Digital Health: The New Frontier of Healthcare – A Different Lens


– I think your average
person in the street thinks we all share
information appropriately and wonderfully in health care. Electronically, seamlessly,
routinely. We don’t. We have many more
problems from an inability to share information than we do from inappropriate sharing of information. (upbeat music) – There is a fair group of people who think that doctors
will become extinct. (upbeat music) I am not in that group. I think that the role
of doctors will change. – You can’t ever replace this
conversation that you’ll have, with a person, with technology. That human connection in
the space of wellbeing, and with young people
is incredibly important. – The doctor can be subject
to fatigue, tiredness, and he or she can make mistakes. The main objective that
we really want to achieve is to help doctors be more
consistent between each other and to improve their accuracy. – We’ve had a whole debate
around My Health Record, and we had 2 million people who opted out. Why are they expressing concerns? Because they weren’t consulted and a lot of people are suspicious about the harvesting of their data. – We are basically these
days in a sort of a war. An unofficial war when it comes
to electronic information. There’s a whole bunch of bad
people trying to access it, and there’s a whole bunch
of good people paid to and constantly exploring
ways to keep them from it. – Probably one of the
challenges is there’s not a universal description and understanding of what digital health means. We don’t have a frame work necessarily, that’s come from United Nations or The World Health Organization, that clearly defines all
the parameters of the space. – If people can’t agree on the terms as used in different
contexts in different ways, then how are we going to
develop effective policies and how are you going to agree on what’s the best way forward to
protect the public’s interests. – I took quite a while
to come to a very simple, one line definition of it. I’ve defined it here as
the use of digital tools and interventions in
wellness and healthcare. – The way we’re solving
seizure prediction, takes advantage of all different ways of doing digital health. So not just basically
developing machine learning or A.I. algorithms but taking
advantage of cloud computing to make data available
for other researchers around the world. We don’t really know what algorithm’s best and so by linking everything together all these digital tools, we’re able to work towards solving a really
hard problem in epilepsy, which is seizure prediction. – What I’m doing with my team, mainly developing computer
programs which can be integrated with the digital images to help doctors to detect diseases more accurately. – Then in Melbourne we
ran a clinical trial of a seizure prediction device
that had the device implanted for up to three years in duration. And so, with that device we
were able to get long term data, and therefore be able
to get enough seizures from each patient to be able to develop reliable algorithms. The predictions we generate
could be used to activate a brain implant that could
electrically stimulate the brain or deliver drugs to stop
seizures from happening. We’re also working with
Melbourne Uni and Seer Medical and supported by Epilepsy
Foundation of America and The Mayo Clinic. – Studies frequently show
that they cannot agree amongst each other. For example, whether
there’s lung cancer or not. Which is actually very scary
from the patients perspective. – We are planning to develop an app which will identify from
a photograph taken from a handheld device the nutritional content of a plate of food. – Monash Malaysia, we have
some new courses introduced for artificial intelligence. Right now, in the unit that
I’m teaching, Computer Vision, there’s more emphasis and more
material being introduced, in deep learning, due to
deep learning becoming such a popular subject. – The use would be extensive, not just for diabetes or hypertension, or obesity, but for normal people like you and I, because it’s such a
common thing for people to take photographs of their food. – We gave them a little
device, what we call a telemonitoring device, where, what they would do is they would
measure their sugar levels. And what we could do is we
could actually analyze it, and give basic advice on diet, basic advice on medications, whether we need to adjust
medications to patients. We could actually send
them advice that said, “Look, your sugar levels
are just too low. You need to do something. So perhaps, you could see a doctor, you should be adjusting your medications.” We became sort of a middle person, between the doctor and the patient. We could be a game changer. – So Ash is a robot, but it’s powered by artificial intelligence
and machine learning. Ash just basically gives
educational information about mental health and wellbeing. So conversation might
be a young person says, “I’m really struggling today, I feel like I’m losing friends.” And then Ash will respond
with supportive comments, like, “Tell me more. What’s happening?” And if they’re at risk
of hurting themselves, there’s an alert trigger in Ash, so, you know, a caring
person, or a teacher, or the wellbeing support team are alerted, and they can step in and
help that young person. – I see the internet as a dangerous space. It’s one where there’s a lot of optimism, but it’s also one where
there’s a lot of dangers, and where people can be
mislead, deceived, manipulated in various ways. – Careful embedding of the technology in the human-to-human contact, so that, we are not actually causing harm is really an important aspect
of the whole situation. These are very vulnerable populations that we are working with. And great damage could be done, if we don’t deal with them
in a sensitive fashion. If, imagine, in a very
heavy handed fashion we come in and say, “Okay,
we’re going to do away with all of this, and
now on, A.I. programs are going to teach homeless
youth about risks of H.I.V. From now on, people who are
going to remind T.B. patients to take their medicine, and all of that, that’s all going to be this A.I. program that’s going to do all of that.” You can imagine that
this can actually cause tremendous harm to the communities. – You don’t know what it’s
like waking up every morning, and the first thing you
look at is Instagram, and super models doing
all sorts of things, and then you turn to
Tumblr to get some help, to cope, and it’s telling you to self-harm. So something like a chatbot
in a school setting, that could give them that
information automatically, where they don’t need
to turn to the internet to get that support, will
hopefully be quite effective. – We have to always remember
that at the end of the day, successful healthcare, especially,
is about helping people. And there’s not question about the value of the personal relationship in that. Still hear a lot of people
talking and writing like, the machines are completely autonomous, and we’re almost pre-Terminator, and they’re just going
to march in and say, “You do not have a job anymore.” – What has happened is
that now it seems like, computers with good algorithms, can actually interpret these things better than a human being can. And over time they get better and better, because they learn. They keep learning. And as they learn, that learning
remains within the system. And then over a period of time, it becomes extremely powerful. – We are not trying to eliminate
the human in the loop. When we are talking
about H.I.V. prevention, and educating youth about
H.I.V. risks, and so forth, it is a social worker
who’s talking to the youth. On the other hand, where the
A.I. technology is useful, is in identifying which
peer leaders to select in the first place. So there’s places where the technology is actually beneficial, and we should identify those places. – A lot of it’s basic public health. We’ve got a large portion
of the population today, at the global level,
and even in Australia, who live in poverty, or don’t have access to clean water or good shelter. Yet we keep pouring money
into the technological fix. Whereas low-tech, or no-tech solutions are the ones we urgently need. – The machine makes a mistake, then who’s going to be blamed? Will the engineer who developed
the machine be blamed? Or will the company who
developed the machine be blamed? Or will the doctor who is in
charge be blamed for that? – In those individuals
that we actually spoke to, so a lot were really optimistic
about digital health. They really see it as
the future of healthcare. However, some did raise some concerns in terms of data privacy. Just the wearables that we are using, a lot of data gets uploaded, online, but really- they were
wondering what really happens to this data? – What do we do with that data, and does it end up getting
held by the school? And what we’re thinking
is, the wellbeing team, will get access to that, and it will be following the
same sort of confidentiality as they would if they
were having a session with a school counsellor. It’s a big challenge. – The biggest companies in the world, they’re in the business of
monetising personal data. I think people have got a
lot to be concerned about, if they’re handing over their information, their health information, which
is very private information. – Health data is much, at a higher risk. If the risk of my illness
is known, you know, I have 50% chance of
getting a heart attack in the next five years. If that data were to go to
corporate companies, you know, let’s say, insurance companies. They can use that against you. They can use that to say,
“We’re not going to insure you.” – If you have all of Australia
putting in their data, and sharing it, and you derive conclusions
from all of that data, then you know, you can
make conclusions that are actually useful. That’s the benefit of doing it. So for me, the benefits
outweigh the security risks. – This is coming so quick and so fast, I think nobody has got
around to understanding it, and how to manage it, how to regulate it. – I think there’s a bit
of an A.I. gold rush on at the minute. And I think a lot of people
are spending a lot of time and effort in that space, running off in lots of directions, because they think there’s
a lot of money to be made. – That A.I. will be incredibly important, in the future, as we move forward in the digital health space. I don’t think we should be afraid of A.I. and what it can produce. – When have we succeeded
in digital health, is when we stop using the word digital. Where now the healthcare we provide, and the wellness care we
provide is just inherently using I.T. and tools like it
to the right safe extent.

Daniel Yohans

Leave a Reply

Your email address will not be published. Required fields are marked *