A recent study in The Lancet Oncology showed that AI found 20 percent more cancers on mammogram screens 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.
Sarah Watts's son Henry was born with spina bifida and can't stand or walk without assistance.
When our son Henry, now six, was diagnosed with spina bifida at his 20-week ultrasound, my husband and I were in shock. It took us more than a few minutes to understand what the doctor was telling us.
When Henry was diagnosed in 2012, postnatal surgery was still the standard of care – but that was about to change.
Neither of us had any family history of birth defects. Our fifteen-month-old daughter, June, was in perfect health.
But more than that, spina bifida – a malformation of the neural tube that eventually becomes the baby's spine – is woefully complex. The defect, the doctor explained, was essentially a hole in Henry's lower spine from which his spinal nerves were protruding – and because they were exposed to my amniotic fluid, those nerves were already permanently damaged. After birth, doctors could push the nerves back into his body and sew up the hole, but he would likely experience some level of paralysis, bladder and bowel dysfunction, and a buildup of cerebrospinal fluid that would require a surgical implant called a shunt to correct. The damage was devastating – and irreversible.
We returned home with June and spent the next few days cycling between disbelief and total despair. But within a week, the maternal-fetal medicine specialist who diagnosed Henry called us up and gave us the first real optimism we had felt in days: There was a new, experimental surgery for spina bifida that was available in just a handful of hospitals around the country. Rather than waiting until birth to repair the baby's defect, some doctors were now trying out a prenatal repair, operating on the baby via c-section, closing the defect, and then keeping the mother on strict bedrest until it was time for the baby to be delivered, just before term.
This new surgery carried risks, he told us – but if it went well, there was a chance Henry wouldn't need a shunt. And because repairing the defect during my pregnancy meant the spinal nerves were exposed for a shorter amount of time, that meant we'd be preventing nerve damage – and less nerve damage meant that there was a chance he'd be able to walk.
Did we want in? the doctor asked.
Had I known more about spina bifida and the history of its treatment, this surgery would have seemed even more miraculous. Not too long ago, the standard of care for babies born with spina bifida was to simply let them die without medical treatment. In fact, it wasn't until the early 1950s that doctors even attempted to surgically repair the baby's defect at all, instead of opting to let the more severe cases die of meningitis from their open wound. (Babies who had closed spina bifida – a spinal defect covered by skin – sometimes survived past infancy, but rarely into adulthood).
But in the 1960s and 1970s, as more doctors started repairing defects and the shunting technology improved, patients with spina bifida began to survive past infancy. When catheterization was introduced, spina bifida patients who had urinary dysfunction, as is common, were able to preserve their renal function into adulthood, and they began living even longer. Within a few decades, spina bifida was no longer considered a death sentence; people were living fuller, happier lives.
When Henry was diagnosed in 2012, postnatal surgery was still the standard of care – but that was about to change. The first major clinical trial for prenatal surgery and spina bifida, called Management of Myelomeningocele (MOMS) had just concluded, and its objective was to see whether repairing the baby's defect in utero would be beneficial. In the trial, doctors assigned eligible women to undergo prenatal surgery in the second trimester of their pregnancies and then followed up with their children throughout the first 30 months of the child's life.
The results were groundbreaking: Not only did the children in the surgery group perform better on motor skills and cognitive tests than did patients in the control group, only 40 percent of patients ended up needing shunts compared to 80 percent of patients who had postnatal surgery. The results were so overwhelmingly positive that the trial was discontinued early (and is now, happily, the medical standard of care). Our doctor relayed this information to us over the phone, breathless, and left my husband and me to make our decision.
After a few days of consideration, and despite the benefits, my husband and I actually ended up opting for the postnatal surgery instead. Prenatal surgery, although miraculous, would have required extensive travel for us, as well as giving birth in a city thousands of miles from home with no one to watch our toddler while my husband worked and I recovered. But other parents I met online throughout our pregnancy did end up choosing prenatal surgery for their children – and the majority of them now walk with little assistance and only a few require shunting.
Even more amazing to me is that now – seven years after Henry's diagnosis, and not quite a decade since the landmark MOMS trial – the standard of care could be about to change yet again.
Regardless of whether they have postnatal or prenatal surgery, most kids with spina bifida still experience some level of paralysis and rely on wheelchairs and walkers to move around. Now, researchers at UC Davis want to augment the fetal surgery with a stem cell treatment, using human placenta-derived mesenchymal stromal cells (PMSCs) and affixing them to a cellular scaffold on the baby's defect, which not only protects the spinal cord from further damage but actually encourages cellular regeneration as well.
The hope is that this treatment will restore gross motor function after the baby is born – and so far, in animal trials, that's exactly what's happening. Fetal sheep, who were induced with spinal cord injuries in utero, were born with complete motor function after receiving prenatal surgery and PMSCs. In 2017, a pair of bulldogs born with spina bifida received the stem cell treatment a few weeks after birth – and two months after surgery, both dogs could run and play freely, whereas before they had dragged their hind legs on the ground behind them. UC Davis researchers hope to bring this treatment into human clinical trials within the next year.
A century ago, a diagnosis of spina bifida meant almost certain death. Today, most children with spina bifida live into adulthood, albeit with significant disabilities. But thanks to research and innovation, it's entirely possible that within my lifetime – and certainly within Henry's – for the first time in human history, the disabilities associated with spina bifida could be a thing of the past.
Dr. Andrew Brooks of RUCDR Infinite Biologics holds up a saliva sample.
The patient tilts back her head and winces as the long swab stick pushes six inches up her nose. The tip twirls around uncomfortably before it's withdrawn.
"Our saliva test can detect the virus in asymptomatic and pre-symptomatic cases."
A gloved and gowned healthcare worker wearing a face shield and mask tells the patient that she will learn whether she is positive for COVID-19 as soon as the lab can process her test.
This is the typical unpleasant scenario for getting a coronavirus test. But times are rapidly changing: Today, for the first time, the U.S. Food and Drug Administration cleared one company to sell saliva collection kits for individuals to use at home.
Scientists at the startup venture, RUCDR Infinite Biologics at Rutgers University in New Jersey, say that saliva testing offers an easier, more useful alternative to the standard nasal swab.
"Our saliva test can detect the virus in asymptomatic and pre-symptomatic cases," said Dr. Andrew Brooks, chief operating officer at RUCDR.
Another venture, Darwin BioSciences in Colorado, has separately developed an innovative method of testing saliva for the coronavirus that causes COVID-19.
Saliva testing can allow earlier detection to identify people who may not know they are contagious, say scientists at both companies. In addition, because patients spit into a tube or cup, saliva testing is safer for healthcare workers than taking swabs. This frees up scarce personal protective equipment (PPE) for use elsewhere. Nasal swabs themselves have been in scarce supply.
Saliva testing, if it becomes widespread, potentially could mean opening society sooner. The more ubiquitous testing becomes across the population, experts say, the more feasible it becomes for public health officials to trace and isolate contacts, especially of asymptomatic cases. Testing early and often will be essential to containing emerging hot spots before a vast outbreak can take root.
Darwin Biosceiences is preparing to seek an FDA Emergency Use Authorization (EUA) this month for its patented "CoVScreen" testing system, which potentially could be available to labs nationally by mid-summer.
Meanwhile, Infinite Biologics will now begin selling kits to consumers for home collection, upon order by a physician. The FDA said that the company's saliva test was as accurate as the nasal swab method used by health care professionals. An FDA summary documenting the company's data reported: "There was 100% positive and negative agreement between the results obtained from testing of saliva and those obtained from nasopharyngeal and oropharyngeal swabs."
The greatest scientific advantage, said Dr. Brooks, is that nasal and oral swabs only collect the surface area where the swab goes, which may not be the place with most viral load. In contrast, the virus occurs throughout a saliva sample, so the test is more trustworthy.
The lab at Rutgers can process 20,000 tests a day, with a 48-hour turnaround. They have 75,000 tests ready to ship now.
The Leap: Detecting Sickness Before You Feel It
"We wanted to create a device that could detect infections before symptoms appeared," explained Nicholas Meyerson, co-founder and CEO of Darwin.
For more than 300 years, he said, "the thermometer was the gold standard for detecting disease because we thought the first sign of illness was a fever. This COVID-19 pandemic has proven that not all pathogens cause a fever. You can be highly contagious without knowing it."
"The question is whether we can scale up fast enough to meet the need. I believe saliva testing can help."
Therefore, Meyerson and co-founder Sara Sawyer from the University of Colorado began to identify RNA biomarkers that can sense when a pathogen first enters a molecule and "sets off alarms." They focused on the nucleic acids concentrated in saliva as the best and easiest place to collect samples for testing.
"The isothermal reaction in saliva takes place at body or room temperature," he said, "so there's no need for complicated testing machinery. The chemical reaction can be read out on a paper strip, like a pregnancy test -- two stripes if you're sick, and one stripe if you're okay."
Before the pandemic, limited but successful human trials were already underway at CU in Boulder and at the CU Anschutz Medical Campus east of Denver. "This was our proof of concept," he said.
Darwin was founded in March and has secured enough venture capital to concentrate protype development on detecting the virus causing COVID-19. So far, said Meyerson, "Everything works."
A small double-blind test of 30 samples at CU produced 100 percent accuracy. "I'm not sure if that will hold true as we go into clinical trials," he said, "but I'm confident we will satisfy all the requirements for at least 95 percent clinical validation."
The specific "CoVStick" test strips will roll out soon, he said: "We hope before the second wave of the pandemic hits."
The broader saliva test-strip product from Darwin, "SickStick," is still one to two years away from deployment by the military and introduction into the consumer drugstore market for home use, said Meyerson. It will affordably and quickly detect a range of viral and bacterial infections.
An illustration of the "CoVStick."
(Darwin Biosciences)
A Potential Game Changer
Society needs widespread testing daily, said George Church, founding core faculty of the Wyss Institute for Biologically Inspired Engineering at Harvard University. Speaking at an online SynBioBeta webinar in April, he urged developing stockpiles of testing kits for home use.
As for any potential of false positives, Church said a much bigger risk is not having enough tests.
"Saliva testing is going to speed up the timeline for opening society a lot," said Meyerson. "People need to self-collect samples at home. A lot more people are going to be willing to spit into a tube than to push a swab six inches up their own nose."
Brooks, of Rutgers, addressed the big picture. "It's critical that we open society as soon as possible to minimize the economic impact of the pandemic. Testing is the surest and safest path. The question is whether we can scale up fast enough to meet the need. I believe saliva testing can help."