5 Ways AI Will Be Effective In The Healthcare Industry

How machine learning will innovate the healthcare industry

25Jan
Discuss this further at the Big Data & Analytics in Healthcare Summit

Artificial intelligence is a burgeoning technological revolution that no one can stop talking about, and it's predicted to impact almost every industry out there. But one, perhaps unexpected, industry is discussed with more excitement and potential than most: the healthcare industry. Applying machine learning to help those at their most vulnerable is a very promising, very worthy use for the technology, and it promises to change the industry for the better. Here are 4 of the areas where AI will be at it's most effective within the healthcare industry over the coming years:

Chatbots

Healthcare providers spend an awful lot of money on customer service. And this budget generally goes towards the wages and the associated costs of customer service representatives, who take patient inquiries on the phone, via email or - more recently - through live chat. Yet in an era of burgeoning AI technology, healthcare customer service has been repeatedly identified as an area where machine learning can step in and increase cost-effectiveness. According to predictions by Juniper Research, introducing chatbots powered by AI could save organizations up to $8 billion a year worldwide.

It's not just about cutting costs either, the same report estimates that there will be significant time savings if chatbots are used, around 4 minutes per customer inquiry. This will not only save between $0.50-$0.70 per interaction, but will also save time, and increased efficiency in healthcare means a better service for those at risk. A chatbot is also able to provide 24/7 service, 365 days a year without needing a wage or benefits and, as always, automation is more reliable, cutting out the risk of human error.

Chatbot functionality tends to operate on four or more structures. The first is natural language processing so they can make sense of the user’s demands, followed by knowledge management which allows them to provide an answer. Deep learning helps the chatbot improve its response to each interaction. Sentiment analysis detects the user’s frustration and transfers them to a human if necessary.

For the time being, only the easier customer service tasks can be undertaken by chatbots, as while the technology is still developing they will struggle to handle more complex tasks. It will also be a while before they are able to empathize as humans are able to. Speaking to Healthcare IT News, Khal Rai, an AI expert, and Senior Vice President, product development and operations at SRS Health, stated 'research in the areas of emotional intelligence is happening. But it is not advanced enough at this moment to put the satisfaction of customers on the line.' Once they have been able to achieve optimal emotional intelligence in chatbots, customer service is likely to be an area that becomes completely automated.

Cybersecurity

As technology improves in the sector, health professionals are better able to tend to patients, but with these life-saving advancements comes the increased risk of cyber threats. With the WannaCry attack in May 2017 still fresh in everyone's minds, this is a real concern for those in healthcare. And rightly so. The Identity Theft Resource Center found the US medical and healthcare sector experienced roughly 336 data breaches in 2017 (as of November 29), and it's a growing problem.

Cybersecurity is one area of healthcare in which AI is already proving to be very effective. Machine learning can be applied to sift through the huge amounts of data created by every event or alert, looking for patterns in behavior versus signatures, as well as being able to take into account the multiple data points from a network, to locate and identify the issue. AI can also be used to analyze the effectiveness of current systems and, if necessary, correct any underlying issues.

Perhaps the most effective way AI can be used to improve cybersecurity is through predictive analysis. In an interview with Wired, Darktrace CEO Nicole Eagan addressed her company's approach to AI cybersecurity. 'The big challenge that the whole security industry and the chief security officers have right now is that they’re always chasing yesterday’s attack,' she said. 'That is kind of the mindset the whole industry has—that if you analyze yesterday’s attack on someone else, you can help predict and prevent tomorrow’s attack on you. It’ssflawed because the attackers keep changing the attack vector. Yet, companies have spent so much money on tools predicated on that false premise. Our approach is fundamentally different: This is just learning in real time what’s going on and using AI to recommend actions to take, even if the attack’s never been seen before. That’s the big transition that Darktrace is trying to get folks…to make: to be in the position of planning forward strategically about cyber risk, not reacting to the past.'

Advances in Smart Prosthetics

Thanks to decades of advances in technology, mechanical limbs today are incredibly nimble and authentic-looking, returning some of the lost ability back to their users. But they are not yet as responsive as the real thing, and the trouble lies in getting past conscious thought. 'Capturing the body’s innate ability to just know what to do is something really lacking from all prosthetics today,' says Mike McLoughlin, who manages the prosthetics program at Johns Hopkins in an interview with Wired. This is where AI can step in and provide the 'intuitiveness' traditional prosthetic cannot.

Researchers at Newcastle University have developed a prosthetic that contains a tiny camera that captures photos of the objects in its view. Using the data collected from the photography, AI then determines an action, for example, 'pick up that glass and raise it to my mouth.' Applying machine learning will mean the arm will literally know what it is grabbing. The developers report it is up to 10 times faster than others on the market, and the impact this will have on amputees' life when the technology becomes widespread will no doubt be phenomenal.

AI in mental health

The National Alliance on Mental Illness report that 1 in 5 adults in the US - that's 43.8 million people - experience mental illness in a given year. The same survey reveals that serious mental illness costs America $193.2 billion per year. Machine learning can effectively assist the diagnosis and treatment of these conditions, limiting the impact it now has on individuals' lives and industry.

According to research by IBM, in just five years time what we say and write will be used as indicators of our mental health and physical wellbeing. New cognitive systems, powered by AI technology, will use analysis of individuals speech and writing habits to discover tell-tale signs of developmental disorders mental illness and degenerative neurological diseases, allowing doctors and patients to better predict, monitor and track these conditions.

Medical diagnosis

Machine learning can also be applied to medical diagnosis, and with more reliable results than human doctors currently produce. Researchers at an Oxford hospital in the UK have developed AI technology that can diagnose scans for heart disease and lung cancer. Cardiologists currently use the timing of the heartbeat in scans to tell if there's a problem, but even the best doctors get it wrong in one out of every five cases. The AI system developed at the John Radcliffe hospital can diagnose much more accurately, picking up on details the doctors are unable to notice, and ruling out the risk of patients going home undiagnosed and suffering a heart attack, or unnecessary surgery taking place. They have also developed a similar AI system that looks for signs of lung cancer, detecting nodules on the organs and immediately identifying whether they are cancerous or not. Human doctors cannot tell from sight alone, so various scans are required to see how the nodules develop. With immediate detection, money can be saved, and more importantly time, reducing the risk of fatalities.

Speaking to the BBC, Dr Timor Kadir estimates that using AI to predict lung cancer will allow more than 4,000 cancer patients a year to be diagnosed earlier, greatly increasing their chance of survival. He also believes that the system could save up to £10bn if it was adopted in the US and the European Union. And it's not just companies traditionally in the healthcare industry who are investing in the technology, firms like Google's Deepmind and CareSkore are attempting to make diagnosis simpler by reducing the time between tests and treatments. The Google Deep Mind Health project, for example, intends to use AI to learn the most effective treatment for different patients. Mining of medical records could happen within minutes, meaning an early diagnosis and treatment.

Smart manufacturing small

Read next:

The Business Case For Smart Manufacturing

i