=Paper=
{{Paper
|id=Vol-1815/paper17
|storemode=property
|title=Challenges for the Similarity-Based Comparison of Human Physical Activities Using Time Series Data
|pdfUrl=https://ceur-ws.org/Vol-1815/paper17.pdf
|volume=Vol-1815
|authors=Tomasz Szczepanski,Kerstin Bach,Agnar Aamodt
|dblpUrl=https://dblp.org/rec/conf/iccbr/SzczepanskiBA16
}}
==Challenges for the Similarity-Based Comparison of Human Physical Activities Using Time Series Data==
https://ceur-ws.org/Vol-1815/paper17.pdf
173
Challenges for the Similarity-Based Comparison
of Human Physical Activities Using Time Series
Data
Tomasz Szczepanski1 , Kerstin Bach1 and Agnar Aamodt1
Department of Computer and Information Science
Norwegian University of Science and Technology, Trondheim, Norway
http://www.idi.ntnu.no
Abstract. In this position paper we present various aspects of compar-
ing human physical activities using time series data within the self-
BACK project. The goal of the project is to develop a decision support
system for patients suffering from non-specific low back pain. The system
will give users advice in form of a self-management plan that is based
on self-reported physical and psychological symptoms as well as activ-
ity stream data collected by a wristband. Here, we discuss the activity
stream representation difficulties as well as various challenges that arise
when comparing the resulting activity streams.
Keywords: Case-Based Reasoning, Time Series Representation, Data
Streams, Similarity Assessment
1 Introduction
selfBACK [2] is a research project developing a decision support system that is
aimed towards helping patients to facilitate, improve and reinforce self-management
of non-specific low back pain. The selfBACK system will constitute a data-
driven, predictive decision support system that uses the Case-Based Reasoning
(CBR) methodology to capture and reuse patient cases in order to suggest the
most suitable activity goals and plans tailored for an individual patient. This
will be based on data from two different types of sources. One is a questionnaire
presented to the patient at suitable intervals in order to capture self-reported
general information and progress of symptoms. The other is a stream of activity
data collected using a wristband. The incoming data will be analyzed to classify
the patients current state and recent activities, and matched against past cases
in order to derive follow-up advices to the patient.
Two of the many challenges during matching and similarity assessment when
using the CBR methodology, is developing a suitable abstraction of the wristband
activity stream and an adequate comparison algorithm for such streams. The
abstractions and the comparison algorithm must catch the important aspects
and compare these in a meaningful way relevant in a non-specific low back pain
domain.
Copyright © 2016 for this paper by its authors. Copying permitted for private and academic purposes.
In Proceedings of the ICCBR 2016 Workshops. Atlanta, Georgia, United States of America
174
We will in the next section shortly present the different aspects and challenges
that need to be considered in order to develop such a comparison. Specifically,
we will focus on the challenges arising when comparing the activity streams of
patients.
2 Challenges
The main goal of the selfBACK project is to give non-specific low back pain
patients personalized advice based on their reported symptoms and activity pat-
terns collected by a wristband. Three major challenges arise when comparing the
activity streams:
1. The activity stream needs to contain the knowledge needed and used for the
treatment of non-specific low back pain.
2. The comparison of such streams must be made in a way that is meaningful
when comparing patients with non-specific low back pain.
3. Comparing activity streams with missing data. The wristbands have limited
battery time and are often not water proof and needs to be taken off during
showering or swimming activities. Users might find it uncomfortable to wear
the wristbands continuously and try to loosen the strap or take breaks in
wearing them, which might introduce gaps in recordings.
Because these problems are very novel, we have very little knowledge on what
to include and how to compare such activity streams. There is some evidence
that sleeping patterns and prolonged periods of inactivity are correlated with an
increase in non-specific low back pain [1], [3], [5]. There is however, very limited
data on what activity types or intensities of activity types are beneficial/harmful
when having non-specific low back pain.
The activity streams that we have experimented with can contain activity
data from periods of up to several weeks. This is because to see the effect of
a change in behaviour, by applying some advice, one often needs to wait for
about 4-6 weeks. Including all kinds of information into an activity stream (type
of activity, intensity, skin temperature, pulse, moisture of the skin, etc.), comes
with an increase in the computational cost when comparing the resulting activity
streams against each other. See figure 1 for examples of abstractions that can
be included.
We have also experimented with different resolutions on our activity streams.
We are able to capture and transform data into 1 second as well as 1 hour
windows to name a few. Decreasing the resolution, lowers the computational
comparison cost, but important data may be lost during the transformation.
Different resolutions can be combined throughout the activity stream to prevent
loss of information. For example a bigger window can be used during prolonged
periods of inactivity or sleep. On the other hand, during periods of activity where
the intensity or activity types changes frequently, a smaller window can be used
to capture and conserve that information.
175
Fig. 1. Three abstractions from collected wristband data: step counts together with
heart rate and skin temperature measures.
Other ideas we have been experimenting with include comparing the activity
streams based on the activity intensity rather than activity type. Clinicians
with experience in the field of non-specific low back pain have suggested, instead
of comparing the streams directly, to compare sleeping patterns and periods
of prolonged inactivity. This way of comparing activity streams produces very
different results than when comparing activity streams containing activity types.
This is shown in figure 2.
Fig. 2. The data from figure 1 abstracted to a higher level and represented as an activity
stream in three different ways. S1 and S2 represents activity intensity throughout the
day (S2 with higher resolution). In S3 the activity stream is transformed, as suggested
by clinical experts, to only show prolonged inactivity periods. The green periods in S3
represent any activity.
As for the comparison algorithms themselves, we are currently experimenting
with a string based approach. We convert the activity stream into a string of
characters. When comparing two strings, we compute the percentage of each
character per string and the number of consecutive sequences of those characters.
We then also compute the longest common sub-sequence, the sequence distance,
the number of similar k-mers and the number of unique similar k-mers. The
sequence distance is the Levenshtein distance which is the least number of single
176
character insertions, deletions or substitutions that is required to transform one
string into the other [4]. A k-mer is a consecutive substring of length k. All these
attributes are combined in to a final similarity metric.
3 Current status and Outlook
We have presented and discussed various challenges that arise when comparing
activity streams in the selfBACK project.
Our current work focuses on the representation and comparison of the activity
stream data that is meaningful in the non-specific low back pain domain. We are
investigating which time resolution is beneficial and which information should
be included in the activity streams. We are also exploring whether and how
domain knowledge from non-specific low back pain experts can be incorporated.
As shown in figure 2, the representation of the activity streams can differ a
lot depending on which information is most relevant (alteration of activities in
S1 and S2 vs. the reduction of pro-longed inactivtiy in S3). Therefore we are
working on how the representation affects the similarity comparison of activity
streams collected from real patients.
Our next steps include the investigation whether activity stream rearrange-
ment could yield towards a more precise comparison. Therefore we investigate if
parts of the activity stream are more important than others (for example sleep
patterns). When the data collection from the planned randomized control trial
is complete, it will also be possible to mine the collected activity streams for
behavioral patterns that contribute to an increase or a decrease in non-specific
low back pain. For example we could discover that a certain amount of time
standing followed by a certain amount of time sitting is increasing non-specific
low back pain etc. Finally we will conduct experiments on how much weight the
activity stream comparison should have compared to the self-reported reported
questionnaire data from the patients.
Acknowledgement The work has been conducted as part of the selfBACK
project, which has received funding from the European Unions Horizon 2020
research and innovation programme under grant agreement No 689043.
References
1. Alsaadi, S.M., McAuley, J.H., Hush, J.M., Lo, S., Lin, C.W., Williams, C.M., Maher,
C.G.: Poor sleep quality is strongly associated with subsequent pain intensity in
patients with acute low back pain. Arthritis Rheumatology 66(5), 1388–94 (2014)
2. Bach, K., Szczepanski, T., Aamodt, A., Gundersen, O.E., Mork, P.J.: Case Rep-
resentation and Similarity Assessment in the selfBACK Decision Support System.
ICCBR 2116 (accepted for publication)
3. Gupta, N., Christiansen, C.S., Hallman, D.M., Korshj, M., Carneiro, I.G., Holter-
mann, A.: Is Objectively Measured Sitting Time Associated with Low Back Pain?
A Cross-Sectional Investigation in the NOMAD study. PLoS ONE 10(3) (2015)
177
4. Navarro, G.: A guided tour to approximate string matching. ACM Computing Sur-
veys 33(1), 31–88 (2001)
5. van Tulder, M., Becker, A., Bekkering, T., Breen, A., del Real, M.T.G., Hutchinson,
A., Koes, B., Laerum, E., Malmivaara, A.: Chapter 3 european guidelines for the
management ofacute nonspecific low back painin primary care. European Spine
Journal 15(2), s169–s191 (2006)