It has become nowhere more apparent than in the healthcare field just how large an impact the AI revolution is having. Applications of automation are expanding rapidly, such as strengthening predictive disease detection, personalized care and virtual health assistants, just to name a few.
Medical imaging is an important step in detecting disease everywhere from broken bones to Alzheimer’s disease. Now, patients can be scanned faster than ever with better quality. AI tools can flag urgent concerns, not as a replacement for radiologists, but as another layer of insight to reduce error and increase accuracy.
The Butterfly Network proved that handheld ultrasound scanners work and even brought medical imaging to less developed countries. Tempus Radiology generates insights from radiology images to diagnose a wide range of conditions. The list goes on.
Another huge sector ripe for innovation is pharmaceuticals and drug discovery. Virtual simulations in a multitude of companies test biological responses across hundreds to thousands of samples and drugs, giving researchers valuable insights and informed decisions faster. Numerion Labs, for instance, a San Francisco AI drug discovery company, has been using structured chemistry models to predict molecular binding and identify safe drug candidates.
As the benefits possible in healthcare are tremendous, so are the risks.
In an Oct. 19 letter, the American Hospital Association (AHA) urged the FDA to hold AI developers as primarily responsible for monitoring the safety of their devices, citing that many hospitals lack the technical information needed to identify flaws.
“Specifically, when it comes to clinical confirmation, providers often lack visibility into an AI model’s underlying weights, making it difficult to determine alignment with real-world patient populations,” wrote the AHA, “Therefore, the agency should clarify device benchmarking and clinical confirmation activities that are the responsibility of the sponsor/device developer.”
Other recommendations include transparency in labeling the extent to which a product is autonomous and setting standards for performance.
“Resource and infrastructure barriers can exacerbate the “digital divide” in certain geographies, where rural and other underserved areas have less access to digital services, including AI tools for clinicians and patients,” the AHA letter continued.
The “digital divide” referenced includes the over two billion people who still lack internet access worldwide, with wealthier countries being better equipped to immediately benefit from AI. Another byproduct of this is algorithmic blind spots where racial minorities may be treated differently due to biased datasets. Digital divides exist in wealthy nations, with lower-income individuals who lack high-speed internet, who while fighting an already existing barrier to e-commerce and digital healthcare, may find themselves further left behind.
At the same time, AI may fill in gaps that make resource allocation, supply chain logistics and other tedious work smoother. This can encourage greater healthcare access by allowing for greater quality patient-clinician time. Uncertainty remains on whether clinical shortcuts serve to save more time or create more work to clean up errors afterwards.
A Harvard trial earlier this year showed that AI systems outperformed human doctors in an emergency high pressure medicinal triage, diagnosing more accurately during potential life and death situations. The system's advantages lay in rapid response answering, diagnosis accuracy and long-term treatment plan development with factors such as reading patients’ visual distress not being tested.
Arjun Manrai, one of the lead authors of the Harvard study, said, “I don’t think our findings mean that AI replaces doctors…I think it does mean that we’re witnessing a really profound change in technology that will reshape medicine.”
Another huge piece to the evolving puzzle is simply how well automation is able to process, hold and manage such vast volumes of data at once. Physician and nurse shortages are chronically worsening amid an aging U.S. population. Burnout from navigating increasingly large patient health records is a cause of worker stress. So, despite concerns of privacy and data quality needing to be addressed, there is optimism that processing data can help.
Eric Cichon, salesman for Medtronic (a large medical manufacturer), works in the endoscopy division selling equipment for GI physicians in hospitals and office spaces.
GI Genius, a product of Medtronic, is the first-to-market computer aided endoscopy module, powered by artificial intelligence.
“The GI genius helps physicians identify possible cancer polyps that could result in colorectal cancer, and we know colorectal cancer is a problem, that’s one of the main diseases that are affecting a lot of people worldwide,” said Cichon, “So it’s catching those, not only catching those but catching them early, right so you wanna catch these polyps, so you can properly treat them and remove them and monitor them.”
The database of such a device is built off of 20 million images, with the AI designed to assist the physician in finding abnormalities as a second set of eyes during routine colonoscopies.
“It’s all of the products now are gonna have some sort of AI capability, it just makes sense if you think about the value of AI. There's no stronger, there’s no stronger need for AI other than healthcare, in my opinion,” said Cichon, “It’s amazing what they’re doing, in the cardiovascular space, I mean just name a sector and it will be affected by AI positively. Telehealth, reading blood samples, I mean everything, it’s gonna be incorporated.”
According to Yahoo Finance, the global artificial intelligence healthcare market is projected to increase from $40.8 billion USD in 2025 to over $861 billion USD in 2034. It can be deduced that skepticism in many ways is a healthy reaction to such huge changes, especially in face of the unknown and the natural capacity for mistakes.
“There’s plenty of reasons to be pessimistic and gloomy about AI, but I think where the biggest benefit is, is healthcare, for sure,” said Cichon, “This will save lives and hopefully will reduce costs as well for patients, and help physicians, help staff treat better and treat more accurately.”
Comments