Welfare Monitoring Archives - Tech4Animals

Automated Observations of Dogs’ Resting Behaviour Patterns using Artificial Intelligence and their Similarity to Behavioural Observations

annazam — Fri, 16 Aug 2024 20:12:16 +0000

This paper presents the development and evaluation of an automated computer system, BlyzerDS, which utilizes convolutional neural networks (CNNs) to monitor and analyze dogs’ sleep patterns, aiming to mitigate the time-consuming and error-prone nature of traditional direct behavioral observations in animal welfare research. The system was developed and trained using 13,688 videos of mixed-breed adult dogs recorded over 130 nights. For evaluation, the system’s performance was compared to direct human observations using 6000 previously unseen frames, successfully classifying 5430 of them. Key findings revealed that the automated system achieved an 89% similarity score to manual observations in detecting and quantifying sleep duration and fragmentation. Although no significant difference was found in the percentage of time dogs spent asleep compared to human observers, the automated system recorded more total sleep time, highlighting its capacity to capture more extensive data and address potential limitations like human observer fatigue. The study concludes that this CNN-based system is a reliable and valuable tool that can potentially enhance animal behavior and welfare research by optimizing data collection, increasing precision, and improving replicability, with future potential for adapting to other behaviors and species

The post Automated Observations of Dogs’ Resting Behaviour Patterns using Artificial Intelligence and their Similarity to Behavioural Observations appeared first on Tech4Animals.

Environmental Enrichments and Data-driven Welfare Indicators for Sheltered Dogs using Telemetric Physiological Measures and Signal Processing

annazam — Sat, 17 Aug 2024 08:13:13 +0000

This study examined how different environmental enrichments affect the welfare of eight male shelter dogs in Italy, using telemetric physiological measures to develop objective welfare indicators. Researchers collected heart rate, muscle activity, and body temperature data over 28 days, introducing enrichments including objects, human presence, and female dog companionship. They developed new welfare metrics: predictability (similarity to baseline), sleep quality, and day-night cyclicity. Results showed all enrichments improved welfare, with female companionship being most effective, positively affecting sleep quality and physiological cyclicity. Human presence also showed benefits even with limited interaction, while entertaining objects had less impact. The study provides objective measures for shelter management, emphasizing the importance of social enrichment, particularly with conspecifics, for improving sheltered dogs’ welfare.

The post Environmental Enrichments and Data-driven Welfare Indicators for Sheltered Dogs using Telemetric Physiological Measures and Signal Processing appeared first on Tech4Animals.

BovineTalk: Machine learning for vocalization analysis of dairy cattle under the negative affective state of isolation

annazam — Fri, 16 Aug 2024 20:22:38 +0000

This study addresses the critical need for developing non-invasive, animal-based indicators of affective states in livestock, specifically focusing on vocalizations in dairy cattle. The study’s primary contribution is providing the largest pre-processed dataset to date of vocalizations from 20 lactating adult multiparous dairy cows, collected under a controlled setting during negative affective states induced by visual isolation. Utilizing this dataset, the researchers developed two computational frameworks—a deep learning-based model and an explainable machine learning-based model—for two key tasks: classifying high-frequency (HF) and low-frequency (LF) cattle calls, and identifying individual cows based on their vocalizations. Their models achieved high accuracy in classifying LF and HF calls, reaching 87.2% for the explainable model and 89.4% for the deep learning model, outperforming previous state-of-the-art approaches and exhibiting less overfitting. For individual cow identification, the models demonstrated 68.9% accuracy with the explainable model and 72.5% with the deep learning model, with HF calls containing more individuality information than LF calls. The study also identified important vocal features for these classifications, such as AMvar, AMrate, AMExtent, Formant dispersal, and Wiener entropy mean for call type, and sound duration for individual identification. These results underscore the potential of machine learning approaches in analyzing cattle vocalizations as a valuable tool for assessing emotional valence and informing precision livestock farming practices.

The post BovineTalk: Machine learning for vocalization analysis of dairy cattle under the negative affective state of isolation appeared first on Tech4Animals.

Machine Learning Approaches to Predict and Detect Early-onset of Digital Dermatitis in Dairy Cows using Sensor Data

annazam — Wed, 16 Aug 2023 20:40:56 +0000

This study explored the use of machine learning algorithms based on sensor behavior data for the early-onset detection and prediction of Digital Dermatitis (DD) in dairy cows. Researchers utilized CowManager® ear tags to continuously record data such as activity, eating time, rumination time, and ear temperature from Holstein cows at the Washington State University Knott Dairy Center. The machine learning model, based on the Tree-Based Pipeline Optimization Tool (TPOT), achieved an accuracy of 79% for DD detection on the day clinical signs appeared. For predicting DD 2 days prior to the first clinical signs, a combination of K-means and TPOT reached an accuracy of 64%. The study identified ‘activity’ as the most important sensor feature for DD detection. The proposed models demonstrate the potential for developing real-time automated tools for monitoring and diagnosing DD in lactating dairy cows, suggesting that alterations in behavioral patterns can serve as inputs for early warning systems to improve herd management and animal welfare. While promising, the authors note limitations, including the number of cases and the need for further studies in diverse farming settings with potentially more complex sensor systems.

The post Machine Learning Approaches to Predict and Detect Early-onset of Digital Dermatitis in Dairy Cows using Sensor Data appeared first on Tech4Animals.

BrachySound: Machine learning based assessment of respiratory sounds in dogs

annazam — Wed, 16 Aug 2023 20:52:14 +0000

This study investigated the use of machine learning (ML) models for the objective analysis and diagnosis of Brachycephalic Obstructive Airway Syndrome (BOAS) in dogs, aiming to overcome the subjectivity and time-intensive nature of traditional diagnostic methods like pharyngolaryngeal auscultation. Researchers analyzed 366 audio samples collected from 69 Pugs and 79 other brachycephalic breeds using an electronic stethoscope during a 15-minute standardized exercise test and at rest. The ML models, which included K-Nearest Neighbors (KNN) and Decision Tree classifiers, were developed to classify BOAS test results, predict outcomes from recordings at rest, and detect laryngeal sounds. The results demonstrated significant potential, with models achieving a peak accuracy of 0.85 (85%) for classifying BOAS test results when using combined post-exercise data from Pugs. Predictions based on recordings at rest showed accuracies of 0.68 (68%) for Pugs and 0.65 (65%) for various brachycephalic breeds. Notably, the detection of laryngeal sounds reached an F1 score of 0.80 (80%). These findings suggest that machine learning can streamline the examination process and provide a more objective and efficient approach to canine health assessment, facilitating earlier and more standardized BOAS diagnostics.

The post BrachySound: Machine learning based assessment of respiratory sounds in dogs appeared first on Tech4Animals.

Measurement of Canine Ataxic Gait Patterns using Body-worn Smartphone Sensor Data

annazam — Wed, 17 Aug 2022 08:01:24 +0000

This study explored the use of body-worn smartphone inertial sensors (accelerometer and gyroscope) to objectively detect ataxic gait patterns in dogs. Researchers collected data from 770 walking sessions, comparing 55 healthy dogs to 23 dogs diagnosed with ataxia. By placing an iPhone SE on the dog’s back using an adjustable harness and applying various machine learning techniques, they achieved a 95% accuracy in distinguishing between healthy and ataxic dogs, with the K-nearest neighbors (KNN) technique performing best. This indicates a significant potential for developing smartphone applications for canine ataxia diagnosis and monitoring of treatment effects in both clinical and domestic settings.

The post Measurement of Canine Ataxic Gait Patterns using Body-worn Smartphone Sensor Data appeared first on Tech4Animals.

Evaluation of Shelter Dog Activity Levels Before and During COVID-19 using Automated Analysis

annazam — Sat, 20 Aug 2022 12:24:37 +0000

This paper evaluates the activity levels of shelter dogs before and during the COVID-19 pandemic using automated video analysis within a large, open-admission animal shelter in New York City. The study compared dog activity from two periods: before COVID-19 restrictions (February-March 2020) and during the COVID-19 quarantine (July 2020). Researchers analyzed video clips for “step count” and “activity presence” using a self-developed automated tool, which achieved over 79% accuracy compared to manual coding. The findings revealed that shelter dogs exhibited significantly higher activity levels and more steps within their kennels during the COVID-19 period than before. While activity decreased in the afternoons before COVID-19, it remained at a constant average during the pandemic, with introduced “nap times” showing the lowest activity. The authors suggest these changes are likely due to alterations in the shelter environment caused by COVID-19 restrictions, such as reduced volunteer interaction and playgroups, which may have led to increased in-kennel activity and potentially stress. The study concludes that automated analysis is a suitable and hands-off method for monitoring shelter dog activity.

The post Evaluation of Shelter Dog Activity Levels Before and During COVID-19 using Automated Analysis appeared first on Tech4Animals.

Analysis of Dogs’ Sleep Patterns Using Convolutional Neural Networks

annazam — Tue, 20 Aug 2019 13:40:17 +0000

This paper describes a system for the automatic analysis of sleeping patterns in kenneled dogs, developed as an indicator of their welfare. Addressing the gap in adequate automatic analysis systems for dogs, especially when using low-quality video, the system uniquely combines convolutional neural networks (CNNs) with classical data processing methods. It processes video footage from simple web or security cameras, even with very low quality, and is capable of detecting multiple dogs in a frame. The core tasks of the system involve localizing dogs within each frame and classifying their state as either awake or asleep. An end-to-end architecture, specifically Faster R-CNN ResNet101, is used for detection and classification, followed by a post-processing module that corrects potential errors in localization and classification. This automated solution provides an efficient and accurate way to quantify sleep parameters—such as total sleep amount, sleep interval count, and sleep interval length—for large volumes of video data, which would otherwise be a tedious and error-prone manual task. The research highlights the potential of neural networks to revolutionize animal behavior and welfare science.

The post Analysis of Dogs’ Sleep Patterns Using Convolutional Neural Networks appeared first on Tech4Animals.

Canine Gait Analysis using Inertial Sensors and Deep Learning for Orthopedic and Neurological Disorders

annazam — Wed, 18 Mar 2026 13:26:29 +0000

Distinguishing between orthopedic and neurological gait abnormalities in dogs is notoriously difficult, as subtle movement changes often elude even experienced clinicians during standard visual inspections. To address this diagnostic challenge, this study equipped 29 dogs with wearable inertial sensors to capture precise movement data while they walked and trotted. Researchers then developed a lightweight deep learning model to automatically analyze the raw accelerometer and gyroscope signals. The model proved highly effective, achieving up to 96% accuracy in classifying the subjects as healthy, orthopedic, or neurological. Crucially, the research demonstrated that a single sensor mounted comfortably on the dog’s collar provided optimal classification performance, establishing an accessible and objective diagnostic aid for veterinary clinics.

The post Canine Gait Analysis using Inertial Sensors and Deep Learning for Orthopedic and Neurological Disorders appeared first on Tech4Animals.

AI-based Prediction and Detection of Early-onset of Digital Dermatitis in Dairy Cows using Infrared Thermography

annazam — Sat, 03 Aug 2024 06:54:25 +0000

This study applied deep learning-based computer vision techniques using infrared thermography (IRT) data for the early detection and prediction of Digital Dermatitis (DD) in dairy cows, a common foot disease that negatively impacts animal welfare, milk production, and fertility. The researchers investigated the role of various inputs, including thermal images of cow feet, statistical color features extracted from IRT images, and manually registered temperature values. Their models achieved an accuracy of above 81% for DD detection on ‘day 0’ (the first appearance of clinical signs) and above 70% accuracy for predicting DD two days prior to clinical signs. The findings indicate that combining IRT images with AI-based predictors shows significant potential for developing future real-time automated tools for monitoring DD in dairy cows, which could lead to improved DD management, more rapid treatments, enhanced animal wellbeing, and reduced negative effects on lactation and reproductive performance. While the study demonstrated promising results, it noted limitations such as a suboptimal number of cows, suggesting that an increased number of monitored animals would likely lead to higher performance for both detection and prediction models.

The post AI-based Prediction and Detection of Early-onset of Digital Dermatitis in Dairy Cows using Infrared Thermography appeared first on Tech4Animals.