A Wearable Sensor System for Lameness Detection in Dairy Cattle †
<p>Motion sensor attached to a cow’s hind left leg taken from [<a href="#B4-mti-02-00027" class="html-bibr">4</a>,<a href="#B5-mti-02-00027" class="html-bibr">5</a>] (<b>a</b>) and sensor box (<b>b</b>).</p> "> Figure 2
<p>Plastic block attached to the outer claw of a cow’s left hind hoof taken from our previous work [<a href="#B4-mti-02-00027" class="html-bibr">4</a>] (<b>a</b>) and member of our team walking behind a cow in the indoor stable of the Ludwig Maximilian University (LMU) Munich during the data collection (<b>b</b>).</p> "> Figure 3
<p>Overview of the <span class="html-italic">Training</span> and <span class="html-italic">Detection</span> phases. The features extracted are used during the <span class="html-italic">training phase</span> to train a machine learning model which is used in the <span class="html-italic">detection phase</span> to classify new feature vectors.</p> "> Figure 4
<p>Motion sensor and Eagle schematics for front and back sides of the Printed Circuit Board (PCB) (<b>a</b>) and the orientation of our device (<b>b</b>). Our device is oriented such that the <span class="html-italic">y</span>-axis represents vertical accelerations, the <span class="html-italic">x</span>-axis is parallel to the cow (i.e., in its walking direction) and the <span class="html-italic">z</span>-axis is lateral (i.e., left and right) to a cow.</p> "> Figure 5
<p>Raw acceleration (<b>a</b>) and low-pass filtered acceleration (<b>b</b>) for four strides. Forward movements during a stride cause a positive acceleration along the <span class="html-italic">x</span>-axis. Hoof impacts with the ground can be seen as peaks in the acceleration, particularly along the <span class="html-italic">x</span>- and <span class="html-italic">y</span>-axes. Periods while the hoof is in contact with the ground (still phases) have almost zero acceleration.</p> "> Figure 6
<p>Initial and trimmed segments detected with our stride segmentation algorithm applied to linear acceleration along the <span class="html-italic">x</span>-axis (<b>a</b>) and illustration of the gait features on the linear acceleration along the <span class="html-italic">x</span>-axis (<b>b</b>).</p> "> Figure 7
<p>Green and red dots represent <span class="html-italic">normal</span> and <span class="html-italic">abnormal</span> stride instances, respectively. <math display="inline"><semantics> <mi>ω</mi> </semantics></math> determines the distance of the boundary to the <span class="html-italic">normal</span> stride instrances. <math display="inline"><semantics> <mi>τ</mi> </semantics></math> is a threshold to the distance of stride instances: stride instances with a distance larger than <math display="inline"><semantics> <mi>τ</mi> </semantics></math> are classified as <span class="html-italic">abnormal</span>.</p> "> Figure 8
<p>Visualization of the average classification performance of our approach for a cow walking normal (<b>a</b>) and after it becomes lame (<b>b</b>).</p> ">
Abstract
:1. Introduction
2. Related Work
3. Study Design
4. Requirements
5. System Design
6. Approach
- Data acquisition. The sensor signals are read from the sensor device and stored in memory.
- Preprocessing. The data is organized in chunks and filtered to eliminate noise.
- Stride segmentation. Cow strides are detected and their boundaries identified.
- Feature extraction. Information describing of a cow’s gait is extracted from each stride.
6.1. Sensor Device
6.2. Data Acquisition
6.3. Preprocessing
6.4. Stride Segmentation
- Every stride has two upper peaks. We detect the highest peak with a peak detection algorithm. We ignore peaks that are less than 60 samples away from a previously detected peak. This also filters out periods when cows did not walk.
- Every stride is preceded by periods of small variance in acceleration. We find these periods by searching for the 9-sample window with smallest variance in acceleration among the 70 samples before and after the detected peak. We call the center of these windows initial stride segments.
- Between two initial stride segments, additional samples are included that might not belong to a stride. Therefore, we trim the stride by shifting the initial stride segments towards the peak detected in step 1. The initial stride segments are shifted until the standard deviation of a 6-sample window centered at the shifted stride segment is larger than a constant . We found empirically.
6.5. Feature Extraction
6.5.1. Feature Normalization
6.5.2. Feature Grouping
6.6. Model Training and Classification
7. Evaluation
- True Positive (TP)
- Amount of abnormal stride instances classified as such.
- True Negative (TN)
- Amount of normal stride instances classified as such.
- False Positive (FP)
- Amount of normal stride instances classified as abnormal.
- False Negative (FN)
- Amount of abnormal stride instances classified as normal.
- Accuracy: The ability of our approach to classify stride instances correctly. It answers the question: “what percent of the classified stride instances is correct?”. Accuracy is calculated as: .
- Specificity: The ability of our approach to identify normal stride instances. It answers the question: “when a cow walks normally, what percent of its stride instances does our approach classify as ‘normal’?”. This is also referred to as “true negative rate” and computed as: .
- Sensitivity: The ability of our approach to identify abnormal stride instances. It answers the question: “when a cow walks abnormally, what percent of its stride instances does our approach classify as ‘abnormal’?”. This is also referred to as “true positive rate” and computed as: .
- We trained the SVM algorithm with N-1 normal stride instances, where N is the total number of normal stride instances for a specific cow.
- We used the model to classify the normal stride instance that was not used to train the algorithm and every abnormal stride instance.
- We repeated steps 1 and 2 N times; each time we left out a different normal stride instance.
- We averaged the accuracies, specificities and sensitivities computed in step 3.
7.1. Results
7.2. Discussion
8. Ethical Considerations
- Respecting and caring for every participant without discrimination. The participants of this experiment were cows of different ages and breeds. We did not harm any of the them or make any discrimination as for the selection of the specific cow subjects or treatment they received during the experiment.
- Garnering participants mediated and contingent consent. We conducted this experiment together with a professional veterinarian team who are the legal representatives of the cows that participated in the experiment. Both veterinarians know the needs and welfare requirements of these cows and gave us their consent to conduct the experiment. Furthermore, they accompanied and supported us throughout the entire experiment to ensure these requirements were met.
- Doing research that is relevant to participants and consistent with their welfare. The results of our research suggest that it is possible to automatically detect a condition that is painful for cows and highly detrimental to their health (e.g., might lead to death if not treated early enough). Therefore, our research has the potential to benefit the individual cows that participated in the experiment, as well as other cows. This research was conducted in the natural environment of the participating cows, an indoor stable located in the outskirts of Munich, Germany.
- Avoiding research procedures that may be harmful to participants. According to the veterinarians that supported us throughout this study, attaching a sensor device and plastic block to cows and encouraging them to walk for less than 10 min did not cause any lasting harm to these cows. Veterinarians trimmed cows before attaching the plastic block to ensure the block was placed and fit properly to the claw. Trimming cow claws is a procedure undertaken to maintain a healthy hoof condition and prevent injury and disease. In addition, we limited the walking sessions to a maximum of 10 min per day and continued the data recording on a different day in order to reduce the level of fatigue caused to the cows.
- Assessing research proposals and obtaining expert support. The cow interventions performed in this study were done by professional veterinarians and were approved by the ethics committee of the Ludwig Maximilian University (LMU) in Munich, Germany to ensure no harm was done to the cows.
9. Conclusions
Author Contributions
Conflicts of Interest
References
- Van Nuffel, A.; Zwertvaegher, I.; Van Weyenberg, S.; Pastell, M.; Thorup, V.M.; Bahr, C.; Sonck, B.; Saeys, W. Lameness detection in dairy cows: Part 2. Use of sensors to automatically register changes in locomotion or behavior. Animals 2015, 5, 861–885. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Nielsen, L.R.; Pedersen, A.R.; Herskin, M.S.; Munksgaard, L. Quantifying walking and standing behaviour of dairy cows using a moving average based on output from an accelerometer. Appl. Anim. Behav. Sci. 2010, 127, 12–19. [Google Scholar] [CrossRef]
- Mazrier, H.; Tal, S.; Aizinbud, E.; Bargai, U. A field investigation of the use of the pedometer for the early detection of lameness in cattle. Can. Vet. J. 2006, 47, 883. [Google Scholar] [PubMed]
- Haladjian, J.; Hodaie, Z.; Nüske, S.; Brügge, B. Gait Anomaly Detection in Dairy Cattle. In Proceedings of the Fourth International Conference on Animal-Computer Interaction; ACM: New York, NY, USA, 2017; pp. 8:1–8:8. [Google Scholar] [CrossRef]
- Haladjian, J.; Brügge, B.; Nüske, S. An approach for early lameness detection in dairy cattle. In Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable Computers, Maui, HI, USA, 11–15 September 2017; pp. 53–56. [Google Scholar]
- Mancini, C. Towards an animal-centred ethics for Animal–Computer Interaction. Int. J. Hum.-Comput. Stud. 2017, 98, 221–233. [Google Scholar] [CrossRef]
- Hertem, T.V.; Maltz, E.; Antler, A.; Romanini, C.E.B.; Viazzi, S.; Bahr, C.; Schlageter-Tello, A.; Lokhorst, C.; Berckmans, D.; Halachmi, I. Lameness detection based on multivariate continuous sensing of milk yield, rumination, and neck activity. J. Dairy Sci. 2013, 96, 4286–4298. [Google Scholar] [CrossRef] [PubMed]
- De Mol, R.M.; André, G.; Bleumer, E.J.B.; der Werf, J.T.N.; De Haas, Y.; Van Reenen, C.G. Applicability of day-to-day variation in behavior for the automated detection of lameness in dairy cows. J. Dairy Sci. 2013, 96, 3703–3712. [Google Scholar] [CrossRef] [PubMed]
- Poursaberi, A.; Bahr, C.; Pluk, A.; Van Nuffel, A.; Berckmans, D. Real-time automatic lameness detection based on back posture extraction in dairy cattle: Shape analysis of cow with image processing techniques. Comput. Electron. Agric. 2010, 74, 110–119. [Google Scholar] [CrossRef]
- Song, X.; Leroy, T.; Vranken, E.; Maertens, W.; Sonck, B.; Berckmans, D. Automatic detection of lameness in dairy cattle-Vision-based trackway analysis in cow’s locomotion. Comput. Electron. Agric. 2008, 64, 39–44. [Google Scholar] [CrossRef]
- Pluk, A.; Bahr, C.; Poursaberi, A.; Maertens, W.; Van Nuffel, A.; Berckmans, D. Automatic measurement of touch and release angles of the fetlock joint for lameness detection in dairy cattle using vision techniques. J. Dairy Sci. 2012, 95, 1738–1748. [Google Scholar] [CrossRef] [PubMed]
- Van Hertem, T.; Viazzi, S.; Steensels, M.; Maltz, E.; Antler, A.; Alchanatis, V.; Schlageter-Tello, A.A.; Lokhorst, K.; Romanini, E.C.B.; Bahr, C.; et al. Automatic lameness detection based on consecutive 3D-video recordings. Biosyst. Eng. 2014, 119, 108–116. [Google Scholar] [CrossRef]
- Eddy, A.L.; Van Hoogmoed, L.M.; Snyder, J.R. The role of thermography in the management of equine lameness. Vet. J. 2001, 162, 172–181. [Google Scholar] [CrossRef] [PubMed]
- Pastell, M.; Kujala, M. A Probabilistic Neural Network Model for Lameness Detection. J. Dairy Sci. 2007, 90, 2283–2292. [Google Scholar] [CrossRef] [PubMed]
- Maertens, W.; Vangeyte, J.; Baert, J.; Jantuan, A.; Mertens, K.C.; De Campeneere, S.; Pluk, A.; Opsomer, G.; Van Weyenberg, S.; Van Nuffel, A. Development of a real time cow gait tracking and analysing tool to assess lameness using a pressure sensitive walkway: The GAITWISE system. Biosyst. Eng. 2011, 110, 29–39. [Google Scholar] [CrossRef]
- Chapinal, N.; Tucker, C.B. Validation of an automated method to count steps while cows stand on a weighing platform and its application as a measure to detect lameness. J. Dairy Sci. 2012, 95, 6523–6528. [Google Scholar] [CrossRef] [PubMed]
- Van Nuffel, A.; Vangeyte, J.; Mertens, K.C.; Pluym, L.; De Campeneere, S.; Saeys, W.; Opsomer, G.; Van Weyenberg, S. Exploration of measurement variation of gait variables for early lameness detection in cattle using the GAITWISE. Livest. Sci. 2013, 156, 88–95. [Google Scholar] [CrossRef]
- Der Tol, P.P.J.; Metz, J.H.M.; Noordhuizen-Stassen, E.N.; Back, W.; Braam, C.R.; Weijs, W.A. The pressure distribution under the bovine claw during square standing on a flat substrate. J. Dairy Sci. 2002, 85, 1476–1481. [Google Scholar] [CrossRef]
- Thorup, V.M.; Munksgaard, L.; Robert, P.E.; Erhard, H.W.; Thomsen, P.T.; Friggens, N.C. Lameness detection via leg-mounted accelerometers on dairy cows on four commercial farms. Animal 2015, 9, 1704–1712. [Google Scholar] [CrossRef] [PubMed]
- Alsaaod, M.; Römer, C.; Kleinmanns, J.; Hendriksen, K.; Rose-Meierhöfer, S.; Plümer, L.; Büscher, W. Electronic detection of lameness in dairy cows through measuring pedometric activity and lying behavior. Appl. Anim. Behav. Sci. 2012, 142, 134–141. [Google Scholar] [CrossRef]
- Yunta, C.; Guasch, I.; Bach, A. Short communication: Lying behavior of lactating dairy cows is influenced by lameness especially around feeding time. J. Dairy Sci. 2012, 95, 6546–6549. [Google Scholar] [CrossRef] [PubMed]
- Pastell, M.; Tiusanen, J.; Hakojärvi, M.; Hänninen, L. A wireless accelerometer system with wavelet analysis for assessing lameness in cattle. Biosyst. Eng. 2009, 104, 545–551. [Google Scholar] [CrossRef]
- Chapinal, N.; De Passillé, A.M.; Rushen, J.; Wagner, S. Automated methods for detecting lameness and measuring analgesia in dairy cattle. J. Dairy Sci. 2010, 93, 2007–2013. [Google Scholar] [CrossRef] [PubMed]
- Chapinal, N.; de Passille, A.M.; Pastell, M.; Hänninen, L.; Munksgaard, L.; Rushen, J. Measurement of acceleration while walking as an automated method for gait assessment in dairy cattle. J. Dairy Sci. 2011, 94, 2895–2901. [Google Scholar] [CrossRef] [PubMed]
- Haladjian, J.; Ermis, A.; Hodaie, Z.; Brügge, B. iPig: Towards Tracking the Behavior of Free-roaming Pigs. In Proceedings of the Fourth International Conference on Animal-Computer Interaction; Milton Keynes, UK, 21–23 November 2017; pp. 10:1–10:5. [Google Scholar] [CrossRef]
- Paci, P.; Mancini, C.; Price, B.A. Towards a wearer-centred framework for animal biotelemetry. In Proceedings of the Measuring Behaviour 2016, Dublin, Ireland, 25–27 May 2016. [Google Scholar]
- Cola, G.; Avvenuti, M.; Vecchio, A.; Yang, G.Z.; Lo, B. An on-node processing approach for anomaly detection in gait. IEEE Sens. J. 2015, 15, 6640–6649. [Google Scholar] [CrossRef]
- Abbate, S.; Avvenuti, M.; Bonatesta, F.; Cola, G.; Corsini, P.; Vecchio, A. A smartphone-based fall detection system. Pervasive Mob. Comput. 2012, 8, 883–899. [Google Scholar] [CrossRef]
Demographics | Data Collected | |||||
---|---|---|---|---|---|---|
# | Age (Years) | Weight (Kg) | Race (DH/FV) | Normal (Minutes/Strides) | Right (Minutes/Strides) | Left (Minutes/Strides) |
1 | 5 | 910 | 31.25/68.75 | 42.3/1136 | 6.1/171 | 6.2/212 |
2 | 5 | 680 | 68.75/31.25 | 28.8/985 | 7.1/181 | 8.2/231 |
3 | 5 | 720 | 43.75/56.25 | 37.7/1166 | 6.5/202 | 5.4/162 |
4 | 7 | 560 | 87.5/12.5 | 5.5/242 | 7.6/290 | 7.8/256 |
5 | 6 | 700 | 31.25/68.75 | 48.8/1276 | 11.9/244 | 12.1/355 |
6 | 4 | 780 | 62.5/37.5 | 20.1/643 | 5.8/191 | 6.2/208 |
7 | 3 | 680 | 0/100 | 26.3/888 | 6.1/250 | 6.4/261 |
8 | 5 | 640 | 100/0 | 44.5/1378 | 8.2/300 | 8.1/310 |
9 | 4 | 610 | 0/100 | 38.8/1410 | 6.1/203 | 7.6/241 |
10 | 3 | 700 | 0/100 | 24.7/824 | 7.8/227 | 7.2/208 |
Feature | Signal | # | |
---|---|---|---|
Gait | peak values | accel | 9 |
rise times | accel | 9 | |
stride duration | accel | 3 | |
Statistical | mean | all | 7 |
median | all | 7 | |
STD | all | 7 | |
ZCR | accel | 3 | |
P2P | all | 7 | |
RMS | all | 7 | |
AAV | all | 7 | |
Total | 66 |
# | Accuracy | Specificity | Sensitivity |
---|---|---|---|
1 | 95.8% | 96.3% | 83.3% |
2 | 94.6% | 95.1% | 78.6% |
3 | 81.9% | 82.0% | 78.6% |
4 | 96.4% | 97.2% | 66.7% |
5 | 97.3% | 97.6% | 80.0% |
6 | 81.3% | 81.7% | 70.6% |
7 | 95.4% | 96.6% | 62.5% |
8 | 92.6% | 93.0% | 77.8% |
9 | 87.2% | 87.6% | 73.1% |
10 | 88.2% | 88.7% | 70.6% |
© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Haladjian, J.; Haug, J.; Nüske, S.; Bruegge, B. A Wearable Sensor System for Lameness Detection in Dairy Cattle. Multimodal Technol. Interact. 2018, 2, 27. https://doi.org/10.3390/mti2020027
Haladjian J, Haug J, Nüske S, Bruegge B. A Wearable Sensor System for Lameness Detection in Dairy Cattle. Multimodal Technologies and Interaction. 2018; 2(2):27. https://doi.org/10.3390/mti2020027
Chicago/Turabian StyleHaladjian, Juan, Johannes Haug, Stefan Nüske, and Bernd Bruegge. 2018. "A Wearable Sensor System for Lameness Detection in Dairy Cattle" Multimodal Technologies and Interaction 2, no. 2: 27. https://doi.org/10.3390/mti2020027