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.

Computer vision model for the detection of canine pododermatitis and neoplasia of the paw.

BACKGROUND: Artificial intelligence (AI) has been used successfully in human dermatology. AI utilises convolutional neural networks (CNN) to accomplish tasks such as image classification, object detection and segmentation, facilitating early diagnosis. Computer vision (CV), a field of AI, has shown great results in detecting signs of human skin diseases. Canine paw skin diseases are a common problem in general veterinary practice, and computer vision tools could facilitate the detection and monitoring of disease processes. Currently, no such tool is available in veterinary dermatology. ANIMALS: Digital images of paws from healthy dogs and paws with pododermatitis or neoplasia were used. OBJECTIVES: We tested the novel object detection model Pawgnosis, a Tiny YOLOv4 image analysis model deployed on a microcomputer with a camera for the rapid detection of canine pododermatitis and neoplasia. MATERIALS AND METHODS: The prediction performance metrics used to evaluate the models included mean average precision (mAP), precision, recall, average precision (AP) for accuracy and frames per second (FPS) for speed. RESULTS: A large dataset labelled by a single individual (Dataset A) used to train a Tiny YOLOv4 model provided the best results with a mean mAP of 0.95, precision of 0.86, recall of 0.93 and 20 FPS. CONCLUSIONS AND CLINICAL RELEVANCE: This novel object detection model has the potential for application in the field of veterinary dermatology.

Comparison of ozonated water and acidified copper sulphate in prevention of digital dermatitis in dairy cows.

Digital dermatitis (DD) is the most significant infectious hoof disorder of cattle in Europe. Hoof baths are one of the most common control methods. Copper sulphate and formalin are commonly used in hoof baths, but their use is problematic in many European countries for health, environmental and safety reasons. Ozonated water and acidified copper sulphate were tested as prevention of DD in a 5-month study. Data were derived from 302 hind feet of Holstein and Estonian Red cows (no. of cows = 151) from a commercial dairy farm in Estonia. Altogether 168 hind feet were included in the acidified copper sulphate group and 134 feet in the ozonated water group. Hoof bathing was carried out three days a week (Mon, Wed, Fri) for two months and then two days a week (Mon, Wed) for three and a half months, in both groups. Ozonated water was sprayed on to the digital skin of hind feet of cows twice a day on treatment days, while the cows were eating. The copper sulphate bath consisted of copper sulphate (2%) mixed with an organic acid compound to acidify and ionize the solution. Cows walked through acidified copper sulphate solution twice a day on treatment days as they were exiting the milking parlor. DD negative and DD positive test results in both groups were compared and statistically tested for differences. The copper sulphate solution was more effective than ozonated water at preventing acute DD lesions. A random maximum likelihood model demonstrated that the odds ratio for DD in the ozonated water group was six times higher compared with DD in the acidified copper sulphate group. Most of the cows that were initially without any DD lesions (M0 + no other severe hoof lesion), remained lesion-free in both groups (copper sulphate group 97% and ozonated water group 88%). Despite trial design deficiencies, the findings indicate that acidified copper sulphate was a more effective solution in preventing DD than ozonated water.

Hot topic: Detecting digital dermatitis with computer vision.

Digital dermatitis (DD) is linked to severe lameness, infertility, and decreased milk production in cattle. Early detection of DD provides an improved prognosis for treatment and recovery; however, this is extremely challenging on commercial dairy farms. Computer vision (COMV) models can help facilitate early DD detection on commercial dairy farms. The aim of this study was to develop and implement a novel COMV tool to identify DD lesions on a commercial dairy farm. Using a database of more than 3,500 DD lesion images, a model was trained using the YOLOv2 architecture to detect the M-stages of DD. The YOLOv2 COMV model detected DD with an accuracy of 71%, and the agreement was quantified as "moderate" by Cohen's kappa when compared with a human evaluator for the internal validation. In the external validation, the YOLOv2 COMV model detected DD with an accuracy of 88% and agreement was quantified as "fair" by Cohen's kappa. Implementation of COMV tools for DD detection provides an opportunity to identify cows for DD treatment, which has the potential to lower DD prevalence and improve animal welfare on commercial dairy farms.

Development of real-time PCR and loop-mediated isothermal amplification (LAMP) assays for the differential detection of digital dermatitis associated treponemes.

Bovine digital dermatitis (DD) is a severe infectious cause of lameness in cattle worldwide, with important economic and welfare consequences. There are three treponeme phylogroups (T. pedis, T. phagedenis, and T. medium) that are implicated in playing an important causative role in DD. This study was conducted to develop real-time PCR and loop-mediated isothermal amplification (LAMP) assays for the detection and differentiation of the three treponeme phylogroups associated with DD. The real-time PCR treponeme phylogroup assays targeted the 16S-23S rDNA intergenic space (ITS) for T. pedis and T. phagedenis, and the flagellin gene (flaB2) for T. medium. The 3 treponeme phylogroup LAMP assays targeted the flagellin gene (flaB2) and the 16S rRNA was targeted for the Treponeme ssp. LAMP assay. The real-time PCR and LAMP assays correctly detected the target sequence of all control strains examined, and no cross-reactions were observed, representing 100% specificity. The limit of detection for each of the three treponeme phylogroup real-time PCR and LAMP assays was ≤ 70 fg/µl. The detection limit for the Treponema spp. LAMP assay ranged from 7-690 fg/µl depending on phylogroup. Treponemes were isolated from 40 DD lesion biopsies using an immunomagnetic separation culture method. The treponeme isolation samples were then subjected to the real-time PCR and LAMP assays for analysis. The treponeme phylogroup real-time PCR and LAMP assay results had 100% agreement, matching on all isolation samples. These results indicate that the developed assays are a sensitive and specific test for the detection and differentiation of the three main treponeme phylogroups implicated in DD.

The DD Check App for prevention and control of digital dermatitis in dairy herds.

Digital dermatitis (DD) is the most important infectious claw disease in the cattle industry causing outbreaks of lameness. The clinical course of disease can be classified using 5 clinical stages. M-stages represent not only different disease severities but also unique clinical characteristics and outcomes. Monitoring the proportions of cows per M-stage is needed to better understand and address DD and factors influencing risks of DD in a herd. Changes in the proportion of cows per M-stage over time or between groups may be attributed to differences in management, environment, or treatment and can have impact on the future claw health of the herd. Yet trends in claw health regarding DD are not intuitively noticed without statistical analysis of detailed records. Our specific aim was to develop a mobile application (app) for persons with less statistical training, experience or supporting programs that would standardize M-stage records, automate data analysis including trends of M-stages over time, the calculation of predictions and assignments of Cow Types (i.e., Cow Types I-III are assigned to cows without active lesions, single and repeated cases of active DD lesions, respectively). The predictions were the stationary distributions of transitions between DD states (i.e., M-stages or signs of chronicity) in a class-structured multi-state Markov chain population model commonly used to model endemic diseases. We hypothesized that the app can be used at different levels of record detail to discover significant trends in the prevalence of M-stages that help to make informed decisions to prevent and control DD on-farm. Four data sets were used to test the flexibility and value of the DD Check App. The app allows easy recording of M-stages in different environments and is flexible in terms of the users' goals and the level of detail used. Results show that this tool discovers trends in M-stage proportions, predicts potential outbreaks of DD, and makes comparisons among Cow Types, signs of chronicity, scorers or pens. The DD Check App also provides a list of cows that should be treated augmented by individual Cow Types to help guide treatment and determine prognoses. Producers can be proactive instead of reactive in controlling DD in a herd by using this app. The DD Check App serves as an example of how technology makes knowledge and advice of veterinary epidemiology widely available to monitor, control and prevent this complex disease.

Short communication: Determination of the ability of Thymox to kill or inhibit various species of microorganisms associated with infectious causes of bovine lameness in vitro.

Infectious claw diseases continue to plague cattle in intensively managed husbandry systems. Poor foot hygiene and constant moist environments lead to the infection and spread of diseases such as digital dermatitis (hairy heel warts), interdigital dermatitis, and interdigital phlegmon (foot rot). Currently, copper sulfate and formalin are the most widely used disinfecting agents in bovine footbaths; however, the industry could benefit from more environmentally and worker friendly substitutes. This study determined the in vitro minimum inhibitory concentrations and minimum bactericidal concentrations of Thymox (Laboratoire M2, Sherbrooke, Québec, Canada) for a selection of microorganisms related to infectious bovine foot diseases. Thymox is a broad-spectrum agricultural disinfectant that is nontoxic, noncorrosive, and readily biodegradable. The values for minimum inhibitory concentration and minimum bactericidal concentration indicated that Thymox inhibited growth and killed the various species of microorganisms under study at much lower concentrations compared with the recommended working concentration of a 1% solution. Overall, the values found in this study of minimum inhibitory concentration and minimum bactericidal concentration of Thymox show its potential as an alternative antibacterial agent used in bovine footbaths; however, field trials are needed to determine its effectiveness for the control and prevention of infectious claw diseases.

Data-driven methods for imputing national-level incidence in global burden of disease studies.

OBJECTIVE: To develop transparent and reproducible methods for imputing missing data on disease incidence at national-level for the year 2005. METHODS: We compared several models for imputing missing country-level incidence rates for two foodborne diseases - congenital toxoplasmosis and aflatoxin-related hepatocellular carcinoma. Missing values were assumed to be missing at random. Predictor variables were selected using least absolute shrinkage and selection operator regression. We compared the predictive performance of naive extrapolation approaches and Bayesian random and mixed-effects regression models. Leave-one-out cross-validation was used to evaluate model accuracy. FINDINGS: The predictive accuracy of the Bayesian mixed-effects models was significantly better than that of the naive extrapolation method for one of the two disease models. However, Bayesian mixed-effects models produced wider prediction intervals for both data sets. CONCLUSION: Several approaches are available for imputing missing data at national level. Strengths of a hierarchical regression approach for this type of task are the ability to derive estimates from other similar countries, transparency, computational efficiency and ease of interpretation. The inclusion of informative covariates may improve model performance, but results should be appraised carefully.

Altered microbiomes in bovine digital dermatitis lesions, and the gut as a pathogen reservoir.

Bovine digital dermatitis (DD) is the most important infectious disease associated with lameness in cattle worldwide. Since the disease was first described in 1974, a series of Treponema species concurrent with other microbes have been identified in DD lesions, suggesting a polymicrobial etiology. However, the pathogenesis of DD and the source of the causative microbes remain unclear. Here we characterized the microbiomes of healthy skin and skin lesions in dairy cows affected with different stages of DD and investigated the gut microbiome as a potential reservoir for microbes associated with this disease. Discriminant analysis revealed that the microbiomes of healthy skin, active DD lesions (ulcerative and chronic ulcerative) and inactive DD lesions (healing and chronic proliferative) are completely distinct. Treponema denticola, Treponema maltophilum, Treponema medium, Treponema putidum, Treponema phagedenis and Treponema paraluiscuniculi were all found to be present in greater relative abundance in active DD lesions when compared with healthy skin and inactive DD lesions, and these same Treponema species were nearly ubiquitously present in rumen and fecal microbiomes. The relative abundance of Candidatus Amoebophilus asiaticus, a bacterium not previously reported in DD lesions, was increased in both active and inactive lesions when compared with healthy skin. In conclusion, our data support the concept that DD is a polymicrobial disease, with active DD lesions having a markedly distinct microbiome dominated by T. denticola, T. maltophilum, T. medium, T. putidum, T. phagedenis and T. paraluiscuniculi. Furthermore, these Treponema species are nearly ubiquitously found in rumen and fecal microbiomes, suggesting that the gut is an important reservoir of microbes involved in DD pathogenesis. Additionally, the bacterium Candidatus Amoebophilus asiaticus was highly abundant in active and inactive DD lesions.

The dynamics of digital dermatitis in populations of dairy cattle: model-based estimates of transition rates and implications for control.

Five groups of dairy cows affected by digital dermatitis were subjected to five different footbath strategies and evaluated at regular 3-weekly intervals. A standard protocol was used to record five different stages of disease from early (M1), acute ulcerative (M2), healing (M3) and chronic lesions (M4) in addition to the negative stage of disease (M0). The effect of the footbathing was evaluated using mathematical modelling for the transmission dynamics of infections and summarized using the reproduction ratio R(0). Sensitivity analysis for a range of parameters in the mathematical model showed that the speed of detecting acute lesions and the efficiency with which those lesions were treated were the key parameters which determined whether lesions became more severe or whether they healed.

Long-term observations on the dynamics of bovine digital dermatitis lesions on a California dairy after topical treatment with lincomycin HCl.

The objective of this study was to observe the dynamics of clinical cure and recurrence of the lesions of bovine digital dermatitis for 11 months after treatment with topical lincomycin HCl. The study was a clinical follow-up of 39 active bovine digital dermatitis lesions (from 29 cows). Cows with active, painful bovine digital dermatitis (BDD) lesions on the interdigital commissure of the rear feet were identified on day 0. On day 1, lesions in all cows were photographed and full-skin thickness 6mm punch biopsies were obtained for histological evaluation. All lesions on all cows were treated with topical lincomycin paste under a light bandage. On days 12 and 23, a subsample of 10 lesions was randomly selected, photographed, and biopsied. On day 37, all lesions on all cows were photographed and biopsied. After day 37, lesions were evaluated on a monthly basis. All lesions were photographed at each observation until day 341 (end of study) but only cows that had macroscopically active lesions were biopsied. Of the 39 lesions treated on day 1, 21 (54%) required re-treatment on at least one occasion before day 341. Macroscopic classification agreed well with histological classification when lesions were small, focal and active (M1 lesions) or large, ulcerative and active (M2), but agreement was variable for lesions that had healed macroscopically (M5) or that were chronic (M4). A transition model showed that M1 and M2 lesions were 27 times more likely to be an M2 lesion on the next observation than to be a healed (M5) lesion.

Modelling the dynamics of intramammary E. coli infections in dairy cows: understanding mechanisms that distinguish transient from persistent infections.

The majority of intramammary infections with Escherichia coli in dairy cows result in transient infections with duration of about 10 days or less, although more persistent infections (2 months or longer) have been identified. We apply a mathematical model to explore the role of an intracellular mammary epithelial cell reservoir in the dynamics of infection. We included biological knowledge of the bovine immune response and known characteristics of the bacterial population in both transient and persistent infections. The results indicate that varying the survival duration of the intracellular reservoir reproduces the data for both transient and persistent infections. Survival in an intracellular reservoir is the most likely mechanism that ensures persistence of E. coli infections in mammary glands. Knowledge of the pathogenesis of persistent infections is essential to develop preventive and treatment programmes for these important infections in dairy cows.