alter. Moreover, radiologists may find themselves debating picture interpretation with patients. Eventually, similar machine-based assessments may be able to detect correlations that human eyes cannot. According to Stevens, iCAD's current AI-based algorithm detects the existence of minute nodules in breast tissue, which experts are beginning to correlate to an increased risk of heart disease. If that link is proven, mammography might be used to measure women's risk of heart disease.

For the time being, adding an AI viewpoint to mammograms may begin to enhance how women's risk of developing is calculated. In the United States, Black women are at greater risk of acquiring more aggressive kinds of breast cancer and are therefore more likely to die from the disease than other women, thus training an AI system to detect the initial indications of these malignancies should lead to improved results.

If an AI computer detects those variations, those women might be sent for extra screening to determine if they are at a higher risk of getting cancer. This might lead to them obtaining treatment sooner, giving them a greater chance of survival. This might also result in less expensive medical treatments, saving time and money for the health system.


RELATED ARTICLE: New Breast Cancer Risk Prediction Tool Can Aid in Earlier Diagnosis, Reduced Mortality Among Black Women

Check out more news and information on Cancer in Science Times.