Veterinärmedizinische Universität Wien Forschungsinformationssystem VetDoc

Grafischer Link zur Startseite der Vetmeduni Vienna

Gewählte Publikation:

Publikationstyp: Zeitschriftenaufsatz
Dokumenttyp: Originalarbeit

Jahr: 2012

AutorInnen: Mangweth, G; Schramel, JP; Peham, C; Gasser, C; Tichy, A; Altenhofer, C; Weber, A; Kofler, J

Titel: [Lameness detection in cows by accelerometric measurement of motion at walk].

Titelvariante: Lameness detection in cows by accelerometric measurement of motion at walk

Quelle: Berl Munch Tierarztl Wochenschr. 2012; 125(9-10):386-396

Autor/innen der Vetmeduni Vienna:

Kofler Johann
Peham Christian
Schramel Johannes
Tichy Alexander

Beteiligte Vetmed-Organisationseinheiten
Universitätsklinik für Pferde, Klinische Abteilung für Interne Medizin Pferde
Universitätsklinik für Pferde, Klinische Abteilung für Pferdechirurgie
Plattform Bioinformatik und Biostatistik

Locomotion scoring (Sprecher et al., 1997) and measurements with a tri-axial accelerometric device were performed in 15 cows before and after digit amputation (Group 1_AMP) as well as in 26 lame cows with different lameness scores (Group 1_LAHM). The evaluated data of Group 1_AMP, documented directly before and on days 2, 5, 10 and 14 after amputation were compared with reference values of 16 non-lame cows (Group 2). For this purpose the Root Mean Square (RMS) as well as the mean minimum (gMin) and maximum (gMax) acceleration were calculated for each measurement (n = 117). Basing on this data, forecast models were performed to predict lameness scores as well as to differentiate between lame and non-lame cows in order to verify the suitability of accelerometry as a technical tool for automated lameness detection in cows. The preoperative locomotion score in Group 1_AMP averaged 4.2. During the period of convalescence, lameness improved significantly (p < 0.01) to a mean score of 1.9 on day 14 after amputation. Statistically, score 1 can be expected on day 20 after surgery (95% CI). On day 35 an amputated cow can be assumed to be non-lame with aprobability of 95%. Acceleration values showed different approximation to those recorded for the reference group. Forecast models enabled prediction of determined scores of lameness with each category of measurement (RMS, gMin, gMax) as well as considering the entire data set with an accuracy of up to 61.7%. Differentiation between lame and non-lame cows was successful with a percentage of up to 91.7%, depending on the applied data. The results of this study showed that accelerometry is a suitable technical tool for automated lameness detection in cows, especially using the acceleration values (RMS, gMin, gMax) and forecast models employed in this study.

Keywords Pubmed: Accelerometry/methods
Cattle Diseases/diagnosis*
Cattle Diseases/physiopathology
Lameness, Animal/diagnosis*
Lameness, Animal/physiopathology
Monitoring, Physiologic/veterinary
Signal Processing, Computer-Assisted

© Veterinärmedizinische Universität Wien Hilfe und Downloads