Can A.I Improve our Breast Cancer Screening?
Healthcare is the final AI for Good Frontier
If you want to support my writing, where I depend on community support, you can do so with a premium subscription here.
If there was a legitimate AI for Good story for artificial intelligence in the 2020s, it would be healthcare. A male dominated Venture Capital sector has skewed the impact of A.I. up until now to a profit-centric degree.
Given that’s the kind of world we live in, I’m always looking for more uplifting stories around the emergence of AI. Recently an MIT researcher who survived breast cancer devised a technique that seems to predict many breast cancer cases (WP paywalled).
In the near future, with artificial intelligence technology, medical professionals can quickly and accurately sort through breast MRIs, radiology screening and mammograms in patients with dense breast tissue to eliminate those without cancer. We need A.I to help improve patient-centric medicine because the cost of our healthcare system is skyrocketing as we age as a population.
The story around Regina Barzilay, 51, who studies artificial intelligence at MIT in Cambridge is truly a feel-good story. Too often A.I. somehow seems to benefit speculative male dominated sectors instead of helping society, such as women’s health.
She along with a student protégé have built an AI that seems able to predicted with unprecedented accuracy whether a healthy person will get breast cancer. This could disrupt how we screen and think about the disease. This year (2021), an estimated 281,550 women in the United States will be diagnosed with invasive breast cancer, and 49,290 women will be diagnosed with non-invasive (in situ) breast cancer.
In the 2020s AI will lean to:
Screen mammographs better
More screening more accessible and affordable
Reduce false positive in Breast Cancer screening
Help predict Breast Cancer risk
Reduce reading time of radiologists
Help triage care with better data
Improve accuracy and reduce the need of unnecessary invasive biopsies
On December 21st, 2021 Washington Post reporter Steve Zeitchik spotlighted Prof. Regina Barzilay and graduate student Adam Yala’s work developing a new AI system, called Mirai, that could transform how breast cancer is diagnosed, “an innovation that could seriously disrupt how we think about the disease.”
Zeitchik writes: “Mirai could transform how mammograms are used, open up a whole new world of testing and prevention, allow patients to avoid aggressive treatments and even save the lives of countless people who get breast cancer.”
We need AI to do more positive in society than negative if we are to augment humanity with technology in a way that enables the species to survive longer. In the 21st century this is seriously in doubt.
Think about how the healthcare system works in many parts of the world. High mortality rates for breast cancer patients are often due to late detection, particularly in rural areas where accessing reliable, affordable screening can be challenging.
You can read more about Mira, in a paper in the Journal of Clinical Oncology. There’s another paper that will be published in Nature Medicine. The conclusion is simple and positive:
Mirai, a mammography-based risk model, maintained its accuracy across globally diverse test sets from seven hospitals across five countries. This is the broadest validation to date of an AI-based breast cancer model and suggests that the technology can offer broad and equitable improvements in care.
So does it take women to lead AI into the AI for Good era? That seems increasingly to be the case. The AI for Good movement needs to be more than just a PR device by companies like Microsoft.
Mira could transform how mammograms are used. Mira has implications in how risk-assessment probability will alter the future of healthcare. This impacts the lives of men and non-binary individuals as well. AI as a predictive tool in our healthcare personalized to each person and patient might show the validation of AI in the field.
According to researchers, AI-based triaging systems and risk-prediction analysis could significantly reduce radiologists’ workload and alter exactly what early prevention means. From ultrasounds to mammograms to MRIs AI is taking the future of radiology by storm.
A weird thing happens when these AIs are trained on the data. Sometimes they have an oracular quality that the designers themselves can’t figure out how it works, but it just does. In pragmatic medicine we do what will save the most lives, even if there are moral, ethical and social questions in the use of AI in healthcare.
What we call Cancer is of course 200 separate diseases, that can express itself differently in each person and at each stage of its progression. We need AI to beat Cancer, since it’s so complex. AI can help us personalized medicine to the patient that’s truly patient-centric. This ultimately will also reduce the cost of healthcare and make it more accessible.
What’s exciting is that there are dozens of concurrent studies related to AI’s impact on radiology. A previous study led by New York University (NYU) researchers created an artificial intelligence tool to improve the accuracy of breast cancer imaging. The computer program was trained to identify patterns among thousands of breast ultrasound images to aid physicians in diagnosing.
Breast cancer is the second most common cancer globally. According to the World Health Organization (WHO), in 2020, 2.3 million women were diagnosed with breast cancer. Google’s AI system interprets computed tomography (CT) scans to predict the likelihood of having lung cancer.
Clearly A.I.’s ability to impact healthcare screening and optimized patient-centric care needs to be a priority in the 2020s to improve global healthcare systems and reduce the cost of its lack of affordability. As we reach for 2022, many of AI’s advances in healthcare are still in a study phase and not yet implemented at scale.
If you want to support my writing so I can continue to do so, please sign up here.
I have started ten Newsletters on substack that you may also be interested in, they are:
My Focus on Ten Newsletters for 2022
China Superpowers -
Crypto Bulls and Bears -
Datascience Learning Center -
IPO Times -
Off the Grid -
Basic Income International
Wishing you a safe holiday season and a great forthcoming year of 2022. Thanks again for your support.