Automatic early detection of induced colic in horses using accelerometer devices.
Abstract: To seek appropriate veterinary attention for horses with colic, owners must recognise early signs. Direct observation of horse behaviour has several drawbacks: it is time-consuming, hard to see subtle and common behavioural signs, and is based on intuition and subjective decisions. Due to recent advances in wearables and artificial intelligence, it may be possible to develop diagnostic software that can automatically detect colic signs. Objective: To develop a software algorithm to aid in the detection of colic signs and levels of pain. Methods: In vivo experiments. Methods: Transient colic was induced in eight experimental mares with luteolytic doses of prostaglandin. Veterinarians observed the horses before and throughout the interventions and assigned pain scores which were used to separate colic episodes into none (pain score ≤5), level 1 (pain score 6-10) or level 2 (pain score ≥11). Accelerometric data and videos were collected throughout the experiments and using accelerometric data, the horse's behaviour was classified into normal and 10 pain-related behaviours and an activity index was calculated. Models were designed that utilised behaviour and activity index characteristics both detecting the presence of colic and assessing its severity. To determine the accuracy of the model, the ground truth, that is the veterinarians' observation of colic signs and assessment of pain level, was compared with the automatic detection system. Results: The cross-validation analysis demonstrated an accuracy of 91.2% for detecting colic and an accuracy of 93.8% in differentiating between level 1 colic and level 2 colic. The model was able to accurately classify 10 pain-related behaviours and distinguish them from normal behaviour with a high accuracy. Conclusions: We included a limited number of horses with severe pain related behaviours in the dataset. This constraint affects the accuracy of categorising colic severity rather than limiting the algorithms' capacity to identify early colic signs. Conclusions: Our system for early detection of colic in horses is unique and innovative, and it can distinguish between colic of varying severity. Unassigned: Um Pferde mit Koliken angemessen tierärztlich behandeln zu lassen, müssen Tierbesitzer frühe Anzeichen erkennen. Die direkte Beobachtung des Pferdeverhaltens birgt einige Nachteile: sie ist zeitaufwendig, es ist schwierig, subtile Verhaltenssymptome zu erkennen, und oftmals basiert die Beurteilung auf Intuition und subjektiven Entscheidungen. Durch jüngste Fortschritte in Wearables (miniaturisierte Chips und Computer) und künstlicher Intelligenz könnte es möglich werden, diagnostische Software für die automatische Erkennung von Koliksymptomen zu entwickeln. Unassigned: Entwicklung eines Software‐Algorithmus zur Hilfe in der Erkennung von Koliksymptomen und Schmerzlevel. Methods: In vivo Experiment. Methods: Transiente Kolik wurde in 8 Versuchsstuten mit luteolytischen Dosen von Prostaglandin induziert. Tierärzte und Tierärztinnen beobachteten die Pferde vor und während den Interventionen und ermittelten Schmerz‐Scores, welche zur Einteilung der Kolikepisoden in keine (Schmerz‐Score ≤5), Stufe 1, (Schmerz‐Score 6–10) oder Stufe 2 (Schmerz‐Score ≥11) genutzt wurden. Daten der Beschleunigungsmesser und Videos wurden im Verlauf des Experiments gesammelt und, unter Verwendung der Daten der Beschleunigungsmesser, das Pferdeverhalten klassifiziert in normal und zehn schmerzbedingte Verhalten. Ein Aktivitätsindex wurde berechnet. Sowohl das Verhalten als auch die Charakteristika der Aktivitäsindizes wurden für die Entwicklung von Modellen verwendet, welche das Vorhandensein von Kolik und den Ausprägungsgrad ermittelten. Um die Genauigkeit des Modells zu bestimmen, wurden die von Tierärzten und Tierärztinnen beobachteten Koliksymptome und deren Beurteilung des Schmerzlevels mit dem automatischen Erkennungssystem verglichen. Unassigned: Die Keuzvalidierungsanalyse ergab eine Genauigkeit von 91.2% bei der Erkennung von Koliken und eine Genauigkeit von 93.8% bei der Unterscheidung zwischen Koliken der Stufe 1 und der Stufe 2. Das Modell war in der Lage, zehn schmerzbedingte Verhaltensweisen genau zu klassifizieren und sie mit hoher Genauigkeit von normalem Verhalten zu unterscheiden. HAUPTEINSCHRÄNKUNGEN: Eine begrenzte Anzahl von Pferden mit schwerem schmerzbedingtem Verhalten wurde in den Datensatz aufgenommen. Diese Einschränkung wirkt sich eher auf die Genauigkeit der Kategorisierung des Kolikschweregrads aus als auf die Fähigkeit der Algorithmen, frühe Kolikanzeichen zu erkennen. Unassigned: Unser System zur Früherkennung von Koliken bei Pferden ist einzigartig und innovativ, und es kann zwischen Koliken unterschiedlichen Schweregrads unterscheiden.
© 2024 EVJ Ltd.
Publication Date: 2024-02-06 PubMed ID: 38318654DOI: 10.1111/evj.14069Google Scholar: Lookup
The Equine Research Bank provides access to a large database of publicly available scientific literature. Inclusion in the Research Bank does not imply endorsement of study methods or findings by Mad Barn.
- Journal Article
Summary
This research summary has been generated with artificial intelligence and may contain errors and omissions. Refer to the original study to confirm details provided. Submit correction.
This study examines the development of a software algorithm that allows for automatic early detection of colic in horses. The software uses data sourced from wearable accelerometers to identify signs of colic and categorize the level of pain.
Research Methodology
- The research involved in vivo experiments on eight horses whose colic was artificially induced using luteolytic doses of prostaglandin.
- Veterinarians evaluated the horses before and throughout the processes, assigning them pain scores. The researchers used these scores to categorize the colic episodes into three levels: none, level 1, or level 2.
- Throughout the experiments, accelerometric data and videos were collected to evaluate the horses’ behavior.
- Utilizing this accelerometric data, the horses’ behavior was classified into normal and ten different types of pain-related behaviors. An activity index was also calculated to evaluate the horses’ level of physical activity.
- From this rich data set, the researchers developed models that used behavior and activity index characteristics to detect the presence of colic and assess its severity.
Findings
- Results of cross-validation tests demonstrated a 91.2% accuracy rate in colic detection and a 93.8% accuracy rate in differentiating between level 1 and level 2 colic.
- The model was capable of accurately classifying the ten types of pain-related behaviors and differentiating them from normal behavior.
Conclusions and Limitations
- The study included a limited number of horses that exhibited severe pain-related behaviors, which affected the accuracy of the severity categorization more than the algorithm’s capacity to detect early signs of colic.
- Despite this limitation, the study concluded that the system developed is unique and innovative, capable of not only detecting early signs of colic in horses but also distinguishing between varying degrees of severity.
Cite This Article
APA
Eerdekens A, Papas M, Damiaans B, Martens L, Govaere J, Joseph W, Deruyck M.
(2024).
Automatic early detection of induced colic in horses using accelerometer devices.
Equine Vet J, 56(6), 1229-1242.
https://doi.org/10.1111/evj.14069 Publication
Researcher Affiliations
- WAVES-Imec, Department of Information Technology, Ghent University-imec, Ghent, Belgium.
- VETMED, Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium.
- VETMED, Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium.
- WAVES-Imec, Department of Information Technology, Ghent University-imec, Ghent, Belgium.
- VETMED, Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Merelbeke, Belgium.
- WAVES-Imec, Department of Information Technology, Ghent University-imec, Ghent, Belgium.
- WAVES-Imec, Department of Information Technology, Ghent University-imec, Ghent, Belgium.
MeSH Terms
- Animals
- Horses
- Horse Diseases / diagnosis
- Colic / veterinary
- Colic / diagnosis
- Female
- Accelerometry / veterinary
- Accelerometry / instrumentation
- Accelerometry / methods
References
This article includes 24 references
- Blikslager AT. Colic prevention to avoid colic surgery: a surgeon's perspective.. J Equine Vet 2019;76:1–5.
- Bowden A, Boynova P, Brennan ML, England GCW, Mair TS, Furness WA. Retrospective case series to identify the most common conditions seen ‘out‐of‐hours’ by first‐opinion equine veterinary practitioners.. Vet Rec 2020;187:404.
- Rhodes DM, Madrigal R. Management of colic in the field.. Vet Clin N Am: Equine Pract 2021;37:421–439.
- Sutton GA, Dahan R, Turner D, Paltiel O. A behaviour‐based pain scale for horses with acute colic: scale construction.. Vet J 2013;196:394–401.
- Burla J‐B, Ostertag A, Westerath HS, Hillmann E. Gait determination and activity measurement in horses using accelerometer.. Comput Electron Agric 2014;102:127–133.
- Maisonpierre IN, Sutton MA, Harris P, Menzies‐Gow N, Weller R, Pfau T. Accelerometer activity tracking in horses and the effect of pasture management on time budget.. Equine Vet J 2019;51:840–845.
- Hartmann C, Lidauer L, Aurich J, Aurich C, Nagel C. Detection of the time of foaling by accelerometer technique in horses (Equus caballlus)—a pilot study.. Reprod Domest Anim 2018;53:1279–1286.
- Schobesberger H, Peham C. Computerized detection of supporting forelimb lameness in the horse using an artificial neural network.. Vet J 2002;163:77–84.
- Andersen PH, Broomé S, Rashid M, Lunblad J, Ask K, Li Z. Towards machine recognition of facial expressions of pain in horses.. Animals 2021;11:1643.
- Casella E, Khamesi AR, Silvestri S. A framework for the recognition of horse gaits through wearable devices.. Pervasive Mobile Comput 2020;67:101213.
- Eerdekens A, Deruyck M, Fontaine J, Damiaans B, Martens L, De Poorter E. Horse jumping and dressage training activity detection using accelerometer data.. Animals 2021;11:2904.
- D.H.B. b.v., S.D. Paardenkamp and F. Diergeneeskunde, EPWA ‐ Equine Pain and Welfare App. Retrieved from https://www.epwa.nl.
- Van Loon J, Dierendonck M. Monitoring acute equine visceral pain with the equine Utrecht University Scale for Composite Pain Assessment (EQUUS‐COMPASS) and the Equine Utrecht University Scale for Facial Assessment of Pain (EQUUS‐FAP): a scale‐construction study.. Vet J 2015;206:356–364.
- M.P.I. for Psycholinguistics, ELAN. Retrieved from https://archive.mpi.nl/tla/elan.
- Eerdekens A, Deruyck M, Fontaine J, Martens L, De Poorter E, Plets D. A framework for energy‐efficient equine activity recognition with leg accelerometers.. Comput Electron Agric 2021;183:106020.
- Freund Y, Schapire RE. A decision‐theoretic generalization of on‐line learning and an application to boosting.. J Comput Syst Sci 1997;55:119–139.
- Breiman L. Classification and regression trees.. London: Routledge; 2017.
- Geurts P, Ernst D, Wehenkel L. Extremely randomized trees.. Mach Learn 2006;63:3–42.
- Friedman JH. Greedy function approximation: a gradient boosting machine.. Ann Stat 2001;29:1189–1232.
- Breiman L. Random forests.. Mach Learn 2001;45:5–32.
- Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O. Scikit‐learn: machine learning in Python.. J Mach Learn Res 2011;12:2825–2830.
- Ruuska S, Hämäläinen W, Kajava S, Mughal M, Matilainen P, Mononen J. Evaluation of the confusion matrix method in the validation of an automated system for measuring feeding behaviour of cattle.. Behav Processes 2018;148:56–62.
- Anderson K, Morrice‐West AV, Walmsley EA, Fisher AD, Whitton RC, Hitchens PL. Validation of inertial measurement units to detect and predict horse behaviour whilst stabled.. Equine Vet J 2022;55:1128–1138.
- Maskato Y, Dugdale AHA, Singer ER, Kelmer G, Sutton GA. Prospective feasibility and revalidation of the Equine Acute Abdominal Pain Scale (EAAPS) in clinical cases of colic in horses.. Animals 2020;10:2242.
Use Nutrition Calculator
Check if your horse's diet meets their nutrition requirements with our easy-to-use tool Check your horse's diet with our easy-to-use tool
Talk to a Nutritionist
Discuss your horse's feeding plan with our experts over a free phone consultation Discuss your horse's diet over a phone consultation
Submit Diet Evaluation
Get a customized feeding plan for your horse formulated by our equine nutritionists Get a custom feeding plan formulated by our nutritionists