Changes in brain function during recovery of stroke patients
Charting lifestyle changes from big data sets
1. Charting lifestyle changes from
big data sets
Treatment of chronic diseases consumes about 75
percent of the healthcare resources of industrialised
countries. A pre-emptive approach to disease and
in check.
Changes in behavioural factors like diet, exercise,
sleep, stress and the use of intoxicants can prevent a
large proportion of chronic diseases and reduce the
morbidity of those who are already ill. However, it is
and ensure that they continue over the long term.
Analysed data on what people eat, when they
exercise and how they sleep is needed as a foundation
for solutions and services aimed at healthy behavioural
changes. The information required can exist in large
sets of data collected in various situations.
Access to unique material
“We used two data sets that were unique in their vol-
ume of data. One was based on photographs taken by
190,000 mobile phone users of the meals they were
about to eat. The other came from Firstbeat Tech-
nologies and showed heartbeat variations of 30,000
SalWe - the Enabler of Joint Research in Health and Wellbeing
Researchers have been analysing big data related to diet, physical activity and stress in
the search to understand how healthy living habits vary between individuals and over time.
The goal of this part of the SalWe Mind and Body programme was to uncover effective ways of
2. people,” says Professor Ilkka Korhonen, Tampere Uni-
versity of Technology.
The data sets were ideal for the study because diet,
physical activity and stress are some of the most im-
to obtain such accurate information, especially over
the long term.
The data was analysed by researchers from Tampere
University of Technology, Firstbeat Technologies, the
University of Jyväskylä, the Institute of Occupational
Health and VTT Technical Research Centre. The uni-
versities of Cornell and South Carolina in the United
States also participated in the work.
Peer diet advice reliable
The diet database had been produced by a smart phone
application. It contained about 8 million pictures con-
tributed by users who took photographs of their meals
and sent them to other users for a verdict on whether
they were eating healthily.
“We wanted to see whether individuals were willing
to use an application like this and whether feedback
from other users was reliable. We observed that most
users soon forgot all about the app and only 3 percent
-
mon problem with free, easy-to-install mobile apps.”
“But on the positive side, we found that peer assess-
ments about the healthiness of the food were as ac-
curate as the views of experts.”
The conclusion is that services to promote healthy
lifestyle changes can be based on peer evaluations
rather than requiring input from health professionals.
In this way the services can provide faster feedback,
their cost can be reduced and their social appeal can
be enhanced.
Assessing exercise from pulse variations
“Data from Firstbeat allowed us to analyse how much
the Finns really exercise and how many follow work-
out recommendations in their everyday lives, when
researchers are not looking on.”
The analysed data is valuable in drawing up norms
and control data against which the measured results
of individuals can be compared.
combining it with other available data. The aim is to
throw new light on human morbidity and physical
performance. The data is also being used to evalu-
ate stress and recovery, and to explore links between
physical activity and stress.”
SalWe as a facilitator
“SalWe made it possible for us to begin the study and
to conduct sustained, challenging research work. It
also played a major role in providing us with access to
American big data and the Firstbeat database.”
SalWe - the Enabler of Joint Research in Health and Wellbeing
More information
Ilkka Korhonen
professor
Tampere University of Technology
+358 40 820 1357
06/2014