The U.K. Is Looking To A.I. To Improve Healthcare Outcomes

The U.K. Is Looking To A.I. To Improve Healthcare Outcomes

One of the United Kingdom’s largest hospitals, the University College London Hospitals (UCLH), is partnering with the Alan Turing Institute to study whether artificial intelligence can carry out key medical tasks normally conducted by human doctors — from diagnosing cancer on CT scans to determining which patients in the ER room should be addressed first. Without the proper oversight, technical proficiency and external awareness from the AI itself, can we trust automation to truly save our skins?

According to Professor Bryan Williams, director of research at University College London Hospitals NHS Foundation Trust, the plans could revolutionize the outcomes of patient care unlike what we’ve seen in the past, drawing comparisons to other technological milestones like Amazon’s delivery service and Google’s all-powerful search engine.

“It’s going to be a game-changer,” Williams told The Guardian. “You can go on your phone and book an airline ticket, decide what movies you’re going to watch or order a pizza … it’s all about AI.” Clarifying their three-year plan, Williams continued: “On the NHS, we’re nowhere near sophisticated enough. We’re still sending letters out, which is extraordinary.”

There’s no denying it’s out of the ordinary. I mean, only the brightest of society’s tech-nerds truly know how those advanced algorithms can produce life-like capabilities — and that could be a problem during this robotic transition. Around March, an Arizona woman lost her life after being struck by one of Uber’s newly introduced self-driving cars. The experimental vehicle raised concerns about Uber, government oversight and the dangers of complicated autonomous vehicles. Whereas the calls for outright bans on new automation seems borderline unreasonable, it’s perfectly reasonable to question how these machines function — and whether they can be used for good without humans at the helm.

As noted by senior editor for MIT Technology Review, Will Knight, it’s unsettling to know modern artificial intelligence often learn to conduct traditionally human-based tasks from observing us. In the case of the self-driving Uber, paired with motion sensors and a network of artificial neurons deriving a million and one interpretations of data against its own unique mechanical bias, how can we assess an accident? Should you know anything about the Turing test, it’s incredibly difficult to decipher the who, what, when, where and why behind artificial intelligence. We can hear what it says, but can the meaning get lost in translation? We can see what a robot does, but can it articulate the rationale behind its own actions? Can humans know it’s genuine in its response? These fundamentals, already complicated enough in humans, is still quite the mystery among A.I.

Thus, can these key medical decisions be left to the likes of artificial intelligence on their own? It was Guruduth Banavar, former Vice President of Cognitive Computing, who saw the positives in the form of “augmented intelligence,” using these complicated algorithms to enhance the human understanding of medical care, among other major day-to-day tasks. Critical reviews from A.I., able to provide new ways of diagnosing disease, risks of illness and directing resources, could improve the species moving forward. Certain studies suggest artificial doctors, during long-term trials, give patients a 95 percent survival chance compared to the 90 percent among varying human doctors. This would surely help medical staff in the analysis of patients, in conducting reports on health status, and even in performing high-pressure life or death surgeries. However, it’s one thing to beta-test these machines in controlled test environments, it’s another to use them in the real world.

According to officials who spoke with The Guardian, the first project will focus on improving the hospital’s accident and emergency department, focusing on meeting wait time targets that are often used as right-wing talking point against the concept of universal healthcare.

“Our performance this year has fallen short of the four-hour wait, which is no reflection on the dedication and commitment of our staff,” Professor Marcel Levi, UCLH chief executive, told the British newspaper. “[It’s] an indicator of some of the other things in the entire chain concerning the flow of acute patients in and out the hospital that are wrong.”

In March, The Guardian reported that around 76.4% of patients needing “urgent care” could only be treated after four hour waiting times, which just so happens to be the lowest since 2010. There is legitimacy behind using the machines to determine the severity of certain medical issues, however this raises questions regarding professional examination from trained entities, record keeping, assurances machines are functional and the cost of medical care driving up over time.

“Machines will never replace doctors,” Levi said, “but the use of data, expertise and technology can radically change how we manage our services — for the better.”