Artificial Intelligence is getting better than humans at detecting breast cancer
Since the early 2000s, AI systems have eliminated more than 1.7 million jobs, and that number will only increase as AI improves. Some research estimates that by 2025, AI will eliminate more than 85 million jobs.
But for all the talk about job security, AI is also proving to be a powerful tool in healthcare—specifically, cancer detection. One recently published study has shown that, remarkably, artificial intelligence was able to detect 20 percent more cancers in imaging scans than radiologists alone.
Published in The Lancet Oncology, the study analyzed the scans of 80,000 Swedish women with a moderate hereditary risk of breast cancer who had undergone a mammogram between April 2021 and July 2022. Half of these scans were read by AI and then a radiologist to double-check the findings. The second group of scans was read by two researchers without the help of AI. (Currently, the standard of care across Europe is to have two radiologists analyze a scan before diagnosing a patient with breast cancer.)
The study showed that the AI group detected cancer in 6 out of every 1,000 scans, while the radiologists detected cancer in 5 per 1,000 scans. In other words, AI found 20 percent more cancers than the highly-trained radiologists.
Scientists have been using MRI images (like the ones pictured here) to train artificial intelligence to detect cancers earlier and with more accuracy. Here, MIT's AI system, MIRAI, looks for patterns in a patient's mammograms to detect breast cancer earlier than ever before. news.mit.edu
But even though the AI was better able to pinpoint cancer on an image, it doesn’t mean radiologists will soon be out of a job. Dr. Laura Heacock, a breast radiologist at NYU, said in an interview with CNN that radiologists do much more than simply screening mammograms, and that even well-trained technology can make errors. “These tools work best when paired with highly-trained radiologists who make the final call on your mammogram. Think of it as a tool like a stethoscope for a cardiologist.”
AI is still an emerging technology, but more and more doctors are using them to detect different cancers. For example, researchers at MIT have developed a program called MIRAI, which looks at patterns in patient mammograms across a series of scans and uses an algorithm to model a patient's risk of developing breast cancer over time. The program was "trained" with more than 200,000 breast imaging scans from Massachusetts General Hospital and has been tested on over 100,000 women in different hospitals across the world. According to MIT, MIRAI "has been shown to be more accurate in predicting the risk for developing breast cancer in the short term (over a 3-year period) compared to traditional tools." It has also been able to detect breast cancer up to five years before a patient receives a diagnosis.
The challenges for cancer-detecting AI tools now is not just accuracy. AI tools are also being challenged to perform consistently well across different ages, races, and breast density profiles, particularly given the increased risks that different women face. For example, Black women are 42 percent more likely than white women to die from breast cancer, despite having nearly the same rates of breast cancer as white women. Recently, an FDA-approved AI device for screening breast cancer has come under fire for wrongly detecting cancer in Black patients significantly more often than white patients.
As AI technology improves, radiologists will be able to accurately scan a more diverse set of patients at a larger volume than ever before, potentially saving more lives than ever.
You read an online article about climate change, then start scanning the comments on Facebook. Right on cue, Seth the Science Denier chimes in with:
The study found that science deniers whose arguments go unchallenged can harm other people's attitudes toward science.
"Humans didn't cause this. Climate is always changing. The earth has always had cycles of warming and cooling—what's happening now isn't new. The idea that humans are causing something that happened long before humans were even around is absurd."
You know he's wrong. You recognize the fallacy in his argument. Do you take the time to engage with him, or write him off and move along?
New research suggests that countering science deniers like Seth is important—not necessarily to change their minds, but to keep them from influencing others.
Looking at Seth's argument, someone without much of a science background might think it makes sense. After all, climate is always changing. The earth has always gone through cycles, even before humans. Without a scientifically sound response, a reader may begin to doubt that human-caused climate change is really a thing.
A study published in Nature found that science deniers whose arguments go unchallenged can harm other people's attitudes toward science. Many people read discussions without actively engaging themselves, and some may not recognize erroneous information when they see it. Without someone to point out how a denier's statements are false or misleading, people are more likely to be influenced by the denier's arguments.
Researchers tested two strategies for countering science denial—by topic (presenting the facts) and by technique (addressing the illogical argument). Rebutting a science denier with facts and pointing out the fallacies in their arguments both had a positive effect on audience attitudes toward legitimate science. A combination of topic and technique rebuttals also had a positive effect.
"In the light of these findings we recommend that advocates for science train in topic and technique rebuttal," the authors wrote. "Both strategies were equally effective in mitigating the influence of science deniers in public debates. Advocates can choose which strategy they prefer, depending on their levels of expertise and confidence."
Who you're really addressing are the lurkers who might be swayed by misinformation if it isn't countered by real science.
So what does that look like? If we were to counter Seth's statements with a topic rebuttal, focusing on facts, it might look something like this:
Yes, climate has always changed due to varying CO2 levels in the atmosphere. Scientists have tracked that data. But they also have data showing that human activity, such as burning fossil fuels, has dramatically increased CO2 levels. Climate change is now happening at a rate that isn't natural and is dangerous for life as we know it.
A technique rebuttal might focus on how Seth is using selective information and leaving out important facts:
Climate has always changed, that's true. But you've omitted important information about why it changes and what's different about the changes we're seeing now.
Ultimately, we could combine the two techniques in something like this:
Climate has always changed, but you've omitted important information about why it changes and what's different about what we're seeing now. Levels of CO2 in the atmosphere are largely what drives natural climate change, but human activity has increased CO2 beyond natural levels. That's making climate change happen faster than it should, with devastating effects for life on Earth.
Remember that the point is not to convince Seth, though it's great if that happens. Who you're really addressing are the lurkers who might be swayed by misinformation if it isn't countered by truth.
It's a wacky world out there, science lovers. Keep on fighting the good fight.
Diagnosed by App: Medical Testing in the Palm of Your Hand
Urinary tract infections aren't life-threatening, but they can be excruciatingly painful and debilitating.
"Overnight, I'd be gripped by this searing pain and I can barely walk," says Ling Koh, a Los Angeles-based bioengineer. But short of going to the ER or urgent care, she'd have to suffer for a few days until she could get in to see her family doctor for an antibiotic prescription.
Smartphones are now able to do on-the-spot diagnostic tests that were previously only able to be performed in a lab.
No longer. Koh, who works for Scanwell Health, was instrumental in the development of the company's smartphone app that is FDA-cleared for urinary tract infection screening. It allows someone to test urine at home using a paper test strip — the same one used by doctors in ERs and labs. The phone app reads a scan card from the test kit that can analyze what's on the strip and then connect the patient to a physician who can make a virtual diagnosis.
Test strips cost $15 for a three-pack and consultation with a doc is about the same as an average co-pay -- $25, and the app matches the quality of clinical laboratory tests, according to the company. Right now, you can get a referral to a telehealth visit with a doctor in California and get a prescription. A national rollout is in the works within the next couple of months.
"It's so easy to use them at home and eliminate the inefficiencies in the process," says Koh. "A telemedicine doctor can look at the test results and prescribe directly to the pharmacy instead of women waiting at home, miserable, and crying in the bathtub."
Scanwell is now involved in an ongoing National Institutes of Health- sponsored study of chronic kidney disease to test a version of the app to identify patients who have the disease, which affects more than 30 million Americans. "Because kidney disease has virtually no symptoms, by the time people realize they're sick, their illness is advanced and they're ready for dialysis," says Koh. "If we can catch it sooner, early intervention can help them avoid kidney failure."
Smartphones have changed society — and now they may change medical care, too. Thanks to the incredible processing capabilities of our smartphones, which come equipped with a camera, access to the internet and are thousands of times faster than the 1960s era NASA computers that ran the Apollo Moon Mission, these pocket-sized powerhouses have become an invaluable tool for managing our health and are even able to do on-the-spot diagnostic tests that were previously only able to be performed in a lab.
This shift to in-home testing is the wave of the future, promising to ease some of the medical care bottlenecks in which patients can have two- to three-week waits to see their family doctors and lift some of the burdens on overworked physicians.
"This is really the democratization of medicine because a lot of the things we used to rely on doctors, hospitals, or labs to do we'll be able to do ourselves," says Dr. Eric Topol, an eminent cardiologist and digital health pioneer at the Scripps Clinic and Research Institute in La Jolla.
But troubling questions remain. Aside from the obvious convenience, are these tests truly as accurate as ones in a doctor's office? And with all this medical information stored and collected by smartphones, will privacy be sacrificed? Will friends, family members, and employers suddenly have access to personal medical information we'd rather keep to ourselves?
The range of what these DIY health care apps can do is mind-boggling, and even more complex tests are on the way.
"I'm really worried about that because we've let our guard down," says Topol. "Data stored on servers is a target for cyber thieves — and data is being breached, hacked, brokered, and sold, and we're complacent."
Still, the apps have come a long way since 2011 when Topol whipped out an experimental smartphone electro-cardiogram that he had been testing on his patients when a fellow passenger on a flight from Washington D.C. was seized with severe chest pains. At 35,000 feet in the air, the app, which uses fingertip sensors to detect heart rate, showed the man was having a heart attack. After an emergency landing, the passenger was rushed to the closest hospital and survived. These days, even the Apple Watch has an FDA-approved app that can monitor your electro-cardiogram readings.
The range of what these DIY health care apps can do is mind-boggling, and even more complex tests are on the way. Phone apps can now monitor sleep quality to detect sleep apnea, blood pressure, weight and temperature. In the future, rapid diagnostic tests for infectious diseases, like flu, Dengue or Zika, and urinalysis will become common.
"There is virtually no limit to the kinds of testing that can be done using a smartphone," says Dr. John Halamka, Executive Director of the Health Technology Exploration Center at Beth Israel Lahey Health. "No one wants to drive to a clinician's office or lab if that same quality testing can be achieved at a lower cost without leaving home."
SkinVision's skin cancer screening tool, for instance, can tell if a suspicious mole is cancerous. Users take three photos, which are then run through the app's algorithm that compares their lesions with more than three million pictures, evaluating such elements as asymmetry, color, and shape, and spits out an assessment within thirty seconds. A team of in-house experts provide a review regardless of whether the mole is high or low risk, and the app encourages users to see their doctors. The Dutch-based company's app has been used by more than a million people globally in the EU, and in New Zealand and Australia, where skin cancer is rampant and early detection can save lives. The company has plans to enter the U.S. market, according to a spokesperson.
Apps like Instant Heart Rate analyze blood flow, which can indicate whether your heart is functioning normally, while uChek examines urine samples for up to 10 markers for conditions like diabetes and urinary tract infections. Some behavioral apps even have sensors that can spot suicide risks if users are less active, indicating they may be suffering from a bout of the blues.
Even more complex tests are in the research pipeline. Apps like ResAppDX could eventually replace x-rays, CT scans, and blood tests in diagnosing severe respiratory infections in kids, while an EU-funded project called i-Prognosis can track a variety of clues — voice changes, facial expressions, hand steadiness — that indicate the onset of Parkinson's disease.
These hand-held testing devices can be especially helpful in developing countries, and there are pilot programs to use smartphone technology to diagnose malaria and HIV infections in remote outposts in Africa.
"In a lot of these places, there's no infrastructure but everyone has a smartphone," says Scanwell's Koh. "We need to leverage the smartphone in a clinically relevant way."
However, patient privacy is an ongoing concern. A 2019 review in the Journal of the American Medical Association conducted by Australian and American researchers looked at three dozen behavioral health apps, mainly for depression and smoking cessation. They found that about 70 percent shared data with third parties, like Facebook and Google, but only one third of them disclosed this in a privacy policy.
"Patients just blindly accept the end user agreements without understanding the implications."
Users need to be vigilant, too. "Patients just blindly accept the end user agreements without understanding the implications," says Hamalka, who is also the Chief Information Officer and Dean for Technology at Harvard Medical School.
And quality control is an issue. Right now, the diagnostic tools currently available have been vetted by the FDA, and overseas companies like Skin Vision have been scrutinized by the U.K.'s National Health Service and the EU. But the danger is that a lot of apps are going to be popping up soon that haven't been properly tested, due to loopholes in the regulations.
"All we want," says Topol, "are rigorous studies to make sure what consumers are using is validated."
[Correction, August 19th, 2019: An earlier version of this story misstated the specifics of SkinVision's service. A team of in-house experts reviews users' submissions, not in-house dermatologists, and the service is not free.]