In the future, will we see medical robots walking our hospital wards, working in theatres and putting in long shifts in GP surgeries?
Speculation on how artificial intelligence (AI) will change society often enters the realm of science fiction. However, the significant strides made in AI technology over the past decade have led to a concentration of some of the globe’s greatest minds on the topic.
At the more pessimistic end of the scale is world-renowned scientist Prof Stephen Hawking, who told the BBC in 2014 that “the development of full artificial intelligence could spell the end of the human race”.
Others such as Mr Ali Parsa, the CEO of digital healthcare company Babylon, heralds the new AI age as a positive time for health. “Machines are able to recall every known disease perfectly when examining symptoms,” he told <em>MIT Technology Review</em> in March 2016. “And unlike human doctors, they don’t have confirmation bias.”
Whatever the long-term impact, it is clear that AI in health is already here. Oncologists at the Memorial Sloan-Kettering Cancer Centre in New York, for example, use the IBM Watson system.
“IBM’s software draws from 600,000 medical evidence reports, 1.5 million patient records and clinical trials, and two million pages of text from medical journals to help doctors develop treatment plans tailored to patients’ individual symptoms, genetics and histories,” according to the <em>Review</em>.
But if this all sounds a little speculative, it is important to note that it’s not only in the US where AI breakthroughs in health have taken place. In Ireland, change is afoot.
“All new technology and innovation is considered under the eHealth Ireland banner and within the <em>Knowledge and Information Plan</em>,” Mr Richard Corbridge, HSE Chief Information Officer and CEO of eHealth Ireland, told the <strong><em>Medical Independent</em></strong> (<strong><em>MI</em></strong>).
<div style=”background: #e8edf0; padding: 10px 15px; margin-bottom: 15px;”>
<h4 class=”subheadMIstyles”>‘Virtual humans’ in mental health</h4>
AI innovation will change all areas of health, but mental healthcare could be one of the sectors altered most profoundly.
“AI is, and will increasingly, bring benefits to the field of mental healthcare,” said Dr David D Luxton, Editor of <em>Artificial Intelligence in Behavioural and Mental Health Care</em> (Academic Press, 2015).
“AI techniques, such as machine learning methods, are used to analyse data about patients and provide information to both healthcare professionals and to patients.
“Expert systems, for example, are a clinical decision-making tool that can help identify emerging risks, such as contraindications of medications or increases in symptoms that confer risk. Data is also collected on intelligent mobile devices, thus increasing the amount and types of data that are available.
“AI methods (ie, machine learning) can identify patterns in data that humans simply cannot do as efficiently or at all.”
But there are also ‘virtual humans’ on the mental health horizon, predicts Dr Luxton.
“Virtual intelligent agents, such as virtual humans, are another example. Virtual human (computer-generated characters) can, in a sense, ‘stand in’ for human care providers.
“They have been developed as coaches — to provide supportive information to care seekers and as simulated patients for training purposes. These systems are constantly improving with advances in computing power and AI, such as language processing and effective computing capabilities — understanding and exhibiting emotions.
“AI applications in robotics and robots are another booming area. Robots are increasingly being used for therapeutic services, such as providing companionship and assisting with cognitive tasks for persons with dementia or following stroke.
“There are a number of benefits of AI applied in mental healthcare. AI technologies can automate many functions such as scanning for clinical information, writing chart notes, etc. [They] therefore free-up human healthcare professionals from some tasks to allow them to focus on other things. Virtual intelligent agent systems can also help to fill gaps in care, thus providing more access to healthcare services.
“People can seek self-care from engaging virtual human therapeutic coaches and then be escalated to more intensive care with human care providers, if needed.”
Dr Luxton said it is impossible to deny the significant changes that AI will bring to mental healthcare. He believes clinicians will have an opportunity to influence these changes.
“I think that AI will most definitely alter the roles of mental healthcare professionals. It is possible, in the future, that virtual humans, for example, will replace some functions currently done by human healthcare professionals,” he said.
“I also think that AI increases opportunities for healthcare professionals. Clinicians are needed to help design these systems, test them and ensure that best practices, both clinical and ethical, are put to use.”
</div>
“We have a small number of AI considerations at the moment. A ‘chat-bot’ solution for patients who suffer from bipolar disorder is in development.
“Medxnote, an Irish healthcare provider, is preparing a solution for a small number of pathfinder hospitals that will deliver elements of an AI solution into their current handover capability.
“Of similar interest, and perhaps moving more quickly for Ireland, is the augmented reality (AR) and virtual reality (VR) fields. Here, we are seeing work underway to deliver AR environments for training and development for clinical and emergency responders staff and VR for patients in hospital with a long-term condition who want to stay in touch with their home environments.”
But if AI is with us for the long-term — and it looks like it most certainly is — there are deeper issues that doctors, patients and policy-makers will have to confront.
“There is a list of ethical issues associated with the use of AI in healthcare,” said Dr David D Luxton, Affiliate Associate Professor at the University of Washington School of Medicine in Seattle, US.
Dr Luxton, a clinical psychologist, is the editor of the recent book <em>Artificial Intelligence in Behavioural and Mental Health Care</em> (Academic Press).
“I’m most concerned about who will own and control the technology,” Dr Luxton told <strong><em>MI</em></strong>.
“Will mental healthcare be under the control of large technology companies such as Google? I’m also concerned about the increasing privacy risks that come with this. Who will control all of the health and behavioural data collected on people and what will they do with it?
“Also, will care-seekers be forced to use virtual professionals in the future and not have the choice to speak to a real human therapist?”
There are also concerns that AI, and other new technologies in health, could help widen the quality of services provided to public and private patients. Last month, Dublin GP and former IMO President Dr Ray Walley told <strong><em>MI</em></strong> he was worried that new technology like telemedicine could see the public healthcare system left behind.
“There is no reason why it needs big business to do this,” said Dr Walley of telemedicine. “These apps can be developed by the State, whereby the State is in ownership of all this rather than having to rely on private developments.”
Dr Walley warned that private companies could end up dominating the telemedicine market and “hold the State potentially to ransom”.
With the rise of AI across the labour market, there is some apprehension that soon, humans ‘need not apply’ for a growing number of professions.
Last year, Chairman of the Pensions Authority Mr David Begg told the Joint Oireachtas Committee on Education and Social Protection that “projections about the impact of artificial intelligence are quite worrying”.
Mr Begg was closely involved with labour issues when he was General Secretary of the Irish Congress of Trade Unions between 2001 and 2015.
“A recently-published report by the World Economic Forum suggests that increased automation and artificial intelligence in the workforce will lead to the loss of 7.1 million jobs over the next five years in 15 leading economies, while helping to create just two million jobs over the same period,” Mr Begg told the Committee.
The World Economic Forum report cited by Mr Begg is far from the only study published in recent years that predicts significant job losses across many sectors because of AI automation. For example, just contemplate the impact mass-produced, self-driving automobiles may have on the number of people employed in the public transport sector, freight transport, taxi and delivery services.
Of course, medicine is dissimilar to many other professions. Nobody is predicting that AI robots will soon be donning their white coats or scrubs, signing-up for IMO, NAGP or IHCA memberships and replacing living and breathing surgeons, physicians and GPs.
For now, AI is regarded by many as a positive addition to a clinician’s tool-set and to the general health system. It is seen as an aid rather than a replacement for doctors.
However, the international experts that <strong><em>MI</em></strong> contacted do predict that the future impact of AI on the doctor-patient relationship, in particular, will be profound.
<div style=”background: #e8edf0; padding: 10px 15px; margin-bottom: 15px;”> <p class=”subheadMIstyles”>‘Humans need not apply’
The upcoming spring exhibition at the Science Gallery in Trinity College Dublin promises to “interrogate the supposedly seismic changes that AI is foisting on society”.
Artists will take on the automated future with the exhibition called ‘Humans Need Not Apply’, which opens on Friday 10 February and runs until 14 May.
“In an automated world, is it nearly time to put humans out to pasture?” asks the exhibition’s guide.
“Does the future resemble a leisure-time utopia or a robot-tended ‘human zoo’? Will the notion of work become a thing of the past if machines really can do everything better, faster and for longer?”
The exhibition will include current demonstrations of machine learning and AI capability and creativity. The works presented in this exhibition explore the idea of computers creating culture, from algorithm-driven artworks to responsive robots and animatronic desk lamps.
One of the exhibition’s highlights include: “<em>Ad infinitum</em> — A parasitical entity that lives off human energy. It attaches electrodes onto visitors and harvests their kinetic energy by electrically persuading them to move their muscles. The only way a visitor can be freed is by seducing another visitor to sit down and take their place.”
<em>The full list of exhibits, including those with a particular medical focus, is available at </em>www.dublin.sciencegallery.com<em> from 9 February</em>.
</div>
Dr Madis Tiik is a GP in Estonia and is the former CEO of the Estonia eHealth Foundation (EEHF). He visited Ireland last September to address the eHealth Summit in Dublin.
“Yes, AI will radically alter the role of the doctor,” Dr Tiik told <strong><em>MI</em></strong>.
“In my opinion, 80 per cent of the health-related problems, questions and concerns can be solved with the help of AI. This means that the hierarchic healthcare system does not fit anymore to the changing environment where the first contact is made with an AI-based system and doctors give second opinions.
“The role of doctors will be more like advisers. We have to keep in mind also the fact that thanks to mobile technology, there will be more customers. So the most difficult and complicated cases will be solved by the doctors, where the rest of the problems will be solved by the AI.”
Dr Tiik envisions that, at first, AI will impact doctors, but in time patients will also see their engagement with the health system change significantly.
“Soon, the focus of AI will change from professional users towards regular people,” he said.
“It will help them to understand health-related issues, guide them through the medical terminology and give them tools to become more healthy and proactive.”
<h3 class=”subheadMIstyles”>US view</h3>
Although Ireland is starting to witness some AI developments, it is in the US where the technology is most advanced and thus where medical experts have given the most detailed consideration to the hopes and concerns created by medical AI.
Prof Robert M Wachter is Professor and Chair of the Department of Medicine at the University of California, San Francisco, US, and author of 2015’s <em>The Digital Doctor: Hope, Hype, and Harm at the Dawn of Medicine’s Computer Age</em> (McGraw-Hill Education).
“The greatest benefit will be to rationalise the practice of medicine,” Prof Wachter told <strong><em>MI</em></strong>.
“To make evidence available to guide decisions by doctors, nurses and patients, and to support a learning health system by using our data to continuously improve.”
However, the California-based physician also says there will be significant challenges posed by the growing utilisation of AI.
“The biggest ones will be, one, what happens when AI helps lesser-trained people do things that they are really not qualified to do?”
<blockquote> <div>
‘Who will control all of the health and behavioral data collected on people and what will they do with it?’
</div> </blockquote>
But even with these challenges, Prof Wachter believes AI will radically alter the central role currently fulfilled by doctors.
“Yes, certainly in the long-term,” he said, predicting significant change. “There are lots of things that doctors, nurses, etc, do that will (and should) be taken over by better technology.
“It’s hard to see how tech will yield productivity gains if it doesn’t.”
Prof Eric Topol is Professor of Genomics at The Scripps Research Institute (TSRI) in La Jolla, California, US.
The cardiologist is also author of<em> The Creative Destruction of Medicine: How the Digital Revolution Will Create Better Health Care</em> (Basic Books).
Prof Topol told <strong><em>MI</em></strong> that he believes the greatest benefits that AI will bring to health will be in the use of patient information, “processing patient-generated data such as sensors, smartphone images, eg, skin lesions, and labs along with lifestyle and environmental data”.
But like many other experts in the field, he does see challenges, including “privacy and security matters”. In the end, the greatest challenge that AI will bring, according to Prof Topol, is that it will lead to the “rebooting [of] the patient-doctor relationship”.
He considers that in the long-term, AI will “radically reconfigure” the role currently fulfilled by doctors in healthcare. “In the mid-term, data added to AI will be empowering consumers, patients and individuals,” he said.
<h3 class=”subheadMIstyles”>Health account</h3>
Back in Europe, experts like Dr Tiik have also considered challenges and possible benefits to health.
“The biggest challenge is to maintain the secure environment with high trust which is under the control of citizens,” said Dr Tiik.
“For that, we need IHAN — an international health account number, like IBAN in banking. Who can we trust to keep our health data and where can a person open a health account? At the moment, there are some examples (Australia, Sweden) where the state has opened the health accounts for all of the citizens.
“In Australia, the state validates the health account providers and everyone has a choice to choose their account holder. In Sweden, the state chooses one provider and opens the account for all the citizens.
“Probably, we also see other candidates for account holders in the future.
“When the health account is established, all AI services can be implemented directly to the consumer. It means that we can solve more and more health-related problems.”
If America is at the forefront of AI development, Irish experts are also looking into the possibility of increasing use of AI in health, including the HSE.
“AI remains to be explored in health with any great certainty,” said Mr Richard Corbridge.
“Basic capabilities that we have seen suggested and are looking at for the Irish healthcare system will enable small AI capability to be used on a mobile device.
“[It] will create a patient-facing healthcare assistant that learns from your behaviours and with other data, can provide assistance that could stop specific episodes from happening or could alert your care provider on behalf of you.”
When asked about any particular challenge, ethical or otherwise, that healthcare faces with the use of AI, Mr Corbridge said it “comes with a whole new selection of engagement challenges”.
“Similar to the autonomous car, ethical decisions need to be reached on who is responsible if the AI makes a decision that has a negative impact.
“As with all digital health solutions that eHealth Ireland is looking at, the patient and clinician need to be at the centre of the project.
“We know of the amazing case studies from IBM’s Watson in the delivery of care to cancer, but for Ireland to avail of this requires a great deal of clinical time. We need to try to get this into the priorities but it is not yet there.”
