Use of thermal imaging in dairy calves: exploring the repeatability and accuracy of measures taken from different anatomical regions

Comparative analysis of computer vision algorithms for the real-time detection of digital dermatitis in dairy cows

Digital dermatitis (DD) is a bovine claw disease responsible for ulcerative lesions on the planar aspect of the hoof. DD is associated with massive herd outbreaks of lameness and influences cattle welfare and production. Early detection of DD can lead to prompt treatment and decrease lameness. Computer vision (CV) provides a unique opportunity to improve early detection. The study aims to train and compare applications for the real-time detection of DD in dairy cows. Eight CV models were trained for detection and scoring, compared using performance metrics and inference time, and the best model was automated for real-time detection using images and video. Images were collected from commercial dairy farms while facing the interdigital space on the plantar surface of the foot. Images were scored for M-stages of DD by a trained investigator using the M-stage DD classification system with distinct labels for hyperkeratosis (H) and proliferations (P). Two sets of images were compiled: the first dataset (Dataset 1) containing 1,177 M0/M4H and 1,050 M2/M2P images and the second dataset (Dataset 2) containing 240 M0, 17 M2, 51 M2P, 114 M4H, and 108 M4P images. Models were trained to detect and score DD lesions and compared for precision, recall, and mean average precision (mAP) in addition to inference time in frame per second (FPS). Seven of the nine CV models performed well compared to the ground truth of labeled images using Dataset 1. The six models, Faster R-CNN, Cascade R-CNN, YOLOv3, Tiny YOLOv3, YOLOv4, Tiny YOLOv4, and YOLOv5s achieved an mAP between 0.964 and 0.998, whereas the other two models, SSD and SSD Lite, yielded an mAP of 0.371 and 0.387 respectively. Overall, YOLOv4, Tiny YOLOv4, and YOLOv5s outperformed all other models with almost perfect precision, perfect recall, and a higher mAP. Tiny YOLOv4 outperformed all other models with respect to inference time at 333 FPS, followed by YOLOv5s at 133 FPS and YOLOv4 at 65 FPS. YOLOv4 and Tiny YOLOv4 performed better than YOLOv5s compared to the ground truth using Dataset 2. YOLOv4 and Tiny YOLOv4 yielded a similar mAP of 0.896 and 0.895, respectively. However, Tiny YOLOv4 achieved both higher precision and recall compared to YOLOv4. Finally, Tiny YOLOv4 was able to detect DD lesions on a commercial dairy farm with high performance and speed. The proposed CV tool can be used for early detection and prompt treatment of DD in dairy cows. This result is a step towards applying CV algorithms to veterinary medicine and implementing real-time DD detection on dairy farms.

Pasture access and eye temperature in dairy cows

Pasture access can benefit dairy cows' behavior, health, and welfare, but herds are increasingly housed indoors full-time. Recent infrared thermal-imaging (thermography) studies suggest that higher eye temperatures may be a physiological indicator of chronic stress. We, therefore, hypothesized that, compared to cows with pasture access, cows housed indoors full-time would have higher eye temperatures. In a two-phase crossover experiment, 29 Holstein-Friesian dairy cows experienced 18 days of overnight pasture access and 18 days of full-time indoor housing. We measured each animal's eye temperature 16 times (eight/phase). During Phase One, cows with pasture access had higher eye temperatures than cows housed indoors full-time (contrary to our hypothesis). However, during Phase Two, cows with pasture access had lower eye temperatures than cows housed indoors full-time. It is, therefore, unclear whether eye temperature reflected disparities in dairy cow welfare between different housing treatments.

Association between a pyroelectric infrared sensor monitoring system and a 3-dimensional accelerometer to assess movement in preweaning dairy calves

The objective of this study was to correlate movement assessed by a pyroelectric infrared sensor system in preweaning dairy calves with lying and standing time assessed by a 3D accelerometer considering the temperature-humidity index (THI). A total of 35 dairy calves (1-7 d of age) were enrolled in the study and 20 calves were included in the final analyses. The lying and standing time of the calves was monitored with a 3D accelerometer (Hobo Pendant G Data Logger, Onset Computer Corporation, USA), which was used as the gold standard reference. The infrared sensor monitoring system (IMS; Calf Monitoring System, Futuro Farming GmbH, Germany) was fixed to the fence of the calf hutch within the calf's reach. Temperature-humidity was monitored with 2 validated THI sensors inside and on outside of each calf hutch. Additionally, one THI sensor was located near the calf hutches. The observation period lasted 14 consecutive days. The average standing time assessed by the 3D accelerometer was 13.4 ± 12.7 (mean ± standard deviation) min/h and the average lying time was 46.6 (±12.7) min/h. The median (25th percentile; 75th percentile) number of movements measured by the IMS was 360 (60; 919) movements per hour. Number of movements per hour measured by the IMS was compared with data obtained with a validated 3D accelerometer. The Pearson correlation coefficient between both standing and lying time and the number of movements was r = 0.85 and r = -0.85, respectively. The Pearson correlation coefficients were only slightly influenced by THI (low THI [<68]: r = 0.86; medium THI [68-72]: r = 0.85; high THI [>72]: r = 0.81). Our data show that the number of movements of dairy calves measured by IMS were highly correlated with the chosen gold standard reference method. High THI slightly affects the measurement accuracy of IMS.

Literature Review on Technological Applications to Monitor and Evaluate Calves' Health and Welfare

Precision livestock farming (PLF) research is rapidly increasing and has improved farmers' quality of life, animal welfare, and production efficiency. PLF research in dairy calves is still relatively recent but has grown in the last few years. Automatic milk feeding systems (AMFS) and 3D accelerometers have been the most extensively used technologies in dairy calves. However, other technologies have been emerging in dairy calves' research, such as infrared thermography (IRT), 3D cameras, ruminal bolus, and sound analysis systems, which have not been properly validated and reviewed in the scientific literature. Thus, with this review, we aimed to analyse the state-of-the-art of technological applications in calves, focusing on dairy calves. Most of the research is focused on technology to detect and predict calves' health problems and monitor pain indicators. Feeding and lying behaviours have sometimes been associated with health and welfare levels. However, a consensus opinion is still unclear since other factors, such as milk allowance, can affect these behaviours differently. Research that employed a multi-technology approach showed better results than research focusing on only a single technique. Integrating and automating different technologies with machine learning algorithms can offer more scientific knowledge and potentially help the farmers improve calves' health, performance, and welfare, if commercial applications are available, which, from the authors' knowledge, are not at the moment.

Infrared thermography as a tool for the measurement of negative emotions in dairy cows

In commercial dairy cows, the conditions in which they are kept may lead to negative emotional states associated with the development of chronic physiological and behavioural abnormalities that may compromise their health, welfare and productivity. Such states include fear, stress or anxiety. Behavioural rather than physiological tests are more likely to be used to indicate these states but can be limited by their subjectivity, need for specialised infrastructure and training (of the operator and sometimes the animal) and the time-consuming nature of data collection. Popularly used physiological measures such as blood cortisol may be more appropriate for acute rather than chronic assessments but are easily confounded, for example by a response to the act of measurement per se. More sophisticated physiological measures such as functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) may be impractical due to cost and time and, like blood cortisol, have the confounding associated with the act of measurement. By contrast, infrared thermography of external body surfaces is remote, non-invasive, easily repeated and follows an objective methodology, allowing longitudinal data acquisition for the inference of changes in chronic emotional state over time. The objective of this review was to investigate the potential of infrared thermography to measure cow emotions. In lactating dairy cows, maximum IRT of the eyes and coronary band of the limbs seem to be most representative of thermoregulatory changes, which are repeatable and correlate with behavioural and physiological indicators of emotional state. IRT methodologies have the potential to become a fundamental tool for the objective assessment of welfare state in dairy cows.

Evaluation of an infrared thermography camera for measuring body temperature in dairy calves

The objective of this diagnostic accuracy study was to validate an infrared thermography (IRT) camera and its software (FLIR One, FLIR, Global) for accuracy and precision for ocular temperature readings to serve as a proxy for rectal temperature in commercially housed calves. A total of 318 male Holstein calves were enrolled into this study from the day of arrival to a calf rearing facility until 14 d later. Researchers took an ocular temperature reading using an IRT camera, and a rectal temperature on each calf each day in the morning. The reference standard method for body temperature in the calves was rectal temperature. We assessed the data for agreement between the IRT and the reference standard using Pearson correlations by calf (accuracy), coefficients of determination (precision), and Bland-Altman plots for bias. In addition, a logistic regression model was built using the reference method as the outcome, with IRT as an explanatory variable to assess the diagnostic accuracy of IRT as an indicator of fever (rectal temperature ≥39.5°C). There was a negligible correlation between the IRT readings and rectal temperature (r = 0.22) and the coefficient of determination for IRT to predict rectal temperature was negligible (R² = 0.05), suggesting poor precision. The average mean difference between the IRT data and rectal temperature was 0.55°C, and the differences between IRT and rectal formed a linear line around the mean difference, suggesting the Bland-Altman analyses showed proportional error and bias. The optimal probability cut-off for IRT readings for fever was at 39.5°C, and had a receiver operating characteristic area under the curve of 0.67, a sensitivity of 61%, a specificity of 71%, and 78% (3,134/4,427) of the samples were correctly labeled as either having a fever or not using IRT readings. In summary, the IRT camera and software were not validated for serving as a proxy for rectal temperature in commercially housed calves due to poor precision, and proportional error partially explained by ambient environmental conditions. We suggest that this infrared thermography system should not replace rectal temperature readings for use in commercially housed calves.

Thermography for disease detection in livestock: A scoping review

Infra-red thermography (IRT) offers potential opportunities as a tool for disease detection in livestock. Despite considerable research in this area, there are no common standards or protocols for managing IRT parameters in animal disease detection research. In this review, we investigate parameters that are essential to the progression of this tool and make recommendations for their use based on the literature found and the veterinary thermography guidelines from the American Academy of Thermology. We analyzed a defined set of 109 articles concerned with the use of IRT in livestock related to disease and from these articles, parameters for accurate IRT were identified and sorted into the fields of camera-, animal- or environment-related categories to assess the practices of each article in reporting parameters. This review demonstrates the inconsistencies in practice across peer-reviewed articles and reveals that some important parameters are completely unreported while others are incorrectly captured and/or under-represented in the literature. Further to this, our review highlights the lack of measured emissivity values for live animals in multiple species. We present guidelines for the standards of parameters that should be used and reported in future experiments and discuss potential opportunities and challenges associated with using IRT for disease detection in livestock.

Comparison of Low- and High-Cost Infrared Thermal Imaging Devices for the Detection of Lameness in Dairy Cattle

Simple Summary

Lameness has high economic and welfare cost to the U.K. dairy industry; accurate and early detection of lameness minimises this cost. Thermal imaging devices can be used as a method of detecting lameness; however, these devices are typically high-cost and fragile, limiting their usefulness in a farm setting. This study looked at the effectiveness of low-cost thermal imaging devices when used as lameness detection aids, by comparing one to a research-specification thermal imaging device. Thermal images were taken of cattle feet, and each cow was assessed for lameness. Both devices tested were able to determine whether the cattle were lame; however, the research-specification device performed marginally better at this function. This minimal difference in effectiveness between these devices suggests that low-cost thermal imaging devices could be used as a lameness detection aid; increased use of these devices by farmers may increase lameness detection rates and benefit animal welfare.

Abstract

Lameness has a high economic cost to the U.K. dairy industry; accurate and early detection of lameness minimises this cost. Infrared thermal imaging (IRT) devices have shown promising results for use as a lameness detection aid in cattle when used in research settings; these devices are typically high-cost, limiting their adoption. This study analysed the effectiveness of low-cost IRT devices (LCDs) as lameness detection aids, by comparing both maximum environmentally adjusted temperature values and hindfeet temperature difference collected by an LCD to the mobility score of the cow; this test was repeated for data collected by a research-specification device. Data collection occurred during routine milking of 83 cattle; each cow’s mobility was scored afterwards. Significant differences were found between lame and sound cows with the LCD, upon analysis of both methods. There was no significant difference between the data captured by differing devices. The maximum sensitivity and specificity values for the LCD were calculated as 66.95 and 64.53, respectively, compared with 70.34 and 70.94, respectively, for the research-specification device; optimum threshold values for these were equivalent for both devices, suggesting IRT lameness identification is not device-dependent. It was concluded that a minimal difference in effectiveness between tested devices suggests that LCDs could be used as a lameness detection aid; consequently, there is potential for widespread adoption as on-farm detection aids.

A Systematic Review of Automatic Health Monitoring in Calves: Glimpsing the Future From Current Practice

Infectious diseases, particularly bovine respiratory disease (BRD) and neonatal calf diarrhea (NCD), are prevalent in calves. Efficient health-monitoring tools to identify such diseases on time are lacking. Common practice (i.e., health checks) often identifies sick calves at a late stage of disease or not at all. Sensor technology enables the automatic and continuous monitoring of calf physiology or behavior, potentially offering timely and precise detection of sick calves. A systematic overview of automated disease detection in calves is still lacking. The objectives of this literature review were hence: to investigate previously applied sensor validation methods used in the context of calf health, to identify sensors used on calves, the parameters these sensors monitor, and the statistical tools applied to identify diseases, to explore potential research gaps and to point to future research opportunities. To achieve these objectives, systematic literature searches were conducted. We defined four stages in the development of health-monitoring systems: (1) sensor technique, (2) data interpretation, (3) information integration, and (4) decision support. Fifty-four articles were included (stage one: 26; stage two: 19; stage three: 9; and stage four: 0). Common parameters that assess the performance of these systems are sensitivity, specificity, accuracy, precision, and negative predictive value. Gold standards that typically assess these parameters include manual measurement and manual health-assessment protocols. At stage one, automatic feeding stations, accelerometers, infrared thermography cameras, microphones, and 3-D cameras are accurate in screening behavior and physiology in calves. At stage two, changes in feeding behaviors, lying, activity, or body temperature corresponded to changes in health status, and point to health issues earlier than manual health checks. At stage three, accelerometers, thermometers, and automatic feeding stations have been integrated into one system that was shown to be able to successfully detect diseases in calves, including BRD and NCD. We discuss these findings, look into potentials at stage four, and touch upon the topic of resilience, whereby health-monitoring system might be used to detect low resilience (i.e., prone to disease but clinically healthy calves), promoting further improvements in calf health and welfare.

Clinical Applications and Factors Involved in Validating Thermal Windows Used in Infrared Thermography in Cattle and River Buffalo to Assess Health and Productivity

Infrared thermography (IRT) is a non-ionizing, non-invasive technique that permits evaluating the comfort levels of animals, a topic of concern due to the growing interest in determining the state of health and welfare of production animals. The operating principle of IRT is detecting the heat irradiated in anatomical regions characterized by a high density of near-surface blood vessels that can regulate temperature gain or loss from/to the environment by modifying blood flow. This is essential for understanding the various vascular thermoregulation mechanisms of different species, such as rodents and ruminants' tails. The usefulness of ocular, nasal, and vulvar thermal windows in the orbital (regio orbitalis), nasal (regio nasalis), and urogenital (regio urogenitalis) regions, respectively, has been demonstrated in cattle. However, recent evidence for the river buffalo has detected discrepancies in the data gathered from distinct thermal regions in these large ruminants, suggesting a limited sensitivity and specificity when used with this species due to various factors: the presence of hair, ambient temperature, and anatomical features, such as skin thickness and variations in blood supplies to different regions. In this review, a literature search was conducted in Scopus, Web of Science, ScienceDirect, and PubMed, using keyword combinations that included "infrared thermography", "water buffalo", "river buffalo" "thermoregulation", "microvascular changes", "lacrimal caruncle", "udder", "mastitis", and "nostril". We discuss recent findings on four thermal windows-the orbital and nasal regions, mammary gland in the udder region (regio uberis), and vulvar in the urogenital region (regio urogenitalis)-to elucidate the factors that modulate and intervene in validating thermal windows and interpreting the information they provide, as it relates to the clinical usefulness of IRT for cattle (Bos) and the river buffalo (Bubalus bubalis).

ASAS-NANP SYMPOSIUM: prospects for interactive and dynamic graphics in the era of data-rich animal science1

Statistical graphics, and data visualization, play an essential but under-utilized, role for data analysis in animal science, and also to visually illustrate the concepts, ideas, or outputs of research and in curricula. The recent rise in web technologies and ubiquitous availability of web browsers enables easier sharing of interactive and dynamic graphics. Interactivity and dynamic feedback enhance human-computer interaction and data exploration. Web applications such as decision support systems coupled with multimedia tools synergize with interactive and dynamic graphics. However, the importance of graphics for effectively communicating data, understanding data uncertainty, and the state of the field of interactive and dynamic graphics is underappreciated in animal science. To address this gap, we describe the current state of graphical methodology and technology that might be more broadly adopted. This includes an explanation of a conceptual framework for effective graphics construction. The ideas and technology are illustrated using publicly available animal datasets. We foresee that many new types of big and complex data being generated in precision livestock farming create exciting opportunities for applying interactive and dynamic graphics to improve data analysis and make data-supported decisions.

Image Analysis and Computer Vision Applications in Animal Sciences: An Overview

Computer Vision, Digital Image Processing, and Digital Image Analysis can be viewed as an amalgam of terms that very often are used to describe similar processes. Most of this confusion arises because these are interconnected fields that emerged with the development of digital image acquisition. Thus, there is a need to understand the connection between these fields, how a digital image is formed, and the differences regarding the many sensors available, each best suited for different applications. From the advent of the charge-coupled devices demarking the birth of digital imaging, the field has advanced quite fast. Sensors have evolved from grayscale to color with increasingly higher resolution and better performance. Also, many other sensors have appeared, such as infrared cameras, stereo imaging, time of flight sensors, satellite, and hyperspectral imaging. There are also images generated by other signals, such as sound (ultrasound scanners and sonars) and radiation (standard x-ray and computed tomography), which are widely used to produce medical images. In animal and veterinary sciences, these sensors have been used in many applications, mostly under experimental conditions and with just some applications yet developed on commercial farms. Such applications can range from the assessment of beef cuts composition to live animal identification, tracking, behavior monitoring, and measurement of phenotypes of interest, such as body weight, condition score, and gait. Computer vision systems (CVS) have the potential to be used in precision livestock farming and high-throughput phenotyping applications. We believe that the constant measurement of traits through CVS can reduce management costs and optimize decision-making in livestock operations, in addition to opening new possibilities in selective breeding. Applications of CSV are currently a growing research area and there are already commercial products available. However, there are still challenges that demand research for the successful development of autonomous solutions capable of delivering critical information. This review intends to present significant developments that have been made in CVS applications in animal and veterinary sciences and to highlight areas in which further research is still needed before full deployment of CVS in breeding programs and commercial farms.

Towards sensor-based calving detection in the rangelands: a systematic review of credible behavioral and physiological indicators

Calving is a critical point in both a cow and calf’s life, when both become more susceptible to disease and risk of death. Ideally, this period is carefully monitored. In extensive grazing systems, however, it is often not economically or physically possible for producers to continuously monitor animals, and thus, calving frequently goes undetected. The development of sensor systems, particularly in these environments, could provide significant benefits to the industry by increasing the quantity and quality of individual animal monitoring. In the time surrounding calving, cows undergo a series of behavioral and physiological changes, which can potentially be detected using sensing technologies. Before developing a sensor-based approach, it is worthwhile considering these behavioral and physiological changes, such that the appropriate technologies can be designed and developed. A systematic literature review was conducted to identify changes in the dam’s behavioral and physiological states in response to a calving event. Articles (n = 104) consisting of 111 independent experiments were assessed following an intensive search of electronic databases. Commonly reported indicators of parturition (n = 38) were identified, and temporal trend graphs were generated for 13 of these changes. The results compare trends in behavioral and physiological changes across a variety of animal-related factors and identifies several reliable indicators of parturition for detection with sensors, namely calf grooming behavior, changes in rumination duration, and lying bouts. This synthesis of literature suggests that variability exists between individuals and thus, combining several calving indicators may result in a more broadly applicable and accurate detection of parturition.

Bovine Respiratory Disease Diagnosis: What Progress Has Been Made in Clinical Diagnosis?

Bovine respiratory disease (BRD) complex is a worldwide health problem in cattle and is a major reason for antimicrobial use in young cattle. Several challenges may explain why it is difficult to make progress in the management of this disease. This article defines the limitation of BRD complex nomenclature, which may not easily distinguish upper versus lower respiratory tract infection and infectious bronchopneumonia versus other types of respiratory diseases. It then discusses the obstacles to clinical diagnosis and reviews the current knowledge of readily available diagnostic test to reach a diagnosis of infectious bronchopneumonia.

Comparison of thermal imaging and rectal temperature in the diagnosis of pyrexia in pre-weaned calves using on farm conditions

Measuring core body temperature is used as part of the diagnostic process in assessing the health of animals. Typically in calves, this is carried out using a rectal thermometer which can be time consuming, stressful to the calf and is invasive by nature. A non-invasive technique that is gaining recognition is thermal imaging. This study investigated the use of thermal imaging as a technique to assess core body temperature in pre-weaned artificially reared calves. A total of 125 male and female calves had rectal temperatures measured daily from day 7 until day 40 of life, and at the same time had a thermal image taken of the area around the medial canthus of the eye. A weak correlation (r = 0.28) was found between calf rectal temperature and thermal image temperature. A multivariable predictive model for core body temperature increased the correlation (r = 0.32) when including the environmental parameters of air temperature (p < .001) and wind speed (p < .001) as well as reconstituted milk replacer consumption (p < .01). The effectiveness of a predictive model including these parameters for the detection of calves with a core body temperature ≥ 39.5 °C was examined and found to have a sensitivity of 0% and a specificity of 100%. The results of this study demonstrate the need to take thermal environmental parameters into consideration when using thermal imaging to assess body temperature. However, the results suggest that accurate measures of core body temperature using thermal imaging cannot be achieved under commercial farm conditions. Further research is needed to determine what other factors could be measured to increase predictive ability.

The effect of calf jacket usage on performance, behaviour and physiological responses of group-housed dairy calves

Poor performance and ill-health of calves in the pre-wean period can affect future productivity. Increasing numbers of producers are opting to use calf jackets as a means of mitigating the potential negative effects of low ambient temperatures, wind speed and precipitation on growth and health. This study aimed to use a range of noninvasive monitoring technologies to investigate the effects of using calf jackets in the first 3 weeks of life on calf performance and behavioural and physiological parameters. Ninety Holstein-Friesian calves were allocated to one of the two treatments: (i) Jacketed until 21 days of age and (J; n = 44) ii. Nonjacketed (NJ; n = 46). Calves were group housed and fed milk replacer (MR) and concentrate solid feed via automatic feeders. Calves were weaned at day 56, and the experiment was completed at day 63. Health assessments were conducted on a daily basis throughout the experiment using predefined faecal and respiratory scoring protocols. A range of novel, noninvasive monitoring technologies were used to examine the activity, heart rate and thermal profiles of calves on an individual basis throughout the experimental period. There were no differences in calf live weight (LWT), average daily gain (ADG) or feed conversion efficiency (FCE) in J and NJ calves between days 5 to 20. However, NJ calves consumed more MR and had more unrewarded visits to the milk feeder than J calves during this period. Although calf LWT was comparable across treatments in the week following jacket removal (days 21 to 28), both ADG and FCE tended to be greater in NJ calves. There were no treatment differences in calf LWT at the end of the study (d63). When measured over a period of 24 h and at a mean ambient temperature of 7.7°C, skin surface temperature was 6.37°C higher in J calves. Core body temperature was higher in J calves between days 5 to 20; however, there were no differences in IR eye or IR rectal temperature. No differences in lying behaviour occurred, with calves spending 18 and 17 h/day lying between days 5 to 20 and days 21 to 28, respectively. Under the climatic and management conditions described, no significant benefits to calf performance were found as a result of the provision of calf jackets to group-housed calves in the first 3 weeks of life. The higher frequency of unrewarded visits to the milk feeder in NJ calves during the first 3 weeks of life could be suggestive of a lack of satiety in these calves.