TECHNOLOGY NEWS 21 March 2016
Smartphones come stuffed with sensors – now doctors are using them to study people's health on a scale impossible before |
Alex Macro/Corbis
IN
BED, at dinner, on your way to work. Our smartphones are always with us, their
sensors capturing data about our lives. That data is priceless to firms like
Google and their customers, advertisers. But medical researchers are getting in
on the act too. Our phones are turning into tools for doing research in
unprecedented detail and scale.
A
team from Seattle-based non-profit Sage Bionetworks, for instance, is gathering
data from thousands of people with Parkinson’s via an app called mPower. They
built the app using Apple’s ResearchKit software, which lets individuals opt in
to studies through their iPhone.
The
team has not yet published any findings, but earlier this month they revealed
some of their numbers when they made the data openly available in the
interests of accelerating research. A few dozen researchers have already signed
up. “We’re very excited by the response,” says Sage’s Andrew Trister.
The
idea behind the smartphone approach is simple. Our phones are packed with
sensors, including accelerometers, microphones, gyroscopes, cameras and GPS.
They can track everything from how much a person moves to variations in their
speech and gait. As speech, gait or activity alter over time, doctors can infer
changes in an underlying disease. Many other
medical devices, such as blood pressure monitors, can now pipe data
straight to our phones too.
This
means that phones hold ever more detailed records on our health. ResearchKit
lets medical researchers gather that information in an ethical way – by
asking for permission. Apple’s system makes it easy to create apps that take
study participants through a consent process, allowing them to share some or
all of their health information anonymously and securely.
Many
ResearchKit-based apps also ask participants to do other tasks: taking pictures
of their moles in the case of the Mole Mapper, say, or going for a walk and
telling MyHeartCounts
about it later.
The
mPower app asks those in the study to complete tasks such as speaking, tapping
on the phone, walking and balancing. The aim is to find out which tasks are the
best measure of Parkinson’s symptoms, then to help people control them.
One
reason for the excitement about phone-based research is that it is a great way
to recruit large numbers of people quickly and cheaply. Just six months after
launching ResearchKit,
Apple announced that the first apps to use it had already recruited more than
100,000 people between them. But there’s a catch – it may not be easy to
hold on to all the volunteers. More than 10,000 people signed up to use mPower,
but only 500 to 600 people use it every week, Trister says. This is still a
large study for this kind of Parkinson’s research, though.
All
trials have people who drop out, says John Torous, a psychiatrist
at Harvard Medical School. But it appears that when signing
up involves little more than downloading an app, people are less
likely to stick with it. “There’s no one to keep them on track,”
he says.
Pestering
volunteers can cause a high dropout rate, as the creators of SleepHealth, an
app designed to track sleep, may be finding out. “If you’re looking for an app
that will keep pestering you with questions, this is one for you,” says the
first review. A new version of mPower aims to be less annoying.
Privacy
is also a major issue. Anonymising
data does not always protect identity, and data from people’s phones
can reveal things about them even they don’t know. The mPower study shows that
phones could help to diagnose Parkinson’s, for instance, although the team
has no plans do this.
It
is early days for such research, but Torous is optimistic. “There’s so much
real-time data that you can gather,” he says. “I think it may be the wave of
the future for research.”
This
article will appear in print under the headline “A lab in every
pocket”
https://www.newscientist.com/article/2081840-iphone-app-tracks-600-patients-in-biggest-ever-parkinsons-study/?utm_source=NSNS
|
No comments:
Post a Comment