Artificial Intelligence will save your life one day: here’s why
Posted: 9 January 2017 | By Darcie Thompson-Fields
A man walks into a dermatologist’s practice…
He’s deeply concerned; he has an illness that the doctor doesn’t recognise. After some research online and a discussion with her colleagues at the practice, it turns out that it’s a rare complaint. The doctor then starts looking at possible medications, comparing drug side effects that might react adversely with the patient’s current prescriptions.
But what if the doctor had a powerful resource at her disposal: a repository of medical information and insights? What if she also had access to smart, accurate clinical decision support and the kind of intelligent predictive analytics that could cut down the time spent looking for answers and help her get straight to diagnosing and treating the patient’s problem…what if they had the help of Artificial Intelligence?
Artificial Intelligence (AI) as a concept has actually existed since the ‘80s, but it’s only now that data, processing and storage have become abundant that there’s been resurgence in interest and investment. AI is software capable of “mimicking” cognitive functions that we associate with human minds, such as learning and problem solving.
Judging by the recent media attention, it may seem like AI lives only in the realm of Go-playing robots and driverless cars, but it is also the future of healthcare. And, as scientists look for better ways to diagnose and cure diseases; as hospitals look to lower costs and increase efficiency; advances in healthcare AI are about to make their presence felt.
“AI’s ability to sift through, remember and learn from vast amounts of information will get us closer to dealing with the complexity of factors that impact our health.”
Attending an interesting Techonomy event in Half Moon Bay recently, and being at the world’s largest consumer tech show CES this week, I’m more than ever convinced, that AI has tremendous potential for better health outcomes.
In combination with the Internet of Things (IoT), AI will no doubt go at least part of the way to addressing global health challenges, from supporting an aging population to the growing demand for chronic care. AI’s ability to sift through, remember and learn from vast amounts of information, will help us determine exactly how clinical, socio-economic, behavioural, genetic and lifestyle factors combine to help us conduct a precise diagnosis and predict the best possible course of a personalized therapy. It will get us closer to dealing with the complexity of factors that impact our health.
Not only will this hugely benefit patients and medical professionals by saving them time and helping them focus on the all-important careaspect of their job, but it could also be especially valuable for care providers that can’t afford specialist teams. It could revolutionize and democratize diagnosis and care, leading to “first-time-right” assessments and precision medicine reducing human error, eliminating the repetitive time-intensive tasks that clog up the system.
With so many medical records incomplete, inconsistent or inaccurate, how do we make sure the data is high quality?
The algorithms used will need to be scientifically validated and embedded into daily practice so that care teams, clinicians and patients can use them to make the best possible choices, sometimes in a split second.
Another challenge is how to accumulate and ‘cleanse’ the data required for application of AI. With so many medical records incomplete, inconsistent or inaccurate, how do we make sure the data is high quality?
Then there’s the need for end-to-end security, privacy regulation and for systems and society to keep up with new types of evidence emerging from the data.
There has also been a lot of debate and increased awareness that the potential power of super intelligence is not to be under estimated. I couldn’t agree more with other business and technology leaders that shared concerns and underpinned that AI should be developed with safety in mind and at the right pace so that we control it and ensure it will benefit humanity and every individual on a global level instead of posing a threat. The public conversation that has now been starting is essential for the further development of AI.
A new report has predicted that healthcare providers and consumers will be spending over $6bn every year on artificial intelligence tools by 2021—a tenfold increase from today.
The major IT powerhouses like Google, Apple, Facebook, Amazon, Microsoft, Baidu and IBM are all investing heavily in the technology. Philips is applying AI software in various forms of clinical decision support, ranging from vital signs analysis, image interpretation to genomics.
To better support patients under intensive care, we introduced a new advanced visualization system that transforms data from patient monitors, electronic medical record systems and other sources into actionable insights.
The system provides context-relevant clinical information on each patient, allowing care professionals to make clinical decisions, based on complex information. In critical care settings where every second counts, advances in AI technology like this could be the difference between life and death.
Last month we launched “Illumeo”, an AI-based radiology solution, that automatically records usage patterns and adapts to the way radiologists work to help identify, quantify and compare lesions in multiple studies. This could dramatically improve the speed and precision of diagnosis.
But not all the applications of medical AI will be for hospitals and doctors. Some of the potential is in helping people manage their own health, whether it’s caring for a chronic condition like diabetes or figuring out the best way to control chronic heart failure.
One of the other Techonomy speakers, Dr. Justin Sanchez, a neuroscientist and program manager in DARPA’s Biological Technologies Office, believes AI could even get to the point where we can control things simply by using our mind – there’s already a mind-boggling example of a brain implant successfully controlling prosthetic arms with the power of thought.
Even non-tech brands are getting in on the action – a leading sports clothing manufacturer will also be using AI and access to millions of pieces of performance data to power a “cognitive coaching system” in an app that provides customized advice for fitness.
Today it would be true to say that AI is still in the early stages of deployment (if you’re thinking in terms of human brain power, we’re probably only at the level of an ant at the moment). However, dramatic increases in computing power, the declining cost of storage and large scale collaboration in open source projects, like OpenAI, have catapulted the technology to the frontier of innovation.
While AI is still years away from matching our own intelligence, it’s already better at recognizing objects in images than human beings. Robots will not replace doctors anytime soon, but today’s supercomputers are capable of crunching vast amounts of data and identifying patterns in a way that people never could. What this means is that AI is already augmenting medical specialists and one day soon it will become an indispensable tool for care givers.
While, there are potential dangerous sides to AI when reaching singularity in the future, today’s first applications of AI in medicine is good news for all of us.
About the author
Jeroen Tas is CEO Connected Care & Health Informatics, Member of the Executive Committee at Royal Philips.
The original article was published January 6 and can be viewed in its entirety here.