Variables of gait inconsistency outperform basic gait variables in detecting mildy lame cows

Annelies Van Nuffel, Wouter Saeys, Bart Sonck, Jürgen Vangeyte, Koen Mertens, Bart De Ketelaere, Stephanie Van Weyenberg

    Onderzoeksoutput: Bijdrage aan tijdschriftA1: Web of Science-artikelpeer review

    Uittreksel

    Tosupportherdsmeninfindingthelamecowsontheirherds,severalautomatedsystems
    that measurelamenessrelatedcowfeaturessuchasgaitpatterns,arebeingdeveloped.
    Most ofthesesystemsareabletodistinguishbetweennon-lameandseverelylamecows.
    Detecting mildlylamecowsinanearlystageoflamenesshoweverseemschallenging.
    Inspiredbytheapproachusedinhumangaitresearch,newvariablesthatmeasurethe
    inconsistencyinstride-to-stridevariablesweretestedusingcowgaitandwereableto
    show differencesbetweenagroupofnon-lameandagroupofmildlylamecows.Inorder
    to investigatetheaddedvalueoftheseinconsistencyvariablesindetectingmildlylame
    cows,twonewlamenessdetectionmodelswerebuild:oneusingsolelybasicgait
    variablesandasecondmodelusingbothbasicandthenewgaitinconsistencyvariables.
    The secondmodelusingthegaitinconsistencyvariablesoutperformedthemodelbased
    on onlybasicgaitvariablesbyfarindetectingthemildlylamecowswithasensitivityof
    88% andaspecificityof87%.Theseresultssupportthesuggestionofincorporatingsuch
    gait inconsistencyvariablesintolamenessdetectionmodels.Furthervalidationofthese
    gait inconsistencyvariablesshouldbeinvestigatedusinglongitudinalstudieswherecows
    developing lamenessandrecoveringfromitaremonitoreddaily.
    Oorspronkelijke taalEngels
    TijdschriftLivestock Science
    Volume177
    Pagina's (van-tot)125-131
    Aantal pagina’s6
    ISSN1871-1413
    PublicatiestatusGepubliceerd - 2015

    Dit citeren