Using Mobile Phones for Activity Recognition in Parkinson’s Patients
Identifieur interne : 000385 ( Ncbi/Merge ); précédent : 000384; suivant : 000386Using Mobile Phones for Activity Recognition in Parkinson’s Patients
Auteurs : Mark Albert [États-Unis] ; Santiago Toledo [États-Unis] ; Mark Shapiro [États-Unis] ; Konrad Kording [États-Unis]Source :
- Frontiers in Neurology ; 2012.
Abstract
Mobile phones with built-in accelerometers promise a convenient, objective way to quantify everyday movements and classify those movements into activities. Using accelerometer data we estimate the following activities of 18 healthy subjects and eight patients with Parkinson’s disease: walking, standing, sitting, holding, or not wearing the phone. We use standard machine learning classifiers (support vector machines, regularized logistic regression) to automatically select, weigh, and combine a large set of standard features for time series analysis. Using cross validation across all samples we are able to correctly identify 96.1% of the activities of healthy subjects and 92.2% of the activities of Parkinson’s patients. However, when applying the classification parameters derived from the set of healthy subjects to Parkinson’s patients, the percent correct lowers to 60.3%, due to different characteristics of movement. For a fairer comparison across populations we also applied subject-wise cross validation, identifying healthy subject activities with 86.0% accuracy and 75.1% accuracy for patients. We discuss the key differences between these populations, and why algorithms designed for and trained with healthy subject data are not reliable for activity recognition in populations with motor disabilities.
Url:
DOI: 10.3389/fneur.2012.00158
PubMed: 23162528
PubMed Central: 3491315
Links toward previous steps (curation, corpus...)
- to stream Pmc, to step Corpus: 000279
- to stream Pmc, to step Curation: 000279
- to stream Pmc, to step Checkpoint: 000091
Links to Exploration step
PMC:3491315Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en">Using Mobile Phones for Activity Recognition in Parkinson’s Patients</title>
<author><name sortKey="Albert, Mark V" sort="Albert, Mark V" uniqKey="Albert M" first="Mark" last="Albert">Mark Albert</name>
<affiliation wicri:level="1"><nlm:aff id="aff1"><institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
<author><name sortKey="Toledo, Santiago" sort="Toledo, Santiago" uniqKey="Toledo S" first="Santiago" last="Toledo">Santiago Toledo</name>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
<author><name sortKey="Shapiro, Mark" sort="Shapiro, Mark" uniqKey="Shapiro M" first="Mark" last="Shapiro">Mark Shapiro</name>
<affiliation wicri:level="1"><nlm:aff id="aff1"><institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
<author><name sortKey="Kording, Konrad" sort="Kording, Konrad" uniqKey="Kording K" first="Konrad" last="Kording">Konrad Kording</name>
<affiliation wicri:level="1"><nlm:aff id="aff1"><institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">PMC</idno>
<idno type="pmid">23162528</idno>
<idno type="pmc">3491315</idno>
<idno type="url">http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3491315</idno>
<idno type="RBID">PMC:3491315</idno>
<idno type="doi">10.3389/fneur.2012.00158</idno>
<date when="2012">2012</date>
<idno type="wicri:Area/Pmc/Corpus">000279</idno>
<idno type="wicri:Area/Pmc/Curation">000279</idno>
<idno type="wicri:Area/Pmc/Checkpoint">000091</idno>
<idno type="wicri:Area/Ncbi/Merge">000385</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title xml:lang="en" level="a" type="main">Using Mobile Phones for Activity Recognition in Parkinson’s Patients</title>
<author><name sortKey="Albert, Mark V" sort="Albert, Mark V" uniqKey="Albert M" first="Mark" last="Albert">Mark Albert</name>
<affiliation wicri:level="1"><nlm:aff id="aff1"><institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
<author><name sortKey="Toledo, Santiago" sort="Toledo, Santiago" uniqKey="Toledo S" first="Santiago" last="Toledo">Santiago Toledo</name>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
<author><name sortKey="Shapiro, Mark" sort="Shapiro, Mark" uniqKey="Shapiro M" first="Mark" last="Shapiro">Mark Shapiro</name>
<affiliation wicri:level="1"><nlm:aff id="aff1"><institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
<author><name sortKey="Kording, Konrad" sort="Kording, Konrad" uniqKey="Kording K" first="Konrad" last="Kording">Konrad Kording</name>
<affiliation wicri:level="1"><nlm:aff id="aff1"><institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
<affiliation wicri:level="1"><nlm:aff id="aff2"><institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</nlm:aff>
<country xml:lang="fr">États-Unis</country>
<wicri:regionArea></wicri:regionArea>
<wicri:regionArea># see nlm:aff region in country</wicri:regionArea>
</affiliation>
</author>
</analytic>
<series><title level="j">Frontiers in Neurology</title>
<idno type="e-ISSN">1664-2295</idno>
<imprint><date when="2012">2012</date>
</imprint>
</series>
</biblStruct>
</sourceDesc>
</fileDesc>
<profileDesc><textClass></textClass>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en"><p>Mobile phones with built-in accelerometers promise a convenient, objective way to quantify everyday movements and classify those movements into activities. Using accelerometer data we estimate the following activities of 18 healthy subjects and eight patients with Parkinson’s disease: walking, standing, sitting, holding, or not wearing the phone. We use standard machine learning classifiers (support vector machines, regularized logistic regression) to automatically select, weigh, and combine a large set of standard features for time series analysis. Using cross validation across all samples we are able to correctly identify 96.1% of the activities of healthy subjects and 92.2% of the activities of Parkinson’s patients. However, when applying the classification parameters derived from the set of healthy subjects to Parkinson’s patients, the percent correct lowers to 60.3%, due to different characteristics of movement. For a fairer comparison across populations we also applied subject-wise cross validation, identifying healthy subject activities with 86.0% accuracy and 75.1% accuracy for patients. We discuss the key differences between these populations, and why algorithms designed for and trained with healthy subject data are not reliable for activity recognition in populations with motor disabilities.</p>
</div>
</front>
<back><div1 type="bibliography"><listBibl><biblStruct><analytic><author><name sortKey="Bao, L" uniqKey="Bao L">L. Bao</name>
</author>
<author><name sortKey="Intille, S" uniqKey="Intille S">S. Intille</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Bieber, G" uniqKey="Bieber G">G. Bieber</name>
</author>
<author><name sortKey="Voskamp, J R" uniqKey="Voskamp J">J. R. Voskamp</name>
</author>
<author><name sortKey="Urban, B" uniqKey="Urban B">B. Urban</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Brezmes, T" uniqKey="Brezmes T">T. Brezmes</name>
</author>
<author><name sortKey="Gorricho, J L" uniqKey="Gorricho J">J.-L. Gorricho</name>
</author>
<author><name sortKey="Cotrina, J" uniqKey="Cotrina J">J. Cotrina</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Chang, C C" uniqKey="Chang C">C.-C. Chang</name>
</author>
<author><name sortKey="Lin, C J" uniqKey="Lin C">C.-J. Lin</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Choudhury, T" uniqKey="Choudhury T">T. Choudhury</name>
</author>
<author><name sortKey="Consolvo, S" uniqKey="Consolvo S">S. Consolvo</name>
</author>
<author><name sortKey="Harrison, B" uniqKey="Harrison B">B. Harrison</name>
</author>
<author><name sortKey="Hightower, J" uniqKey="Hightower J">J. Hightower</name>
</author>
<author><name sortKey="Lamarca, A" uniqKey="Lamarca A">A. Lamarca</name>
</author>
<author><name sortKey="Legrand, L" uniqKey="Legrand L">L. Legrand</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Dibble, L E" uniqKey="Dibble L">L. E. Dibble</name>
</author>
<author><name sortKey="Addison, O" uniqKey="Addison O">O. Addison</name>
</author>
<author><name sortKey="Papa, E" uniqKey="Papa E">E. Papa</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Dorsey, E R" uniqKey="Dorsey E">E. R. Dorsey</name>
</author>
<author><name sortKey="Constantinescu, R" uniqKey="Constantinescu R">R. Constantinescu</name>
</author>
<author><name sortKey="Thompson, J P" uniqKey="Thompson J">J. P. Thompson</name>
</author>
<author><name sortKey="Biglan, K M" uniqKey="Biglan K">K. M. Biglan</name>
</author>
<author><name sortKey="Holloway, R G" uniqKey="Holloway R">R. G. Holloway</name>
</author>
<author><name sortKey="Kieburtz, K" uniqKey="Kieburtz K">K. Kieburtz</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Ellis, T" uniqKey="Ellis T">T. Ellis</name>
</author>
<author><name sortKey="Cavanaugh, J T" uniqKey="Cavanaugh J">J. T. Cavanaugh</name>
</author>
<author><name sortKey="Earhart, G M" uniqKey="Earhart G">G. M. Earhart</name>
</author>
<author><name sortKey="Ford, M P" uniqKey="Ford M">M. P. Ford</name>
</author>
<author><name sortKey="Foreman, K B" uniqKey="Foreman K">K. B. Foreman</name>
</author>
<author><name sortKey="Dibble, L E" uniqKey="Dibble L">L. E. Dibble</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Fernandes, H L" uniqKey="Fernandes H">H. L. Fernandes</name>
</author>
<author><name sortKey="Albert, M V" uniqKey="Albert M">M. V. Albert</name>
</author>
<author><name sortKey="Kording, K P" uniqKey="Kording K">K. P. Kording</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Gelb, D J" uniqKey="Gelb D">D. J. Gelb</name>
</author>
<author><name sortKey="Oliver, E" uniqKey="Oliver E">E. Oliver</name>
</author>
<author><name sortKey="Gilman, S" uniqKey="Gilman S">S. Gilman</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Goodwin, V A" uniqKey="Goodwin V">V. A. Goodwin</name>
</author>
<author><name sortKey="Richards, S H" uniqKey="Richards S">S. H. Richards</name>
</author>
<author><name sortKey="Taylor, R S" uniqKey="Taylor R">R. S. Taylor</name>
</author>
<author><name sortKey="Taylor, A H" uniqKey="Taylor A">A. H. Taylor</name>
</author>
<author><name sortKey="Campbell, J L" uniqKey="Campbell J">J. L. Campbell</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Gyorbiro, N" uniqKey="Gyorbiro N">N. Gyorbiro</name>
</author>
<author><name sortKey="Fabian, A" uniqKey="Fabian A">A. Fabian</name>
</author>
<author><name sortKey="Homanyi, G" uniqKey="Homanyi G">G. Homanyi</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Jankovic, J" uniqKey="Jankovic J">J. Jankovic</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Jeannet, P Y" uniqKey="Jeannet P">P.-Y. Jeannet</name>
</author>
<author><name sortKey="Aminian, K" uniqKey="Aminian K">K. Aminian</name>
</author>
<author><name sortKey="Bloetzer, C" uniqKey="Bloetzer C">C. Bloetzer</name>
</author>
<author><name sortKey="Najafi, B" uniqKey="Najafi B">B. Najafi</name>
</author>
<author><name sortKey="Paraschiv Ionescu, A" uniqKey="Paraschiv Ionescu A">A. Paraschiv-Ionescu</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Joundi, R A" uniqKey="Joundi R">R. A. Joundi</name>
</author>
<author><name sortKey="Brittain, J S" uniqKey="Brittain J">J.-S. Brittain</name>
</author>
<author><name sortKey="Jenkinson, N" uniqKey="Jenkinson N">N. Jenkinson</name>
</author>
<author><name sortKey="Green, A L" uniqKey="Green A">A. L. Green</name>
</author>
<author><name sortKey="Aziz, T" uniqKey="Aziz T">T. Aziz</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Kavanagh, J J" uniqKey="Kavanagh J">J. J. Kavanagh</name>
</author>
<author><name sortKey="Menz, H B" uniqKey="Menz H">H. B. Menz</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Krishnan, N C" uniqKey="Krishnan N">N. C. Krishnan</name>
</author>
<author><name sortKey="Panchanathan, S" uniqKey="Panchanathan S">S. Panchanathan</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Krishnapuram, B" uniqKey="Krishnapuram B">B. Krishnapuram</name>
</author>
<author><name sortKey="Carin, L" uniqKey="Carin L">L. Carin</name>
</author>
<author><name sortKey="Figueiredo, M A T" uniqKey="Figueiredo M">M. A. T. Figueiredo</name>
</author>
<author><name sortKey="Hartemink, A J" uniqKey="Hartemink A">A. J. Hartemink</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Kwapisz, J R" uniqKey="Kwapisz J">J. R. Kwapisz</name>
</author>
<author><name sortKey="Weiss, G M" uniqKey="Weiss G">G. M. Weiss</name>
</author>
<author><name sortKey="Moore, S A" uniqKey="Moore S">S. A. Moore</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Lee, M" uniqKey="Lee M">M. Lee</name>
</author>
<author><name sortKey="Kim, J" uniqKey="Kim J">J. Kim</name>
</author>
<author><name sortKey="Kim, K" uniqKey="Kim K">K. Kim</name>
</author>
<author><name sortKey="Lee, I" uniqKey="Lee I">I. Lee</name>
</author>
<author><name sortKey="Jee, S H" uniqKey="Jee S">S. H. Jee</name>
</author>
<author><name sortKey="Yoo, S K" uniqKey="Yoo S">S. K. Yoo</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Long, X" uniqKey="Long X">X. Long</name>
</author>
<author><name sortKey="Yin, B" uniqKey="Yin B">B. Yin</name>
</author>
<author><name sortKey="Aarts, R M" uniqKey="Aarts R">R. M. Aarts</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Mannini, A" uniqKey="Mannini A">A. Mannini</name>
</author>
<author><name sortKey="Sabatini, A M" uniqKey="Sabatini A">A. M. Sabatini</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Mathie, M J" uniqKey="Mathie M">M. J. Mathie</name>
</author>
<author><name sortKey="Celler, B G" uniqKey="Celler B">B. G. Celler</name>
</author>
<author><name sortKey="Lovell, N H" uniqKey="Lovell N">N. H. Lovell</name>
</author>
<author><name sortKey="Coster, A C" uniqKey="Coster A">A. C. Coster</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Mathie, M J" uniqKey="Mathie M">M. J. Mathie</name>
</author>
<author><name sortKey="Coster, A C F" uniqKey="Coster A">A. C. F. Coster</name>
</author>
<author><name sortKey="Lovell, N H" uniqKey="Lovell N">N. H. Lovell</name>
</author>
<author><name sortKey="Celler, B G" uniqKey="Celler B">B. G. Celler</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Maurer, U" uniqKey="Maurer U">U. Maurer</name>
</author>
<author><name sortKey="Smailagic, A" uniqKey="Smailagic A">A. Smailagic</name>
</author>
<author><name sortKey="Siewiorek, D P" uniqKey="Siewiorek D">D. P. Siewiorek</name>
</author>
<author><name sortKey="Deisher, M" uniqKey="Deisher M">M. Deisher</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Najafi, B" uniqKey="Najafi B">B. Najafi</name>
</author>
<author><name sortKey="Aminian, K" uniqKey="Aminian K">K. Aminian</name>
</author>
<author><name sortKey="Paraschiv Ionescu, A" uniqKey="Paraschiv Ionescu A">A. Paraschiv-Ionescu</name>
</author>
<author><name sortKey="Loew, F" uniqKey="Loew F">F. Loew</name>
</author>
<author><name sortKey="Bula, C J" uniqKey="Bula C">C. J. Bula</name>
</author>
<author><name sortKey="Robert, P" uniqKey="Robert P">P. Robert</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Palmer, S S" uniqKey="Palmer S">S. S. Palmer</name>
</author>
<author><name sortKey="Mortimer, J A" uniqKey="Mortimer J">J. A. Mortimer</name>
</author>
<author><name sortKey="Webster, D D" uniqKey="Webster D">D. D. Webster</name>
</author>
<author><name sortKey="Bistevins, R" uniqKey="Bistevins R">R. Bistevins</name>
</author>
<author><name sortKey="Dickinson, G L" uniqKey="Dickinson G">G. L. Dickinson</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Ravi, N" uniqKey="Ravi N">N. Ravi</name>
</author>
<author><name sortKey="Dandekar, N" uniqKey="Dandekar N">N. Dandekar</name>
</author>
<author><name sortKey="Mysore, P" uniqKey="Mysore P">P. Mysore</name>
</author>
<author><name sortKey="Littman, M" uniqKey="Littman M">M. Littman</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Ryder, J" uniqKey="Ryder J">J. Ryder</name>
</author>
<author><name sortKey="Longstaff, B" uniqKey="Longstaff B">B. Longstaff</name>
</author>
<author><name sortKey="Reddy, S" uniqKey="Reddy S">S. Reddy</name>
</author>
<author><name sortKey="Estrin, D" uniqKey="Estrin D">D. Estrin</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Salarian, A" uniqKey="Salarian A">A. Salarian</name>
</author>
<author><name sortKey="Russmann, H" uniqKey="Russmann H">H. Russmann</name>
</author>
<author><name sortKey="Vingerhoets, F J G" uniqKey="Vingerhoets F">F. J. G. Vingerhoets</name>
</author>
<author><name sortKey="Burkhard, P R" uniqKey="Burkhard P">P. R. Burkhard</name>
</author>
<author><name sortKey="Aminian, K" uniqKey="Aminian K">K. Aminian</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Schenkman, M" uniqKey="Schenkman M">M. Schenkman</name>
</author>
<author><name sortKey="Cutson, T M" uniqKey="Cutson T">T. M. Cutson</name>
</author>
<author><name sortKey="Kuchibhatla, M" uniqKey="Kuchibhatla M">M. Kuchibhatla</name>
</author>
<author><name sortKey="Chandler, J" uniqKey="Chandler J">J. Chandler</name>
</author>
<author><name sortKey="Pieper, C F" uniqKey="Pieper C">C. F. Pieper</name>
</author>
<author><name sortKey="Ray, L" uniqKey="Ray L">L. Ray</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Stone, A A" uniqKey="Stone A">A. A. Stone</name>
</author>
<author><name sortKey="Shiffman, S" uniqKey="Shiffman S">S. Shiffman</name>
</author>
<author><name sortKey="Schwartz, J E" uniqKey="Schwartz J">J. E. Schwartz</name>
</author>
<author><name sortKey="Broderick, J E" uniqKey="Broderick J">J. E. Broderick</name>
</author>
<author><name sortKey="Hufford, M R" uniqKey="Hufford M">M. R. Hufford</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Surangsrirat, D" uniqKey="Surangsrirat D">D. Surangsrirat</name>
</author>
<author><name sortKey="Thanawattano, C" uniqKey="Thanawattano C">C. Thanawattano</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Tapia, E M" uniqKey="Tapia E">E. M. Tapia</name>
</author>
<author><name sortKey="Intille, S S" uniqKey="Intille S">S. S. Intille</name>
</author>
<author><name sortKey="Haskell, W" uniqKey="Haskell W">W. Haskell</name>
</author>
<author><name sortKey="Larson, K" uniqKey="Larson K">K. Larson</name>
</author>
<author><name sortKey="Wright, J" uniqKey="Wright J">J. Wright</name>
</author>
<author><name sortKey="King, A" uniqKey="King A">A. King</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Wang, Y" uniqKey="Wang Y">Y. Wang</name>
</author>
<author><name sortKey="Lin, J" uniqKey="Lin J">J. Lin</name>
</author>
<author><name sortKey="Annavaram, M" uniqKey="Annavaram M">M. Annavaram</name>
</author>
<author><name sortKey="Jacobson, Q A" uniqKey="Jacobson Q">Q. A. Jacobson</name>
</author>
<author><name sortKey="Hong, J" uniqKey="Hong J">J. Hong</name>
</author>
<author><name sortKey="Krishnamachari, B" uniqKey="Krishnamachari B">B. Krishnamachari</name>
</author>
</analytic>
</biblStruct>
<biblStruct><analytic><author><name sortKey="Yang, J" uniqKey="Yang J">J. Yang</name>
</author>
</analytic>
</biblStruct>
</listBibl>
</div1>
</back>
</TEI>
<pmc article-type="research-article"><pmc-dir>properties open_access</pmc-dir>
<front><journal-meta><journal-id journal-id-type="nlm-ta">Front Neurol</journal-id>
<journal-id journal-id-type="iso-abbrev">Front Neurol</journal-id>
<journal-id journal-id-type="publisher-id">Front. Neur.</journal-id>
<journal-title-group><journal-title>Frontiers in Neurology</journal-title>
</journal-title-group>
<issn pub-type="epub">1664-2295</issn>
<publisher><publisher-name>Frontiers Media S.A.</publisher-name>
</publisher>
</journal-meta>
<article-meta><article-id pub-id-type="pmid">23162528</article-id>
<article-id pub-id-type="pmc">3491315</article-id>
<article-id pub-id-type="doi">10.3389/fneur.2012.00158</article-id>
<article-categories><subj-group subj-group-type="heading"><subject>Neuroscience</subject>
<subj-group><subject>Original Research</subject>
</subj-group>
</subj-group>
</article-categories>
<title-group><article-title>Using Mobile Phones for Activity Recognition in Parkinson’s Patients</article-title>
</title-group>
<contrib-group><contrib contrib-type="author"><name><surname>Albert</surname>
<given-names>Mark V.</given-names>
</name>
<xref ref-type="aff" rid="aff1"><sup>1</sup>
</xref>
<xref ref-type="aff" rid="aff2"><sup>2</sup>
</xref>
<xref ref-type="author-notes" rid="fn001">*</xref>
</contrib>
<contrib contrib-type="author"><name><surname>Toledo</surname>
<given-names>Santiago</given-names>
</name>
<xref ref-type="aff" rid="aff2"><sup>2</sup>
</xref>
</contrib>
<contrib contrib-type="author"><name><surname>Shapiro</surname>
<given-names>Mark</given-names>
</name>
<xref ref-type="aff" rid="aff1"><sup>1</sup>
</xref>
<xref ref-type="aff" rid="aff2"><sup>2</sup>
</xref>
</contrib>
<contrib contrib-type="author"><name><surname>Kording</surname>
<given-names>Konrad</given-names>
</name>
<xref ref-type="aff" rid="aff1"><sup>1</sup>
</xref>
<xref ref-type="aff" rid="aff2"><sup>2</sup>
</xref>
</contrib>
</contrib-group>
<aff id="aff1"><sup>1</sup>
<institution>Sensory Motor Performance Program, Rehabilitation Institute of Chicago</institution>
<country>Chicago, IL, USA</country>
</aff>
<aff id="aff2"><sup>2</sup>
<institution>Department of Physical Medicine and Rehabilitation, Northwestern University</institution>
<country>Chicago, IL, USA</country>
</aff>
<author-notes><fn fn-type="edited-by"><p>Edited by: Ryuji Kaji, Tokushima University Hospital, Japan</p>
</fn>
<fn fn-type="edited-by"><p>Reviewed by: Pedro Chana, Universidad de Santiago de Chile, Chile; Benzi Kluger, University of Colorado Denver, USA; Maria Fiorella Contarino, Academic Medical Center, Netherlands</p>
</fn>
<corresp id="fn001">*Correspondence: Mark V. Albert, Sensory Motor Performance Program, Rehabilitation Institute of Chicago, 345 E Superior Street room 1479, Chicago, IL 60611, USA. e-mail: <email xlink:type="simple">markvalbert@gmail.com</email>
</corresp>
<fn fn-type="other" id="fn002"><p>This article was submitted to Frontiers in Movement Disorders, a specialty of Frontiers in Neurology.</p>
</fn>
</author-notes>
<pub-date pub-type="epub"><day>07</day>
<month>11</month>
<year>2012</year>
</pub-date>
<pub-date pub-type="collection"><year>2012</year>
</pub-date>
<volume>3</volume>
<elocation-id>158</elocation-id>
<history><date date-type="received"><day>18</day>
<month>4</month>
<year>2012</year>
</date>
<date date-type="accepted"><day>19</day>
<month>10</month>
<year>2012</year>
</date>
</history>
<permissions><copyright-statement>Copyright © 2012 Albert, Toledo, Shapiro and Kording.</copyright-statement>
<copyright-year>2012</copyright-year>
<license license-type="open-access" xlink:href="http://www.frontiersin.org/licenseagreement"><license-p>This is an open-access article distributed under the terms of the <uri xlink:type="simple" xlink:href="http://creativecommons.org/licenses/by/3.0/">Creative Commons Attribution License</uri>
, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc.</license-p>
</license>
</permissions>
<abstract><p>Mobile phones with built-in accelerometers promise a convenient, objective way to quantify everyday movements and classify those movements into activities. Using accelerometer data we estimate the following activities of 18 healthy subjects and eight patients with Parkinson’s disease: walking, standing, sitting, holding, or not wearing the phone. We use standard machine learning classifiers (support vector machines, regularized logistic regression) to automatically select, weigh, and combine a large set of standard features for time series analysis. Using cross validation across all samples we are able to correctly identify 96.1% of the activities of healthy subjects and 92.2% of the activities of Parkinson’s patients. However, when applying the classification parameters derived from the set of healthy subjects to Parkinson’s patients, the percent correct lowers to 60.3%, due to different characteristics of movement. For a fairer comparison across populations we also applied subject-wise cross validation, identifying healthy subject activities with 86.0% accuracy and 75.1% accuracy for patients. We discuss the key differences between these populations, and why algorithms designed for and trained with healthy subject data are not reliable for activity recognition in populations with motor disabilities.</p>
</abstract>
<kwd-group><kwd>mobile phone</kwd>
<kwd>accelerometer</kwd>
<kwd>activity recognition</kwd>
<kwd>Parkinson’s disease</kwd>
</kwd-group>
<counts><fig-count count="2"></fig-count>
<table-count count="6"></table-count>
<equation-count count="0"></equation-count>
<ref-count count="36"></ref-count>
<page-count count="7"></page-count>
<word-count count="5730"></word-count>
</counts>
</article-meta>
</front>
<body><sec><title>Introduction</title>
<p>Accurately tracking the activities of patients with motor disabilities has the potential to better inform patient care. With more precise, objective measures treatment alternatives can be evaluated more definitively. This is particularly important in motor disabilities such as Parkinson’s disease (PD) that respond to an increasing variety of treatment options, including drugs and various exercise therapies (Palmer et al., <xref ref-type="bibr" rid="B27">1986</xref>
; Schenkman et al., <xref ref-type="bibr" rid="B31">1998</xref>
; Goodwin et al., <xref ref-type="bibr" rid="B11">2008</xref>
; Dibble et al., <xref ref-type="bibr" rid="B6">2009</xref>
). Quantifying symptoms both in the clinic and at home has the potential to provide functional measures better associated with quality of life (Ellis et al., <xref ref-type="bibr" rid="B8">2011</xref>
).</p>
<p>Current means of evaluating patient mobility are limited. Clinical evaluations require a patient to travel to the location of a health care provider, and testing is expensive in terms of money and clinician time. This limits the frequency of clinical evaluations to only a few times per year for personal needs and only a few times per week during research studies. For better temporal resolution, studies often turn to patient journaling. Asking patients to periodically indicate their activities suffers from a number of problems. First, it is subjective, leading to changes based on the mental state of the subject. Also, because of the inconvenience to patients, it is difficult to achieve high compliance, with one study indicating only an 11% compliance rate using journaling (Stone et al., <xref ref-type="bibr" rid="B32">2003</xref>
). It would be helpful to develop a measure of evaluating patient mobility that is both frequent and convenient for the subject.</p>
<p>Dedicated accelerometers are an inexpensive, standardized way of measuring movement (Mathie et al., <xref ref-type="bibr" rid="B24">2004b</xref>
). These components can cost as little as one dollar, and can measure movement in any direction as well as orientation relative to gravity. For example, by attaching an accelerometer to a shoe, one can estimate the amount of time running and walking based on the periodic movements. For activity recognition, there have been many studies which place these accelerometers at specific locations on the body – including the head, chest, arm, foot, and thigh (reviewed in Kavanagh and Menz, <xref ref-type="bibr" rid="B16">2008</xref>
). The advantage of placing accelerometers at such locations is more consistent signals across individuals. Although such accelerometers are inexpensive, they are often dedicated equipment that has to remain attached at a particular location on the body to be effective. This is often an impractical inconvenience for long-term use.</p>
<p>Modern mobile phones have built-in accelerometers that can be used to track movements without the need for an additional device (Brezmes et al., <xref ref-type="bibr" rid="B3">2009</xref>
; Gyorbiro et al., <xref ref-type="bibr" rid="B12">2009</xref>
; Ryder et al., <xref ref-type="bibr" rid="B29">2009</xref>
; Fernandes et al., <xref ref-type="bibr" rid="B9">2011</xref>
). They have their own power sources, memory storage capabilities, and can transmit data wirelessly. Patients could simply download an app onto their smartphone enabling data collection and analysis. Mobile phones allow automatic, convenient, real-time monitoring, and recording, which can be invaluable to large-scale studies and personal patient health monitoring.</p>
<p>Unfortunately, applying current mobile phone strategies to populations with motor disabilities is challenging. For example, PD symptoms include tremor, slowed motion (bradykinesia), rigid muscles, loss of common automatic movements, and impaired posture (Jankovic, <xref ref-type="bibr" rid="B13">2008</xref>
). These symptoms all can adversely affect activity recognition. Activity recognition strategies have been tailored specifically for the elderly (Najafi et al., <xref ref-type="bibr" rid="B26">2003</xref>
), individuals with muscular dystrophy (Jeannet et al., <xref ref-type="bibr" rid="B14">2011</xref>
), and even PD (Salarian et al., <xref ref-type="bibr" rid="B30">2007</xref>
), but each of these studies were done with accelerometers at standardized locations or using multiple sensors throughout the body. Performing activity recognition for a population with a motor disability when carrying the phone naturally in pockets or belt clips provides additional challenges.</p>
<p>Here we show how modern machine learning techniques can quantify the movements of PD patients that carry mobile phones. We first collect data from both healthy subjects and PD patients performing a standard set of activities. From this data, we analyze the precision of activity recognition both within and across the two groups. Ultimately, we demonstrate an approach to activity recognition using mobile phone devices for patient populations, allowing us to better monitor patient responses to treatment in future studies.</p>
</sec>
<sec sec-type="materials|methods" id="s1"><title>Materials and Methods</title>
<p>Eighteen healthy subjects and eight PD patients were recruited for this study. The eighteen healthy subjects, having had no previous history of a movement disorder, came from two groups, 13 younger subjects (6M/7F, 25.1 ± 3.0 years), and five older subjects (5F, 53.4 ± 7.4 years). The PD patients were mild and moderately affect patients in Hoehn and Yahr stage 1–3 (7F/1M, 67.0 ± 8.1 years, median ± range). Patients were recorded while taking their usual medications (ON-med condition). Many patients presented mild dyskinesias during the course of the experiment. Written, informed consent was obtained for all subjects. The Northwestern University Institutional Review Board approved this study.</p>
<p>All subjects were instructed to carry T-mobile G1 phones running Android OS version 1.6 in their front pockets. These phones have a standard built-in tri-axial accelerometer with a range of ±2.8 g. The sampling rate was variable between 15 and 25 Hz depending upon the amount of movement.</p>
<p>Subjects were instructed to perform a number of different activities, each for at least 1 min. Before each activity, the subject would select the activity on a specially designed phone app (Figure <xref ref-type="fig" rid="F1">1</xref>
) with the experimenter present to minimize errors. The accelerations were labeled according to the activity they were performing. The activities were performed in a laboratory setting in the order shown below. For a few initial subjects the activities began and ended with an additional lying down activity, but the protocol was simplified by removing this activity for all subjects later on. We also repeated activities to get more recordings with the phone in slightly different positions on the subjects.</p>
<list list-type="order"><list-item><p>sitting</p>
</list-item>
<list-item><p>standing</p>
</list-item>
<list-item><p>holding the phone (standing with arms bent forward)</p>
</list-item>
<list-item><p>walking</p>
</list-item>
<list-item><p>not wearing (set on a table)</p>
</list-item>
<list-item><p>walking</p>
</list-item>
<list-item><p>holding the phone</p>
</list-item>
<list-item><p>standing</p>
</list-item>
<list-item><p>sitting</p>
</list-item>
</list>
<fig id="F1" position="float"><label>Figure 1</label>
<caption><p><bold>Recording device and software</bold>
. <bold>(A)</bold>
The subjects carried T-mobile G1 android phones in their pockets. <bold>(B)</bold>
The axes of the accelerometer relative to the orientation of the phone in <bold>(A)</bold>
. <bold>(C)</bold>
The screen which subjects selected which activity they were performing.</p>
</caption>
<graphic xlink:href="fneur-03-00158-g001"></graphic>
</fig>
<sec><title>Data processing and classification</title>
<p>The accelerometer signals were preprocessed using the following procedure: the recordings were segmented into 10 s clips performing a given activity for the entire duration. The first and last samples were removed from the ends of the recording to allow time for the phones to enter and be removed from the pockets of subjects. Three thousand three hundred and eighty eight samples were recorded from healthy subjects, while 2184 total samples were recorded from PD patients. The subject and researcher both observed labeling during recording, ensuring validity of training, and test data. Misclassified samples were only checked for artifacts. The phone accelerometer values were linearly interpolated from a variable rate between 15 and 25 Hz to match 20 Hz.</p>
<p>The subjects naturally placed the phone in their pockets in the following possible orientations, due to the elongated rectangular shape of the phones.</p>
<list list-type="order"><list-item><p>screen in/right side up</p>
</list-item>
<list-item><p>screen in/upside down</p>
</list-item>
<list-item><p>screen out/right side up</p>
</list-item>
<list-item><p>screen out/upside down</p>
</list-item>
</list>
<p>In the accelerometer readings, each of these orientations only differs by the signs of the axes. For example, flipping the phone right side up to upside down (e.g., orientation #1–#2) changes the sign of the <italic>x</italic>
and <italic>y</italic>
axes, while turning the screen inward (e.g., #1–#3) changes the sign of axes <italic>x</italic>
and <italic>z</italic>
axes. To correct for these different phone orientations, we generated three additional samples for each recorded sample by a coordinate transform that effectively flips the phone 180° along each of its three axes.</p>
<p>From these 10 s clips features were extracted, as summarized in Table <xref ref-type="table" rid="T1">1</xref>
.</p>
<table-wrap id="T1" position="float"><label>Table 1</label>
<caption><p><bold>Features used for activity recognition</bold>
.</p>
</caption>
<table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1">Description</th>
<th align="left" rowspan="1" colspan="1">Total number of values</th>
</tr>
</thead>
<tbody><tr><td align="left" rowspan="1" colspan="1">Mean, absolute value of the mean</td>
<td align="left" rowspan="1" colspan="1">6</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Moments: standard deviation, skew, kurtosis</td>
<td align="left" rowspan="1" colspan="1">9</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">For the change in acceleration: mean, standard deviation, skew, kurtosis</td>
<td align="left" rowspan="1" colspan="1">12</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Root mean square</td>
<td align="left" rowspan="1" colspan="1">3</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Smoothed root mean square (5 pt kernel, 10 pt kernel)</td>
<td align="left" rowspan="1" colspan="1">6</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Extremes: min, max, abs min, abs max</td>
<td align="left" rowspan="1" colspan="1">12</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Histogram: includes counts for −4 to 4 <italic>z</italic>
-score bins</td>
<td align="left" rowspan="1" colspan="1">27</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Fourier components: 32 samples for each axis</td>
<td align="left" rowspan="1" colspan="1">96</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Overall mean acceleration</td>
<td align="left" rowspan="1" colspan="1">1</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Cross product means: <italic>xy</italic>
, <italic>xz</italic>
, <italic>yz</italic>
</td>
<td align="left" rowspan="1" colspan="1">3</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Abs mean of the cross products</td>
<td align="left" rowspan="1" colspan="1">3</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>Two popular algorithms were used for classification: support vector machines (SVM; Chang and Lin, <xref ref-type="bibr" rid="B4">2011</xref>
) and sparse (regularized) multinomial logistic regression sparse multinomial logistic regression (SMLR; Krishnapuram et al., <xref ref-type="bibr" rid="B18">2005</xref>
). Both techniques have been successfully applied in a large number of machine learning classification problems with a great deal of practical success on large feature sets. The hyperparameters for both classifiers were found by a grid search of 10× where <italic>x</italic>
is an integer between −5 and 5 and selecting the maximum cross-validated error in predicting the healthy subject labeled activities. Given the size of the data set used for cross validation this procedure was not expected not lead to noticeable over-fitting for the hyperparameters. For SMLR, the coefficient for the regularization term during optimization, λ, was 0.0001. For SVM, we normalized each feature to have 0 mean and unit variance. We applied radial basis functions, giving us two hyperparameters – the soft slack variable, <italic>C</italic>
, and the size of the Gaussian kernel, γ. The values found by cross validation were <italic>C</italic>
= 1 and γ = 0.1 for the across subjects validation and <italic>C</italic>
= 10 and γ = 1 for the 10-fold validation.</p>
</sec>
</sec>
<sec><title>Results</title>
<p>To examine our ability to classify activities in a patient population, we collected data on both PD patients and healthy subjects, as detailed in the Section <xref ref-type="sec" rid="s1">“Materials and Methods.”</xref>
Subjects were instructed to carry mobile phones in their front pants pocket while performing a series of activities (Figure <xref ref-type="fig" rid="F1">1</xref>
C). We applied two different classifiers, SVM and SMLR, to classify the activities. Our intention is to demonstrate the importance of using classifiers trained with data specifically from patient populations.</p>
<p>The activities subjects performed have characteristic patterns in the accelerometer signals. Recordings were made from the three-axis accelerometers in the phones. The orientation of the phone determined the orientation of the accelerometer axes (Figure <xref ref-type="fig" rid="F1">1</xref>
B). There are also visible differences in the data of PD patients compared to healthy subjects. Example clips from these activities (Figure <xref ref-type="fig" rid="F2">2</xref>
) show the presence of dyskinesias in one PD patient. It can also be observed in the accelerations that walking is often less periodic for PD patients than healthy subjects. Such differences can lead to errors in classification if features such as periodicity or vibrations are used for prediction. Although the movements are related between groups, the exact characteristics of those movements have enough differences between the populations to examine the effect of these differences on classification accuracy.</p>
<fig id="F2" position="float"><label>Figure 2</label>
<caption><p><bold>Typical examples of accelerometer readings for Parkinson’s patients and healthy subjects for the four activities studied</bold>
. Red, green, and blue lines are the <italic>x</italic>
, <italic>y</italic>
, and <italic>z</italic>
-axis accelerations, as specified in Figure <xref ref-type="fig" rid="F1">1</xref>
B. The patient shown here exhibited dyskinesia in the arm that is clearly visible while holding the phone and somewhat visible during standing and sitting. The patient also had an irregular gait cycle during walking.</p>
</caption>
<graphic xlink:href="fneur-03-00158-g002"></graphic>
</fig>
<p>Unlike many studies which classify signals based on only a few, specific features, we did not seek individual features that could be used for classification. For example, knowing if someone is standing, sitting, or holding the phone can depend not only on the orientation of the phone, but also on the amount of vibration in the movement. Instead of searching for particular features with clear, independent differences between activities, we chose to apply the standard, state-of-the-art machine learning approach: we constructed a large feature set and had the algorithms select how to combine and weigh the features appropriately.</p>
<p>First, we wanted to compute a classification accuracy measure that can be compared across studies. To do this, we used 10-fold cross validation, selecting every 10th sample for the test set. This accuracy is expected to be fairly high considering movement patterns specific to individual subjects were in both the training and test sets. For SVM classification, this lead to a 96.1% accuracy for healthy subjects (Table <xref ref-type="table" rid="T2">2</xref>
), and a 92.2% accuracy for PD patients (Table <xref ref-type="table" rid="T3">3</xref>
). Similar results were found for SMLR (89.7% for healthy, and 84.7% for PD patients) so only the SVM results are shown for clarity.</p>
<table-wrap id="T2" position="float"><label>Table 2</label>
<caption><p><bold>Classification matrix for healthy subjects with 10-fold cross validation</bold>
.</p>
</caption>
<table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1">Activity</th>
<th align="left" rowspan="1" colspan="1">Walking</th>
<th align="left" rowspan="1" colspan="1">Standing</th>
<th align="left" rowspan="1" colspan="1">Holding</th>
<th align="left" rowspan="1" colspan="1">Sitting</th>
<th align="left" rowspan="1" colspan="1">Not wearing</th>
</tr>
</thead>
<tbody><tr><td align="left" rowspan="1" colspan="1">Walking</td>
<td align="left" rowspan="1" colspan="1"><bold>908</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Standing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>708</bold>
</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Holding</td>
<td align="left" rowspan="1" colspan="1">2</td>
<td align="left" rowspan="1" colspan="1">22</td>
<td align="left" rowspan="1" colspan="1"><bold>768</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">4</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Sitting</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">32</td>
<td align="left" rowspan="1" colspan="1">48</td>
<td align="left" rowspan="1" colspan="1"><bold>652</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Not wearing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1"><bold>220</bold>
</td>
</tr>
</tbody>
</table>
<table-wrap-foot><p><italic>96.1% overall accuracy for 18 healthy subjects</italic>
.</p>
</table-wrap-foot>
</table-wrap>
<table-wrap id="T3" position="float"><label>Table 3</label>
<caption><p><bold>Classification matrix for PD patients with 10-fold cross validation</bold>
.</p>
</caption>
<table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1">Activity</th>
<th align="left" rowspan="1" colspan="1">Walking</th>
<th align="left" rowspan="1" colspan="1">Standing</th>
<th align="left" rowspan="1" colspan="1">Holding</th>
<th align="left" rowspan="1" colspan="1">Sitting</th>
<th align="left" rowspan="1" colspan="1">Not wearing</th>
</tr>
</thead>
<tbody><tr><td align="left" rowspan="1" colspan="1">Walking</td>
<td align="left" rowspan="1" colspan="1"><bold>662</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">6</td>
<td align="left" rowspan="1" colspan="1">10</td>
<td align="left" rowspan="1" colspan="1">34</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Standing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>400</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">16</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Holding</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>376</bold>
</td>
<td align="left" rowspan="1" colspan="1">32</td>
<td align="left" rowspan="1" colspan="1">4</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Sitting</td>
<td align="left" rowspan="1" colspan="1">10</td>
<td align="left" rowspan="1" colspan="1">24</td>
<td align="left" rowspan="1" colspan="1">2</td>
<td align="left" rowspan="1" colspan="1"><bold>396</bold>
</td>
<td align="left" rowspan="1" colspan="1">4</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Not wearing</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">12</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1"><bold>180</bold>
</td>
</tr>
</tbody>
</table>
<table-wrap-foot><p><italic>92.2% overall accuracy for eight PD subjects</italic>
.</p>
</table-wrap-foot>
</table-wrap>
<p>We sought to quantify the effect of differences between movements made by Parkinson’s patients and healthy subject on the classification algorithm. Unlike previous studies that have analyzed the difference from healthy subjects for particular features (Salarian et al., <xref ref-type="bibr" rid="B30">2007</xref>
; Jeannet et al., <xref ref-type="bibr" rid="B14">2011</xref>
), we directly trained the classifiers using healthy subject data, and applied the classifiers to patient recordings to observe the effect of those differences. From this, we achieved a much lower accuracy of 60.3% using SVM’s (Table <xref ref-type="table" rid="T4">4</xref>
) and 63.5% using SMLR. The lower accuracy is indicative of the difference between the populations and the need to find another, more accurate way to identify patient activities.</p>
<table-wrap id="T4" position="float"><label>Table 4</label>
<caption><p><bold>Classification matrix for PD patients using healthy subject training data</bold>
.</p>
</caption>
<table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1">Activity</th>
<th align="left" rowspan="1" colspan="1">Walking</th>
<th align="left" rowspan="1" colspan="1">Standing</th>
<th align="left" rowspan="1" colspan="1">Holding</th>
<th align="left" rowspan="1" colspan="1">Sitting</th>
<th align="left" rowspan="1" colspan="1">Not wearing</th>
</tr>
</thead>
<tbody><tr><td align="left" rowspan="1" colspan="1">Walking</td>
<td align="left" rowspan="1" colspan="1"><bold>428</bold>
</td>
<td align="left" rowspan="1" colspan="1">96</td>
<td align="left" rowspan="1" colspan="1">44</td>
<td align="left" rowspan="1" colspan="1">120</td>
<td align="left" rowspan="1" colspan="1">24</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Standing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>308</bold>
</td>
<td align="left" rowspan="1" colspan="1">36</td>
<td align="left" rowspan="1" colspan="1">72</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Holding</td>
<td align="left" rowspan="1" colspan="1">36</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>180</bold>
</td>
<td align="left" rowspan="1" colspan="1">184</td>
<td align="left" rowspan="1" colspan="1">16</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Sitting</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">56</td>
<td align="left" rowspan="1" colspan="1">104</td>
<td align="left" rowspan="1" colspan="1"><bold>276</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Not wearing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">48</td>
<td align="left" rowspan="1" colspan="1">32</td>
<td align="left" rowspan="1" colspan="1"><bold>124</bold>
</td>
</tr>
</tbody>
</table>
<table-wrap-foot><p><italic>60.3% overall accuracy for 18 healthy subjects for training and eight PD patients for testing</italic>
.</p>
</table-wrap-foot>
</table-wrap>
<p>One entry in the table is made for every 10 s sample. The rows represent the activity being performed and the columns represent the prediction according to the algorithm. Overall accuracy is the sum of the correct classification (diagonal in bold) compared to the sum of all entries in the entire table.</p>
<p>We sought to better understand the large difference in accuracy for the within vs. across population results. Ten-fold or leave-one-out cross validation techniques do not remove the effect of the same individual that may have significantly distinct movement pattern from others. Using 10-fold cross validation would still allow such idiosyncratic movements of individuals to be part of the training and test sets. We wanted a measure that would indicate the accuracy of the algorithm if it were applied to subjects after training. For this we performed subject-wise cross validation for the 18 healthy subjects and found an accuracy of 86.0% for SVM’s and 85.2% for SMLR. Note that because much of the variation in movements is across subjects, this accuracy is much lower than that of the 10-fold cross validation. Table <xref ref-type="table" rid="T5">5</xref>
presents a breakdown of the classification for SVM’s, with SMLR appearing similar.</p>
<table-wrap id="T5" position="float"><label>Table 5</label>
<caption><p><bold>Classification matrix for healthy subjects with subject-wise cross validation</bold>
.</p>
</caption>
<table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1">Activity</th>
<th align="left" rowspan="1" colspan="1">Walking</th>
<th align="left" rowspan="1" colspan="1">Standing</th>
<th align="left" rowspan="1" colspan="1">Holding</th>
<th align="left" rowspan="1" colspan="1">Sitting</th>
<th align="left" rowspan="1" colspan="1">Not wearing</th>
</tr>
</thead>
<tbody><tr><td align="left" rowspan="1" colspan="1">Walking</td>
<td align="left" rowspan="1" colspan="1"><bold>900</bold>
</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">4</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Standing</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1"><bold>676</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">36</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Holding</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">88</td>
<td align="left" rowspan="1" colspan="1"><bold>624</bold>
</td>
<td align="left" rowspan="1" colspan="1">40</td>
<td align="left" rowspan="1" colspan="1">44</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Sitting</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">120</td>
<td align="left" rowspan="1" colspan="1">120</td>
<td align="left" rowspan="1" colspan="1"><bold>492</bold>
</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Not wearing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>220</bold>
</td>
</tr>
</tbody>
</table>
<table-wrap-foot><p><italic>86.0% overall accuracy for 18 healthy subjects</italic>
.</p>
</table-wrap-foot>
</table-wrap>
<p>To consider the ability of this approach to be adapted to PD patients by using patient data, we also analyzed the PD patients separately. Similar to the healthy subjects, we applied subject-wise cross validation on the PD patient data alone. Using SVM’s, the accuracy was 75.1% (Table <xref ref-type="table" rid="T6">6</xref>
) and using SMLR it was 76.0%. Although this is lower than the previous percentage for healthy subjects, this is expected as PD patients movements vary more significantly across subjects. Most importantly, when considering predictions across subjects, training using patient data led to a significantly better prediction than training using healthy data alone.</p>
<table-wrap id="T6" position="float"><label>Table 6</label>
<caption><p><bold>Classification matrix for PD patients with subject-wise cross validation</bold>
.</p>
</caption>
<table frame="hsides" rules="groups"><thead><tr><th align="left" rowspan="1" colspan="1">Activity</th>
<th align="left" rowspan="1" colspan="1">Walking</th>
<th align="left" rowspan="1" colspan="1">Standing</th>
<th align="left" rowspan="1" colspan="1">Holding</th>
<th align="left" rowspan="1" colspan="1">Sitting</th>
<th align="left" rowspan="1" colspan="1">Not wearing</th>
</tr>
</thead>
<tbody><tr><td align="left" rowspan="1" colspan="1">Walking</td>
<td align="left" rowspan="1" colspan="1"><bold>588</bold>
</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1">80</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">32</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Standing</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1"><bold>400</bold>
</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1">0</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Holding</td>
<td align="left" rowspan="1" colspan="1">20</td>
<td align="left" rowspan="1" colspan="1">4</td>
<td align="left" rowspan="1" colspan="1"><bold>296</bold>
</td>
<td align="left" rowspan="1" colspan="1">56</td>
<td align="left" rowspan="1" colspan="1">40</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Sitting</td>
<td align="left" rowspan="1" colspan="1">8</td>
<td align="left" rowspan="1" colspan="1">84</td>
<td align="left" rowspan="1" colspan="1">132</td>
<td align="left" rowspan="1" colspan="1"><bold>208</bold>
</td>
<td align="left" rowspan="1" colspan="1">4</td>
</tr>
<tr><td align="left" rowspan="1" colspan="1">Not wearing</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1">56</td>
<td align="left" rowspan="1" colspan="1">0</td>
<td align="left" rowspan="1" colspan="1"><bold>148</bold>
</td>
</tr>
</tbody>
</table>
<table-wrap-foot><p><italic>75.1% overall accuracy for eight PD subjects</italic>
.</p>
</table-wrap-foot>
</table-wrap>
</sec>
<sec sec-type="discussion"><title>Discussion</title>
<p>We applied machine learning to signals from mobile phones to classify the activities of people with PD. Instead of handpicking the most relevant features and comparing them, we used a large feature set and had the relevant features selected by the machine learning algorithms. Because many algorithms depend on the training data, these methods were not expected to test well for populations with unique movement patterns. This was done using mobile phones carried naturally in pants pockets. Though this natural way of carrying was expected to lower accuracy values, it is more indicative of expected accuracy if this research is to be applied to studies with patient groups.</p>
<p>The most accurate methods of data collection for activity recognition rely on multiple sensors. Often this involves accelerometers, as they are small, relatively inexpensive, and register both movement and orientation to gravity. Some systems have integrated temperature, compass, light, and sound sensors on the waist (Choudhury et al., <xref ref-type="bibr" rid="B5">2008</xref>
) or a similar collection of multimodal sensors on the wrist (Maurer et al., <xref ref-type="bibr" rid="B25">2006</xref>
; Gyorbiro et al., <xref ref-type="bibr" rid="B12">2009</xref>
). For accelerometer-only arrays, multiple sensors may be placed throughout the body – anywhere from three to five locations (Bao and Intille, <xref ref-type="bibr" rid="B1">2004</xref>
; Tapia et al., <xref ref-type="bibr" rid="B34">2007</xref>
; Krishnan and Panchanathan, <xref ref-type="bibr" rid="B17">2008</xref>
) or more. Mannini and Sabatini (<xref ref-type="bibr" rid="B22">2010</xref>
) provide a review of these approaches.</p>
<p>There are simpler alternatives to using multiple sensors, improving the convenience, cost, and compliance rates. The most common approach is to use a single, waist-mounted accelerometer. This approach has been analyzed on very specific sets of instructed activities with over 98% accuracy (Mathie et al., <xref ref-type="bibr" rid="B23">2004a</xref>
; Mathie et al., <xref ref-type="bibr" rid="B24">2004b</xref>
; Ravi et al., <xref ref-type="bibr" rid="B28">2005</xref>
; Lee et al., <xref ref-type="bibr" rid="B20">2009</xref>
). High accuracy ratings were possible in part due to the fixed location of the accelerometers on the body, the use of within-subject vs. across-subject cross validation, and the artificial nature of instructed movements. Signals from walking, sitting, and standing are necessarily more repeatable when in a consistent lab setting following instruction. For comparison, when subjects simply wore such a device for 24 h, with more natural activities, accuracy was closer to 80% (Long et al., <xref ref-type="bibr" rid="B21">2009</xref>
). Single, waist-worn accelerometers have been well-studied in the domain of activity recognition, but may need consistent placement for high accuracy.</p>
<p>Unlike dedicated accelerometers, some people already consistently carry mobile phones, making them a convenient platform for recording movements. Most smartphones have built-in accelerometers and are often worn on the person, similar in principle to previous work on accelerometry. Mobile phones have built-in communication protocols that allow simple data logging to the device and wireless transmission. This permits real-time response, or in an experimental setting, compliance verification. Because mobile phones are widely adopted, compliance without verification is already high, as people are used to carrying them. Due to these advantages, mobile phones have the promise to provide a convenient, inexpensive, and objective means to detect the activities of people.</p>
<p>Mobile phones have been used to classify activities of healthy subjects (Bieber et al., <xref ref-type="bibr" rid="B2">2009</xref>
; Brezmes et al., <xref ref-type="bibr" rid="B3">2009</xref>
; Gyorbiro et al., <xref ref-type="bibr" rid="B12">2009</xref>
; Ryder et al., <xref ref-type="bibr" rid="B29">2009</xref>
; Wang et al., <xref ref-type="bibr" rid="B35">2009</xref>
; Yang, <xref ref-type="bibr" rid="B36">2009</xref>
; Kwapisz et al., <xref ref-type="bibr" rid="B19">2011</xref>
). Common activities include walking, jogging/running, standing, sitting, and using stairs. The choice of activities influences accuracy rates, and also because most rates in these previous studies are not subject-wise cross-validated, applicability across subjects is more difficult to interpret. In Kwapisz et al. (<xref ref-type="bibr" rid="B19">2011</xref>
), healthy subjects were instructed to carry the phone in their left pocket and perform a specific set of activities; all activities except stair climbing were classified with at least 90% accuracy. Other studies found similarly high accuracy but with different classification techniques (Brezmes et al., <xref ref-type="bibr" rid="B3">2009</xref>
; Ryder et al., <xref ref-type="bibr" rid="B29">2009</xref>
; Yang, <xref ref-type="bibr" rid="B36">2009</xref>
). In Wang et al. (<xref ref-type="bibr" rid="B35">2009</xref>
), classes were divided as still, walking, running, or in a vehicle, which simplified classification which was done using microphones and GPS as well as accelerometer readings. In Yang (<xref ref-type="bibr" rid="B36">2009</xref>
), a preprocessing technique was used which converted the axes from phone-specific to phone-independent coordinates based on orientation of gravity, providing 88–90% accuracy. While our results on healthy subjects are in line with previous studies, the central contribution of our paper is the careful analysis of precision of activity recognition in the context of PD.</p>
<sec><title>Patient populations</title>
<p>We chose to analyze the PD population for various reasons. Millions of people throughout the world are suffering from diseases that affect mobility. Many diseases, such as stroke, heart disease, or depression affect large populations but have a wide variety of causes, types, and symptoms. PD, on the other hand, is characterized by a number of common characteristics, which makes analysis easier across subjects (Gelb et al., <xref ref-type="bibr" rid="B10">1999</xref>
). Common symptoms such as tremor are visible in movements and lend themselves well to analyses using accelerometers in mobile phones (Joundi et al., <xref ref-type="bibr" rid="B15">2011</xref>
; Surangsrirat and Thanawattano, <xref ref-type="bibr" rid="B33">2012</xref>
). The PD population is also an important subgroup to consider as it also effects a relatively large population – approximately four million people globally (Dorsey et al., <xref ref-type="bibr" rid="B7">2007</xref>
).</p>
<p>There is another study that automatically classified and characterized postures and activities for a population of PD patients (Salarian et al., <xref ref-type="bibr" rid="B30">2007</xref>
). However, their results used within-subject cross validation and thus cannot speak to the across-subject generalization issue we are discussing here. Moreover, they used a set of accelerometers and gyroscopes instead of mobile phones. Our paper demonstrates the ability to use mobile phone recordings of acceleration to enable quality activity recognition with PD patients.</p>
<p>There are a few limitations to the interpretation of our results to address. For our healthy subjects, we used a population of both younger and older subjects, instead of age-matched controls. Some of the difference between the groups can be age-related, however we believe this effect was minor compared to the effect of PD on patient movements. Also, the PD group was relatively small (eight subjects) and heterogeneous (Hoehn and Yahr stage 1–3), however even from this heterogeneous group we note a significant improvement by using PD training data. Lastly, because we had both the researcher and subject observing, we relied on the recording procedure for accurate activity labeling. Subjects did not always perform the instructed actions in a typical fashion (e.g., moving feet while standing, stopping briefly while walking, etc). Instead of removing possible inconsistencies by hand, and thus affecting the validity of this approach, we retained all samples in the data set. Despite these limitations, the main conclusions of this study are supported.</p>
<p>There were two main goals for this study. First, we demonstrate how machine learning can be used to infer the activities of PD populations; the focus is not on particular, hand-picked features of movement, but on automated methods of weighing and combining those features. The second major goal was to highlight and quantify the effect of applying classifiers designed for healthy subjects on a PD patient population. A demonstrable drop in classification accuracy from 92.2 to 60.3% makes this point clear; it is important to use tools and analyses designed for specific patient populations. Although this study is not thorough enough to validate this classification method for clinical practice, it does demonstrate a strong benefit of machine learning, and a caution for clinicians who may want to use any activity recognition methods designed for healthy subjects.</p>
<p>The ultimate objective of therapies is to improve patient quality of life and activity tracking is an additional way of quantifying this. Such quantitative evaluation techniques could help clinicians test and optimize aspects of many therapies for motor disorders. By only downloading an application, mobile phones can record a person’s movements, greatly simplifying the study design and improving compliance. This information can be of personal or community medical use, improving evaluation of patient outcomes in therapeutic interventions. It is clear that populations with motor impairments require special consideration in approaches that analyze movement patterns. Mobile phones provide a means of tracking movements in an objective, convenient, and inexpensive way. The extent to which leveraging these qualities can improve and enable new therapeutic approaches is an area of further research.</p>
</sec>
</sec>
<sec><title>Conflict of Interest Statement</title>
<p>The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.</p>
</sec>
</body>
<back><ack><p>This work was supported in part by the National Parkinson Foundation and the U.S. National Institutes of Health under Grants P01NS044393 and R01NS063399. The Authors would like to thank Andrew Cichowski and Aaron Yang for their help in recruiting subjects and data collection.</p>
</ack>
<ref-list><title>References</title>
<ref id="B1"><mixed-citation publication-type="book"><person-group person-group-type="author"><name><surname>Bao</surname>
<given-names>L.</given-names>
</name>
<name><surname>Intille</surname>
<given-names>S.</given-names>
</name>
</person-group>
(<year>2004</year>
). <article-title>“Activity recognition from user-annotated acceleration data,”</article-title>
in <source>Pervasive Computing</source>
, eds <person-group person-group-type="editor"><name><surname>Ferscha</surname>
<given-names>A.</given-names>
</name>
<name><surname>Mattern</surname>
<given-names>F.</given-names>
</name>
</person-group>
(<publisher-loc>Berlin/Heidelberg</publisher-loc>
: <publisher-name>Springer</publisher-name>
), <fpage>1</fpage>
–<lpage>17</lpage>
</mixed-citation>
</ref>
<ref id="B2"><mixed-citation publication-type="book"><person-group person-group-type="author"><name><surname>Bieber</surname>
<given-names>G.</given-names>
</name>
<name><surname>Voskamp</surname>
<given-names>J. R.</given-names>
</name>
<name><surname>Urban</surname>
<given-names>B.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>“Activity recognition for everyday life on mobile phones. Universal access,”</article-title>
in <source>Human-Computer Interaction. Intelligent and Ubiquitous Interaction Environments</source>
, ed. <person-group person-group-type="editor"><name><surname>Stephanidis</surname>
<given-names>C.</given-names>
</name>
</person-group>
(<publisher-loc>Berlin/Heidelberg</publisher-loc>
: <publisher-name>Springer</publisher-name>
), <fpage>289</fpage>
–<lpage>296</lpage>
</mixed-citation>
</ref>
<ref id="B3"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Brezmes</surname>
<given-names>T.</given-names>
</name>
<name><surname>Gorricho</surname>
<given-names>J.-L.</given-names>
</name>
<name><surname>Cotrina</surname>
<given-names>J.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>“Activity recognition from accelerometer data on a mobile phone,”</article-title>
in <conf-name>IWANN ‘09 Proceedings of the 10th International Work-Conference on Artificial Neural Networks: Part II: Distributed Computing, Artificial Intelligence, Bioinformatics, Soft Computing, and Ambient Assisted Living</conf-name>
, eds <person-group person-group-type="editor"><name><surname>Omatu</surname>
<given-names>S.</given-names>
</name>
<name><surname>Rocha</surname>
<given-names>M.</given-names>
</name>
<name><surname>Bravo</surname>
<given-names>J.</given-names>
</name>
<name><surname>Fernández</surname>
<given-names>F.</given-names>
</name>
<name><surname>Corchado</surname>
<given-names>E.</given-names>
</name>
<name><surname>Bustillo</surname>
<given-names>A.</given-names>
</name>
<name><surname>Corchado</surname>
<given-names>J.</given-names>
</name>
</person-group>
(<conf-loc>Berlin/Heidelberg</conf-loc>
: <conf-sponsor>Springer</conf-sponsor>
), <fpage>796</fpage>
–<lpage>799</lpage>
</mixed-citation>
</ref>
<ref id="B4"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Chang</surname>
<given-names>C.-C.</given-names>
</name>
<name><surname>Lin</surname>
<given-names>C.-J.</given-names>
</name>
</person-group>
(<year>2011</year>
). <article-title>LIBSVM: a library for support vector machines</article-title>
. <source>ACM Trans. Intell. Syst. Technol.</source>
<volume>2</volume>
, <fpage>1</fpage>
–<lpage>27</lpage>
<pub-id pub-id-type="doi">10.1145/1961189.1961199</pub-id>
</mixed-citation>
</ref>
<ref id="B5"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Choudhury</surname>
<given-names>T.</given-names>
</name>
<name><surname>Consolvo</surname>
<given-names>S.</given-names>
</name>
<name><surname>Harrison</surname>
<given-names>B.</given-names>
</name>
<name><surname>Hightower</surname>
<given-names>J.</given-names>
</name>
<name><surname>Lamarca</surname>
<given-names>A.</given-names>
</name>
<name><surname>Legrand</surname>
<given-names>L.</given-names>
</name>
<etal></etal>
</person-group>
(<year>2008</year>
). <article-title>The mobile sensing platform: an embedded activity recognition system</article-title>
. <source>Pervasive Comput. IEEE</source>
<volume>7</volume>
, <fpage>32</fpage>
–<lpage>41</lpage>
<pub-id pub-id-type="doi">10.1109/MPRV.2008.42</pub-id>
</mixed-citation>
</ref>
<ref id="B6"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dibble</surname>
<given-names>L. E.</given-names>
</name>
<name><surname>Addison</surname>
<given-names>O.</given-names>
</name>
<name><surname>Papa</surname>
<given-names>E.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>The effects of exercise on balance in persons with Parkinson’s disease: a systematic review across the disability spectrum</article-title>
. <source>J. Neurol. Phys. Ther.</source>
<volume>33</volume>
, <fpage>14</fpage>
–<lpage>26</lpage>
<pub-id pub-id-type="pmid">19265767</pub-id>
</mixed-citation>
</ref>
<ref id="B7"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dorsey</surname>
<given-names>E. R.</given-names>
</name>
<name><surname>Constantinescu</surname>
<given-names>R.</given-names>
</name>
<name><surname>Thompson</surname>
<given-names>J. P.</given-names>
</name>
<name><surname>Biglan</surname>
<given-names>K. M.</given-names>
</name>
<name><surname>Holloway</surname>
<given-names>R. G.</given-names>
</name>
<name><surname>Kieburtz</surname>
<given-names>K.</given-names>
</name>
<etal></etal>
</person-group>
(<year>2007</year>
). <article-title>Projected number of people with Parkinson disease in the most populous nations, 2005 through 2030</article-title>
. <source>Neurology</source>
<volume>68</volume>
, <fpage>384</fpage>
–<lpage>386</lpage>
<pub-id pub-id-type="doi">10.1212/01.wnl.0000260699.09720.ad</pub-id>
<pub-id pub-id-type="pmid">17082464</pub-id>
</mixed-citation>
</ref>
<ref id="B8"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ellis</surname>
<given-names>T.</given-names>
</name>
<name><surname>Cavanaugh</surname>
<given-names>J. T.</given-names>
</name>
<name><surname>Earhart</surname>
<given-names>G. M.</given-names>
</name>
<name><surname>Ford</surname>
<given-names>M. P.</given-names>
</name>
<name><surname>Foreman</surname>
<given-names>K. B.</given-names>
</name>
<name><surname>Dibble</surname>
<given-names>L. E.</given-names>
</name>
</person-group>
(<year>2011</year>
). <article-title>Which measures of physical function and motor impairment best predict quality of life in Parkinson‚ Äôs disease?</article-title>
<source>Parkinsonism Relat. Disord.</source>
<volume>17</volume>
, <fpage>693</fpage>
–<lpage>697</lpage>
<pub-id pub-id-type="doi">10.1016/j.parkreldis.2011.07.004</pub-id>
<pub-id pub-id-type="pmid">21820940</pub-id>
</mixed-citation>
</ref>
<ref id="B9"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fernandes</surname>
<given-names>H. L.</given-names>
</name>
<name><surname>Albert</surname>
<given-names>M. V.</given-names>
</name>
<name><surname>Kording</surname>
<given-names>K. P.</given-names>
</name>
</person-group>
(<year>2011</year>
). <article-title>Measuring generalization of visuomotor perturbations in wrist movements using mobile phones</article-title>
. <source>PLoS ONE</source>
<volume>6</volume>
, <fpage>e20290</fpage>
<pub-id pub-id-type="doi">10.1371/journal.pone.0020290</pub-id>
<pub-id pub-id-type="pmid">21629659</pub-id>
</mixed-citation>
</ref>
<ref id="B10"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Gelb</surname>
<given-names>D. J.</given-names>
</name>
<name><surname>Oliver</surname>
<given-names>E.</given-names>
</name>
<name><surname>Gilman</surname>
<given-names>S.</given-names>
</name>
</person-group>
(<year>1999</year>
). <article-title>Diagnostic criteria for Parkinson disease</article-title>
. <source>Arch. Neurol.</source>
<volume>56</volume>
, <fpage>33</fpage>
–<lpage>39</lpage>
<pub-id pub-id-type="doi">10.1001/archneur.56.1.33</pub-id>
<pub-id pub-id-type="pmid">9923759</pub-id>
</mixed-citation>
</ref>
<ref id="B11"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Goodwin</surname>
<given-names>V. A.</given-names>
</name>
<name><surname>Richards</surname>
<given-names>S. H.</given-names>
</name>
<name><surname>Taylor</surname>
<given-names>R. S.</given-names>
</name>
<name><surname>Taylor</surname>
<given-names>A. H.</given-names>
</name>
<name><surname>Campbell</surname>
<given-names>J. L.</given-names>
</name>
</person-group>
(<year>2008</year>
). <article-title>The effectiveness of exercise interventions for people with Parkinson’s disease: a systematic review and meta-analysis</article-title>
. <source>Mov. Disord.</source>
<volume>23</volume>
, <fpage>631</fpage>
–<lpage>640</lpage>
<pub-id pub-id-type="doi">10.1002/mds.21922</pub-id>
<pub-id pub-id-type="pmid">18181210</pub-id>
</mixed-citation>
</ref>
<ref id="B12"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Gyorbiro</surname>
<given-names>N.</given-names>
</name>
<name><surname>Fabian</surname>
<given-names>A.</given-names>
</name>
<name><surname>Homanyi</surname>
<given-names>G.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>An activity recognition system for mobile phones</article-title>
. <source>Mobile Netw. Appl.</source>
<volume>14</volume>
, <fpage>82</fpage>
–<lpage>91</lpage>
<pub-id pub-id-type="doi">10.1007/s11036-008-0112-y</pub-id>
</mixed-citation>
</ref>
<ref id="B13"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jankovic</surname>
<given-names>J.</given-names>
</name>
</person-group>
(<year>2008</year>
). <article-title>Parkinson’s disease: clinical features and diagnosis</article-title>
. <source>J. Neurol. Neurosurg. Psychiatr.</source>
<volume>79</volume>
, <fpage>368</fpage>
–<lpage>376</lpage>
<pub-id pub-id-type="doi">10.1136/jnnp.2007.131045</pub-id>
<pub-id pub-id-type="pmid">18344392</pub-id>
</mixed-citation>
</ref>
<ref id="B14"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jeannet</surname>
<given-names>P.-Y.</given-names>
</name>
<name><surname>Aminian</surname>
<given-names>K.</given-names>
</name>
<name><surname>Bloetzer</surname>
<given-names>C.</given-names>
</name>
<name><surname>Najafi</surname>
<given-names>B.</given-names>
</name>
<name><surname>Paraschiv-Ionescu</surname>
<given-names>A.</given-names>
</name>
</person-group>
(<year>2011</year>
). <article-title>Continuous monitoring and quantification of multiple parameters of daily physical activity in ambulatory Duchenne muscular dystrophy patients</article-title>
. <source>Eur. J. Paediatr. Neurol.</source>
<volume>15</volume>
, <fpage>40</fpage>
–<lpage>47</lpage>
<pub-id pub-id-type="doi">10.1016/j.ejpn.2010.07.002</pub-id>
<pub-id pub-id-type="pmid">20719551</pub-id>
</mixed-citation>
</ref>
<ref id="B15"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Joundi</surname>
<given-names>R. A.</given-names>
</name>
<name><surname>Brittain</surname>
<given-names>J.-S.</given-names>
</name>
<name><surname>Jenkinson</surname>
<given-names>N.</given-names>
</name>
<name><surname>Green</surname>
<given-names>A. L.</given-names>
</name>
<name><surname>Aziz</surname>
<given-names>T.</given-names>
</name>
</person-group>
(<year>2011</year>
). <article-title>Rapid tremor frequency assessment with the iPhone accelerometer</article-title>
. <source>Parkinsonism Relat. Disord.</source>
<volume>17</volume>
, <fpage>288</fpage>
–<lpage>290</lpage>
<pub-id pub-id-type="doi">10.1016/j.parkreldis.2011.01.001</pub-id>
<pub-id pub-id-type="pmid">21300563</pub-id>
</mixed-citation>
</ref>
<ref id="B16"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kavanagh</surname>
<given-names>J. J.</given-names>
</name>
<name><surname>Menz</surname>
<given-names>H. B.</given-names>
</name>
</person-group>
(<year>2008</year>
). <article-title>Accelerometry: a technique for quantifying movement patterns during walking</article-title>
. <source>Gait Posture</source>
<volume>28</volume>
, <fpage>1</fpage>
–<lpage>15</lpage>
<pub-id pub-id-type="doi">10.1016/S0966-6362(08)70001-8</pub-id>
<pub-id pub-id-type="pmid">18178436</pub-id>
</mixed-citation>
</ref>
<ref id="B17"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Krishnan</surname>
<given-names>N. C.</given-names>
</name>
<name><surname>Panchanathan</surname>
<given-names>S.</given-names>
</name>
</person-group>
(<year>2008</year>
). <article-title>“Analysis of low resolution accelerometer data for continuous human activity recognition,”</article-title>
in <conf-name>International 28 Conference on, Acoustics, Speech, and Signal Processing, ICASSP 2008 IEEE</conf-name>
, <conf-loc>Las Vegas</conf-loc>
, <fpage>3337</fpage>
–<lpage>3340</lpage>
</mixed-citation>
</ref>
<ref id="B18"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Krishnapuram</surname>
<given-names>B.</given-names>
</name>
<name><surname>Carin</surname>
<given-names>L.</given-names>
</name>
<name><surname>Figueiredo</surname>
<given-names>M. A. T.</given-names>
</name>
<name><surname>Hartemink</surname>
<given-names>A. J.</given-names>
</name>
</person-group>
(<year>2005</year>
). <article-title>Sparse multinomial logistic regression: fast algorithms and generalization bounds</article-title>
. <source>IEEE Trans. Pattern Anal. Mach. Intell.</source>
<volume>27</volume>
, <fpage>957</fpage>
–<lpage>968</lpage>
<pub-id pub-id-type="doi">10.1109/TPAMI.2005.127</pub-id>
<pub-id pub-id-type="pmid">15943426</pub-id>
</mixed-citation>
</ref>
<ref id="B19"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kwapisz</surname>
<given-names>J. R.</given-names>
</name>
<name><surname>Weiss</surname>
<given-names>G. M.</given-names>
</name>
<name><surname>Moore</surname>
<given-names>S. A.</given-names>
</name>
</person-group>
(<year>2011</year>
). <article-title>Activity recognition using cell phone accelerometers</article-title>
. <source>SIGKDD Explor.</source>
<volume>12</volume>
, <fpage>74</fpage>
–<lpage>82</lpage>
<pub-id pub-id-type="doi">10.1145/1964897.1964918</pub-id>
</mixed-citation>
</ref>
<ref id="B20"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Lee</surname>
<given-names>M.</given-names>
</name>
<name><surname>Kim</surname>
<given-names>J.</given-names>
</name>
<name><surname>Kim</surname>
<given-names>K.</given-names>
</name>
<name><surname>Lee</surname>
<given-names>I.</given-names>
</name>
<name><surname>Jee</surname>
<given-names>S. H.</given-names>
</name>
<name><surname>Yoo</surname>
<given-names>S. K.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>“Physical activity recognition using a single tri-axis accelerometer,”</article-title>
in <conf-name>Proceedings of the World Congress on Engineering and Computer Science</conf-name>
, <conf-loc>San Francisco</conf-loc>
</mixed-citation>
</ref>
<ref id="B21"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Long</surname>
<given-names>X.</given-names>
</name>
<name><surname>Yin</surname>
<given-names>B.</given-names>
</name>
<name><surname>Aarts</surname>
<given-names>R. M.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>“Single-accelerometer-based daily physical activity classification,”</article-title>
in <conf-name>Annual International Conference of the IEEE Engineering in Medicine and Biology Society EMBC 2009</conf-name>
, <conf-name>Minneapolis</conf-name>
, <fpage>6107</fpage>
–<lpage>6110</lpage>
<pub-id pub-id-type="pmid">19965261</pub-id>
</mixed-citation>
</ref>
<ref id="B22"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mannini</surname>
<given-names>A.</given-names>
</name>
<name><surname>Sabatini</surname>
<given-names>A. M.</given-names>
</name>
</person-group>
(<year>2010</year>
). <article-title>Machine learning methods for classifying human physical activity from on-body accelerometers</article-title>
. <source>Sensors</source>
<volume>10</volume>
, <fpage>1154</fpage>
–<lpage>1175</lpage>
<pub-id pub-id-type="doi">10.3390/s100201154</pub-id>
<pub-id pub-id-type="pmid">22205862</pub-id>
</mixed-citation>
</ref>
<ref id="B23"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mathie</surname>
<given-names>M. J.</given-names>
</name>
<name><surname>Celler</surname>
<given-names>B. G.</given-names>
</name>
<name><surname>Lovell</surname>
<given-names>N. H.</given-names>
</name>
<name><surname>Coster</surname>
<given-names>A. C.</given-names>
</name>
</person-group>
(<year>2004a</year>
). <article-title>Classification of basic daily movements using a triaxial accelerometer</article-title>
. <source>Med. Biol. Eng. Comput.</source>
<volume>42</volume>
, <fpage>679</fpage>
–<lpage>687</lpage>
<pub-id pub-id-type="doi">10.1007/BF02347551</pub-id>
<pub-id pub-id-type="pmid">15503970</pub-id>
</mixed-citation>
</ref>
<ref id="B24"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mathie</surname>
<given-names>M. J.</given-names>
</name>
<name><surname>Coster</surname>
<given-names>A. C. F.</given-names>
</name>
<name><surname>Lovell</surname>
<given-names>N. H.</given-names>
</name>
<name><surname>Celler</surname>
<given-names>B. G.</given-names>
</name>
</person-group>
(<year>2004b</year>
). <article-title>Accelerometry: providing an integrated, practical method for long-term, ambulatory monitoring of human movement</article-title>
. <source>Physiol. Meas.</source>
<volume>25</volume>
, <fpage>R1</fpage>
<pub-id pub-id-type="doi">10.1088/0967-3334/25/1/001</pub-id>
<pub-id pub-id-type="pmid">15132305</pub-id>
</mixed-citation>
</ref>
<ref id="B25"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Maurer</surname>
<given-names>U.</given-names>
</name>
<name><surname>Smailagic</surname>
<given-names>A.</given-names>
</name>
<name><surname>Siewiorek</surname>
<given-names>D. P.</given-names>
</name>
<name><surname>Deisher</surname>
<given-names>M.</given-names>
</name>
</person-group>
(<year>2006</year>
). <article-title>“Activity recognition and monitoring (using) multiple sensors on different body positions,”</article-title>
in <conf-name>International Workshop on Wearable and Implantable Body Sensor Networks, BSN 2006</conf-name>
, <conf-loc>Aachen</conf-loc>
, <fpage>113</fpage>
–<lpage>116</lpage>
</mixed-citation>
</ref>
<ref id="B26"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Najafi</surname>
<given-names>B.</given-names>
</name>
<name><surname>Aminian</surname>
<given-names>K.</given-names>
</name>
<name><surname>Paraschiv-Ionescu</surname>
<given-names>A.</given-names>
</name>
<name><surname>Loew</surname>
<given-names>F.</given-names>
</name>
<name><surname>Bula</surname>
<given-names>C. J.</given-names>
</name>
<name><surname>Robert</surname>
<given-names>P.</given-names>
</name>
</person-group>
(<year>2003</year>
). <article-title>Ambulatory system for human motion analysis using a kinematic sensor: monitoring of daily physical activity in the elderly</article-title>
. <source>IEEE Trans. Biomed. Eng</source>
. <volume>50</volume>
, <fpage>711</fpage>
–<lpage>723</lpage>
<pub-id pub-id-type="pmid">12814238</pub-id>
</mixed-citation>
</ref>
<ref id="B27"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Palmer</surname>
<given-names>S. S.</given-names>
</name>
<name><surname>Mortimer</surname>
<given-names>J. A.</given-names>
</name>
<name><surname>Webster</surname>
<given-names>D. D.</given-names>
</name>
<name><surname>Bistevins</surname>
<given-names>R.</given-names>
</name>
<name><surname>Dickinson</surname>
<given-names>G. L.</given-names>
</name>
</person-group>
(<year>1986</year>
). <article-title>Exercise therapy for Parkinson’s disease</article-title>
. <source>Arch. Phys. Med. Rehabil.</source>
<volume>67</volume>
, <fpage>741</fpage>
–<lpage>745</lpage>
<pub-id pub-id-type="doi">10.1016/0003-9993(86)90007-9</pub-id>
<pub-id pub-id-type="pmid">3767624</pub-id>
</mixed-citation>
</ref>
<ref id="B28"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Ravi</surname>
<given-names>N.</given-names>
</name>
<name><surname>Dandekar</surname>
<given-names>N.</given-names>
</name>
<name><surname>Mysore</surname>
<given-names>P.</given-names>
</name>
<name><surname>Littman</surname>
<given-names>M.</given-names>
</name>
</person-group>
(<year>2005</year>
). <article-title>“Activity recognition from accelerometer data,”</article-title>
in <conf-name>Proceedings of the Seventeenth Conference on Innovative Applications of Artificial Intelligence</conf-name>
, (<conf-loc>Pittsburgh</conf-loc>
: <conf-sponsor>AAAI Press</conf-sponsor>
), <fpage>1541</fpage>
–<lpage>1546</lpage>
</mixed-citation>
</ref>
<ref id="B29"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Ryder</surname>
<given-names>J.</given-names>
</name>
<name><surname>Longstaff</surname>
<given-names>B.</given-names>
</name>
<name><surname>Reddy</surname>
<given-names>S.</given-names>
</name>
<name><surname>Estrin</surname>
<given-names>D.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>“Ambulation: a tool for monitoring mobility patterns over time using mobile phones,”</article-title>
in <conf-name>International Conference on Computational Science and Engineering CSE‘09</conf-name>
, <conf-loc>Vancouver</conf-loc>
, <fpage>927</fpage>
–<lpage>931</lpage>
</mixed-citation>
</ref>
<ref id="B30"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Salarian</surname>
<given-names>A.</given-names>
</name>
<name><surname>Russmann</surname>
<given-names>H.</given-names>
</name>
<name><surname>Vingerhoets</surname>
<given-names>F. J. G.</given-names>
</name>
<name><surname>Burkhard</surname>
<given-names>P. R.</given-names>
</name>
<name><surname>Aminian</surname>
<given-names>K.</given-names>
</name>
</person-group>
(<year>2007</year>
). <article-title>Ambulatory monitoring of physical activities in patients with Parkinson’s disease</article-title>
. <source>IEEE Trans. Biomed. Eng.</source>
<volume>54</volume>
, <fpage>2296</fpage>
–<lpage>2299</lpage>
<pub-id pub-id-type="doi">10.1109/TBME.2007.896591</pub-id>
<pub-id pub-id-type="pmid">18075046</pub-id>
</mixed-citation>
</ref>
<ref id="B31"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Schenkman</surname>
<given-names>M.</given-names>
</name>
<name><surname>Cutson</surname>
<given-names>T. M.</given-names>
</name>
<name><surname>Kuchibhatla</surname>
<given-names>M.</given-names>
</name>
<name><surname>Chandler</surname>
<given-names>J.</given-names>
</name>
<name><surname>Pieper</surname>
<given-names>C. F.</given-names>
</name>
<name><surname>Ray</surname>
<given-names>L.</given-names>
</name>
<etal></etal>
</person-group>
(<year>1998</year>
). <article-title>Exercise to improve spinal flexibility and function for people with Parkinson’s disease: a randomized, controlled trial</article-title>
. <source>J. Am. Geriatr. Soc.</source>
<volume>46</volume>
, <fpage>1207</fpage>
–<lpage>1216</lpage>
<pub-id pub-id-type="pmid">9777901</pub-id>
</mixed-citation>
</ref>
<ref id="B32"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Stone</surname>
<given-names>A. A.</given-names>
</name>
<name><surname>Shiffman</surname>
<given-names>S.</given-names>
</name>
<name><surname>Schwartz</surname>
<given-names>J. E.</given-names>
</name>
<name><surname>Broderick</surname>
<given-names>J. E.</given-names>
</name>
<name><surname>Hufford</surname>
<given-names>M. R.</given-names>
</name>
</person-group>
(<year>2003</year>
). <article-title>Patient compliance with paper and electronic diaries</article-title>
. <source>Control. Clin. Trials</source>
<volume>24</volume>
, <fpage>182</fpage>
–<lpage>199</lpage>
<pub-id pub-id-type="doi">10.1016/S0197-2456(02)00320-3</pub-id>
<pub-id pub-id-type="pmid">12689739</pub-id>
</mixed-citation>
</ref>
<ref id="B33"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Surangsrirat</surname>
<given-names>D.</given-names>
</name>
<name><surname>Thanawattano</surname>
<given-names>C.</given-names>
</name>
</person-group>
(<year>2012</year>
). <article-title>“Android application for spiral analysis in Parkinson’s Disease,”</article-title>
in <conf-name>Proceedings of IEEE Southeastcon</conf-name>
, <conf-loc>Orlando</conf-loc>
, <fpage>1</fpage>
–<lpage>6</lpage>
</mixed-citation>
</ref>
<ref id="B34"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Tapia</surname>
<given-names>E. M.</given-names>
</name>
<name><surname>Intille</surname>
<given-names>S. S.</given-names>
</name>
<name><surname>Haskell</surname>
<given-names>W.</given-names>
</name>
<name><surname>Larson</surname>
<given-names>K.</given-names>
</name>
<name><surname>Wright</surname>
<given-names>J.</given-names>
</name>
<name><surname>King</surname>
<given-names>A.</given-names>
</name>
<etal></etal>
</person-group>
(<year>2007</year>
). <article-title>“Real-time recognition of physical activities and their intensities using wireless accelerometers and a heart rate monitor,”</article-title>
in <conf-name>Proceedings of the 2007 11th IEEE International Symposium on Wearable Computers. IEEE Computer Society</conf-name>
, <conf-loc>Boston</conf-loc>
</mixed-citation>
</ref>
<ref id="B35"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Wang</surname>
<given-names>Y.</given-names>
</name>
<name><surname>Lin</surname>
<given-names>J.</given-names>
</name>
<name><surname>Annavaram</surname>
<given-names>M.</given-names>
</name>
<name><surname>Jacobson</surname>
<given-names>Q. A.</given-names>
</name>
<name><surname>Hong</surname>
<given-names>J.</given-names>
</name>
<name><surname>Krishnamachari</surname>
<given-names>B.</given-names>
</name>
<etal></etal>
</person-group>
(<year>2009</year>
). <article-title>“A framework of energy efficient mobile sensing for automatic user state recognition,”</article-title>
in <conf-name>Proceedings of the 7th international conference on Mobile systems, applications, and services</conf-name>
, <conf-loc>Krakow, Poland</conf-loc>
: <conf-sponsor>ACM</conf-sponsor>
</mixed-citation>
</ref>
<ref id="B36"><mixed-citation publication-type="confproc"><person-group person-group-type="author"><name><surname>Yang</surname>
<given-names>J.</given-names>
</name>
</person-group>
(<year>2009</year>
). <article-title>“Toward physical activity diary: motion recognition using simple acceleration features with mobile phones,”</article-title>
in <conf-name>Proceedings of the 1st international workshop on Interactive multimedia for consumer electronics</conf-name>
, <conf-loc>Beijing, China</conf-loc>
: <conf-sponsor>ACM</conf-sponsor>
</mixed-citation>
</ref>
</ref-list>
</back>
</pmc>
<affiliations><list><country><li>États-Unis</li>
</country>
</list>
<tree><country name="États-Unis"><noRegion><name sortKey="Albert, Mark V" sort="Albert, Mark V" uniqKey="Albert M" first="Mark" last="Albert">Mark Albert</name>
</noRegion>
<name sortKey="Albert, Mark V" sort="Albert, Mark V" uniqKey="Albert M" first="Mark" last="Albert">Mark Albert</name>
<name sortKey="Kording, Konrad" sort="Kording, Konrad" uniqKey="Kording K" first="Konrad" last="Kording">Konrad Kording</name>
<name sortKey="Kording, Konrad" sort="Kording, Konrad" uniqKey="Kording K" first="Konrad" last="Kording">Konrad Kording</name>
<name sortKey="Shapiro, Mark" sort="Shapiro, Mark" uniqKey="Shapiro M" first="Mark" last="Shapiro">Mark Shapiro</name>
<name sortKey="Shapiro, Mark" sort="Shapiro, Mark" uniqKey="Shapiro M" first="Mark" last="Shapiro">Mark Shapiro</name>
<name sortKey="Toledo, Santiago" sort="Toledo, Santiago" uniqKey="Toledo S" first="Santiago" last="Toledo">Santiago Toledo</name>
</country>
</tree>
</affiliations>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Santé/explor/JankovicV1/Data/Ncbi/Merge
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000385 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Ncbi/Merge/biblio.hfd -nk 000385 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Wicri/Santé |area= JankovicV1 |flux= Ncbi |étape= Merge |type= RBID |clé= PMC:3491315 |texte= Using Mobile Phones for Activity Recognition in Parkinson’s Patients }}
Pour générer des pages wiki
HfdIndexSelect -h $EXPLOR_AREA/Data/Ncbi/Merge/RBID.i -Sk "pubmed:23162528" \ | HfdSelect -Kh $EXPLOR_AREA/Data/Ncbi/Merge/biblio.hfd \ | NlmPubMed2Wicri -a JankovicV1
![]() | This area was generated with Dilib version V0.6.19. | ![]() |