If there’s one thing that the future of healthcare can’t seem to escape, it’s the promise and potential of artificial intelligence (AI).
Though the field has been around for quite some time, it wouldn’t be until IBM’s Deep Blue beat Garry Kasparov in a series of seven chess games in 1997 that people would really begin taking machine learning and AI seriously.
After two decades, IBM would be pushing another AI machine into the public eye, this time dubbed Watson, and presented as a game-changer for healthcare. Unfortunately, the hype hasn’t quite lived up to the reality, prompting Chamath Palihapitiya, an influential tech investor who founded the VC firm Social Capital, to call Watson is “a joke” in May 2017 via CNBC.
The good news is that, according to David H. Freedman writing for Technology Review, “most of the criticism of Watson … doesn’t seem rooted in any particular flaw in the technology. Instead, it’s a reaction to IBM’s overly optimistic claims of how far along Watson would be by now. In fact, it still seems likely that Watson Health will be a leader in applying AI to health care’s woes.”
So what’s the holdup? Where does AI currently stand in the healthcare field, and where do we see it going in the future?
Digital assistants and programs with names like Watson, Siri, Cortana, and Alexa have all become commonplace in recent years. The physical embodiments of these software personalities are varied, previously existing only in phones and tablets, but now packaged with Amazon Echo and Google Home devices. These voice-enabled devices themselves are predicted to have an impact on the healthcare industry, with MD Connect’s Jonathan Catley writing:
“In-home, voice-enabled devices open a realm of new possibilities for medical organizations. Voice assistants could help patients book rides to and from a hospital through ride-sharing services, or communicate discharge instructions for post-surgery care once a patient is no longer under the effects of anesthesia or heavy painkillers. This technology also improves things on the clinician side; patients can get answers to common medical questions while doctors collect data remotely, eliminating unnecessary follow-up visits.”
Catley continues, mentioning that the elderly, who aren’t always savvy with keyboards and mice, could benefit greatly from voice-search. “Patients can simply speak to a device as they would to a medical assistant, getting the answers they need with less hassle,” he writes.
He also mentions that “some of the more aspirational uses of this technology are a longer way off,” citing the evolution of HIPAA regulations as a major determining factor for the pace of adoption. Interestingly, however, automation and AI could help to reduce some of the most common HIPAA violations if properly employed.
Automation & Security
The experts at the University of Cincinnati Online write that several HIPAA violations are common among all health organizations, “such as failing to store private health care information properly, failing to obtain written consent from patients and sharing photos of patients on social media.”
These types of violations are many things — irresponsible, inexcusable, and avoidable — but most of all, they are the products of human imperfection. AI and automation promise to do away with human mistakes born of fatigue, inattention, and even malice, by performing wrote tasks such as data entry. More importantly, AI can protect against HIPAA violations via cybersecurity, the lack of which has plagued the healthcare industry for the last couple of years.
Unfortunately, while AI-powered cybersecurity systems are being deployed with some success, they aren’t foolproof. Reg Harnish, founder of GreyCastle Security and a fellow at the National Cybersecurity Institute, told Healthcare IT News that AI-powered security analytics tools often have big vulnerabilities of their own, and that great care should be taken when CISOs and other IT execs are looking for technologies in which to invest.
“The fact that we’ve deployed AI does not mean we can shut the lights off and walk out of the room,” he said. “It still requires care and feeding. In some ways, in addition to the end goal we still have to monitor – ‘Is there an intrusion on my network?’ – we also have to monitor the quality of the decision-making around that question … It’s not robots building robots, it’s humans building robots. We’re a generation away from codes that we can actually trust,” Harnish added. “That said, we’ve got to start somewhere, and the faster we can get to robots building robots, the more predictable those technologies will be.”
While still a work-in-progress, these systems are on their way to being able to protect against some of the worst attacks, including the $55,000 attack on Hancock Health that occurred in January earlier this year.
The Future: Robots in the Theatre
As certain as AI in the healthcare field either seems or actually is, the fact that it’s still a work in progress is an absolute certainty. There’s a lot of hope, and even proof of concept: In 2016, Eliza Strickland wrote about the potential for AI-powered robots working independently in the operating theatre.
“In a robotic surgery breakthrough, a bot stitched up a pig’s small intestines using its own vision, tools, and intelligence to carry out the procedure,” she write. “What’s more, the Smart Tissue Autonomous Robot (STAR) did a better job on the operation than human surgeons who were given the same task.”
We’re a long way from deploying these automatons in actual operating theatres with human patients. The fact is, we haven’t even fully figured out how to put AI behind the wheel of a car yet while ensuring safety of drivers, passengers, and bystanders — but we’re close. GoHealth Urgent Care has partnered with Uber to facilitate what’s being called “mobile health”. They write, in their post, “Benefits of the Latest, Emerging Medical Technologies”:
“Say you have a health condition that renders you unable to drive to medical help. A self-driving car would quite possibly save lives in that scenario. There are also talks of implementing wearable sensors in the cabin – around the seats, over the headrests, and within the dashboard and center-console – to monitor your vitals. This means the car itself can be a point-of-care, versus being just a care delivery method, much like an ambulance with their EMTs onboard.”
Just as self-driving cars could save hundreds of thousands of lives per year, AI in the field of healthcare could do the same — particularly when paired with AI in other fields, as the above example shows.
It’s truly only a matter of time before we put these truly innovative programs into use and action. Until that day when AI is able to save patients, we’re just going to have to wait… patiently.