It’s no secret that AI is revolutionising the way we get things done. From automated factories building our cars to intelligent algorithms answering our emails, organisations – both public and private – are increasingly investing in learned machines.
But what about the world of healthcare? Can AI go beyond hype and idealism, and fulfil the promise of utilising the world’s most advanced technology to enhance the way we care? With $47 trillion of global economic impact of chronic disease estimated by 2030, surely AI can play a role in better equipping healthcare systems for the increased reliance on their services.
Joan Van Loon, Belux Enterprise Business Unit Leader of Public, Life Sciences, Telco & Utilities for IBM Belgium, spoke to us about the promise of AI in healthcare and how IBM is working to bring this high potential technology into the world of medicine.
The Age of Data
By 2020, we will have staggering amounts of data at our fingertips – Gartner predicts there will be 25 billion so-called Internet-of-Things devices, such as wearables and connected home appliances, collecting data about our every move. At the same time, we are increasingly living longer – there will be 2 billion people over the age of 60 by 2050. It’s no surprise then, that we’re seeing a shortage in resources to care for the number of people alive: “By 2035, it is estimated there will be a 12.9 million expected global shortage of healthcare workers”, Van Loon explains. With so much data available to help lighten the load on healthcare workers, the cost of not investing in better managing time and focus is potentially catastrophic.Today is gift from god that's why it's called PRESENT Click To Tweet
One example Van Loon provides, of where IBM is leveraging AI in reducing the mental and time load of clinicians, is in the field of cancer. They have been working with Memorial Sloan Kettering to create a cognitive computing system to help physicians make better – and quicker – decisions about cancer therapies for individual patients, pairing the world’s knowledge with theirs.
Speeding up the Journey of a Drug
Van Loon also tells us about the huge data issues in the space of drug development: “It costs, on average, $2billion to develop a new pharmaceutical drug, and less than 10% of drugs currently in development will even make it to market.” More thorough analysis of data – saving both time and money to weed out failed drugs earlier and accelerate the better ones quicker – can be achieved by utilising methods far beyond the capacity of the human brain. An example is IBM’s Quest platform, which was able to find 5 proteins, and hence 5 new targets, for ALS, by using data analytics techniques which surpasses human, and standard computational, ability. AI technologies like this can shave years of trial and error off the search for drugs.
And it’s not just analysis of vast datasets in the lab. With 1 in 10 clinical trials in the cancer space shutting down from lack of participation, the need for better management of supply of and demand for patients is critical; a problem easily solved when you consider it as a data organisation problem. There’s also much data already being collected by devices already in patients today which isn’t investigated and utilised in the search for better treatments. Van Loon explains: “An example is insulin pumps in Diabetes patients, for which IBM partnered with Medtronic. Using our AI platform Watson, we analysed the mountains of information being collected to gain insights on how to find a better – automated – treatment for Diabetes beyond the more manual approach taken nowadays.”
Van Loon also spoke to us about the change in consumer attitudes that is fuelling a change in expectations in the healthcare industry: “We as citizens are looking differently at the world – we want products to be personally applicable to us. We want to connect quicker. The way we interact with hospital doctors has therefore changed.” The intolerance of mass-market, impersonalized approaches, coupled with the demand for choice and expectation of immediate results means that people are increasingly perceiving health care systems as inefficient, slow and ‘not for me’. As a result, many personalized devices utilizing AI are emerging to provide individuals with insights based on their data alone – Amiko’s Respiro smart inhaler being one such example.
But how does the broader healthcare industry realise the potential of AI beyond these specific examples when faced with such complex, far reaching problems? For Van Loon, it’s all about the quality of the data: “If you fill the system with bad information, you get bad outcomes”. It seems being vigilant in collecting better information about patients’ health and organising it in such a way we can truly gain insight, is at the heart of optimising the world of health. With better data, comes better analysis, and the opportunities for utilising intelligent technologies is – arguably – endless.
But What About our Jobs?
Van Loon believes in the importance of treating AI as an augmenting technology, not a replacement for humans. He talks about the potential of AI to predict occurrence and severity of schizophrenia through advanced image analysis; to detect and classify diabetic retinopathy based on retinal scans; and even to help detect heart failure before it’s too late. These are tasks we desperately need to be able to do – tasks that we simply cannot do without machines, as humans simply don’t have the informational processing capacity to do them naturally. As Van Loon put it: “AI helps people make better decisions, using the abilities the computer alone has.”
Some AI technologies are still in their infancy, but the potential for more intelligent, better managed, and less costly healthcare systems is huge when we consider what unique abilities machines can offer humans, as opposed retreating for fear of replacement. With the challenges in healthcare increasing across the board, we’d do well to get on board with AI now, and let the artificial intelligence shed some desperately needed light.