Improving lameness detection in cows: A machine learning algorithm application

ABSTRACT: The deployment of diverse data-generating technologies in livestock farming holds the promise of early disease detection and improved animal well-being. In this paper, we combine routinely collected dairy farm and herd data with weather and high-frequency sensor data from 6 farms to predic...

Full description

Saved in:
Bibliographic Details
Main Authors: Elma Dervić, Caspar Matzhold, Christa Egger-Danner, Franz Steininger, Peter Klimek
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Journal of Dairy Science
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S0022030224011457
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850220603265515520
author Elma Dervić
Caspar Matzhold
Christa Egger-Danner
Franz Steininger
Peter Klimek
author_facet Elma Dervić
Caspar Matzhold
Christa Egger-Danner
Franz Steininger
Peter Klimek
author_sort Elma Dervić
collection DOAJ
description ABSTRACT: The deployment of diverse data-generating technologies in livestock farming holds the promise of early disease detection and improved animal well-being. In this paper, we combine routinely collected dairy farm and herd data with weather and high-frequency sensor data from 6 farms to predict new lameness events in various future periods, spanning from the following day to 3 wk. A Random Forest classifier, using input features selected by the Boruta algorithm, was used for the prediction task; effects of individual features were further assessed using partial dependence plots. We achieve precision scores of up to 93% when predicting lameness for the next 3 wk and when using information from the last 3 wk, combined with a balanced accuracy of 79%. Removing sensor data results has a tendency to reduce the precision for predictions, especially when using information from the last 1, 2, or 3 wk. Moving to a larger dataset (without sensor data) of 44 farms keeps the similar balanced accuracy but reduces precision by more than 30%, revealing a substantial a trade-off in model quality between false positives (false lameness alerts) and false negatives (missed lameness events). Sensor data holds promise to further improve the precision of these models, but can be partially compensated by high-resolution data from other systems, such as automated milking systems.
format Article
id doaj-art-2413d00f1eed49399d60fcf24645ea80
institution OA Journals
issn 0022-0302
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Journal of Dairy Science
spelling doaj-art-2413d00f1eed49399d60fcf24645ea802025-08-20T02:07:00ZengElsevierJournal of Dairy Science0022-03022024-12-0110712115501156210.3168/jds.2024-24730Improving lameness detection in cows: A machine learning algorithm applicationElma Dervić0Caspar Matzhold1Christa Egger-Danner2Franz Steininger3Peter Klimek4Complexity Science Hub Vienna, 1080 Vienna, Austria; Supply Chain Intelligence Institute Austria, 1080 Vienna, Austria; Medical University of Vienna, Section for Science of Complex Systems, CeMSIIS, 1090 Vienna, AustriaMedical University of Vienna, Section for Science of Complex Systems, CeMSIIS, 1090 Vienna, Austria; ZuchtData EDV-Dienstleistungen GmbH, 1200 Vienna, AustriaZuchtData EDV-Dienstleistungen GmbH, 1200 Vienna, AustriaZuchtData EDV-Dienstleistungen GmbH, 1200 Vienna, AustriaComplexity Science Hub Vienna, 1080 Vienna, Austria; Supply Chain Intelligence Institute Austria, 1080 Vienna, Austria; Medical University of Vienna, Section for Science of Complex Systems, CeMSIIS, 1090 Vienna, Austria; Corresponding authorABSTRACT: The deployment of diverse data-generating technologies in livestock farming holds the promise of early disease detection and improved animal well-being. In this paper, we combine routinely collected dairy farm and herd data with weather and high-frequency sensor data from 6 farms to predict new lameness events in various future periods, spanning from the following day to 3 wk. A Random Forest classifier, using input features selected by the Boruta algorithm, was used for the prediction task; effects of individual features were further assessed using partial dependence plots. We achieve precision scores of up to 93% when predicting lameness for the next 3 wk and when using information from the last 3 wk, combined with a balanced accuracy of 79%. Removing sensor data results has a tendency to reduce the precision for predictions, especially when using information from the last 1, 2, or 3 wk. Moving to a larger dataset (without sensor data) of 44 farms keeps the similar balanced accuracy but reduces precision by more than 30%, revealing a substantial a trade-off in model quality between false positives (false lameness alerts) and false negatives (missed lameness events). Sensor data holds promise to further improve the precision of these models, but can be partially compensated by high-resolution data from other systems, such as automated milking systems.http://www.sciencedirect.com/science/article/pii/S0022030224011457data integrationdisease predictionmachine learningprecision livestock farminglameness
spellingShingle Elma Dervić
Caspar Matzhold
Christa Egger-Danner
Franz Steininger
Peter Klimek
Improving lameness detection in cows: A machine learning algorithm application
Journal of Dairy Science
data integration
disease prediction
machine learning
precision livestock farming
lameness
title Improving lameness detection in cows: A machine learning algorithm application
title_full Improving lameness detection in cows: A machine learning algorithm application
title_fullStr Improving lameness detection in cows: A machine learning algorithm application
title_full_unstemmed Improving lameness detection in cows: A machine learning algorithm application
title_short Improving lameness detection in cows: A machine learning algorithm application
title_sort improving lameness detection in cows a machine learning algorithm application
topic data integration
disease prediction
machine learning
precision livestock farming
lameness
url http://www.sciencedirect.com/science/article/pii/S0022030224011457
work_keys_str_mv AT elmadervic improvinglamenessdetectionincowsamachinelearningalgorithmapplication
AT casparmatzhold improvinglamenessdetectionincowsamachinelearningalgorithmapplication
AT christaeggerdanner improvinglamenessdetectionincowsamachinelearningalgorithmapplication
AT franzsteininger improvinglamenessdetectionincowsamachinelearningalgorithmapplication
AT peterklimek improvinglamenessdetectionincowsamachinelearningalgorithmapplication