Diagnostic validation of vertebral heart score machine learning algorithm for canine lateral chest radiographs

OBJECTIVES

The vertebral heart score is a measurement used to index heart size relative to thoracic vertebra. Vertebral heart score can be a useful tool for identifying and staging heart disease and providing prognostic information. The purpose of this study is to validate the use of a vertebral heart score algorithm compared to manual vertebral heart scoring by three board-certified veterinary cardiologists.

MATERIALS AND METHODS

A convolutional neural network centred around semantic segmentation of relevant anatomical features was developed to predict heart size and vertebral bodies. These predictions were used to calculate the vertebral heart score. An external validation study consisting of 1200 canine lateral radiographs was randomly selected to match the underlying distribution of vertebral heart scores. Three American College of Veterinary Internal Medicine board-certified cardiologists were enrolled to manually score 400 images each using the traditional Buchanan method. Post-scoring, the cardiologists evaluated the algorithm for misaligned anatomic landmarks and overall image quality.

RESULTS

The 95th percentile absolute difference between the cardiologist vertebral heart score and the algorithm vertebral heart score was 1.05 vertebrae (95% confidence interval: 0.97 to 1.20 vertebrae) with a mean bias of -0.09 vertebrae (95% confidence interval: -0.12 to -0.05 vertebrae). In addition, the model was observed to be well calibrated across the predictive range.

CLINICAL SIGNIFICANCE

We have found the performance of the vertebral heart score algorithm comparable to three board-certified cardiologists. While validation of this vertebral heart score algorithm has shown strong performance compared to veterinarians, further external validation in other clinical settings is warranted before use in those settings.

Asymptomatic carriage of canine infectious respiratory disease complex pathogens among healthy dogs

OBJECTIVES

The objectives of this study were to determine the prevalence of canine infectious respiratory disease pathogens among asymptomatic client-owned dogs, and to compare the risks of asymptomatic pathogen carriage between client-owned dogs and dogs in an animal shelter.

MATERIALS AND METHODS

Pooled tonsillar, conjunctival and nasal cavity swabs from asymptomatic client-owned dogs (n=133) were tested using a real-time polymerase chain reaction canine respiratory panel. Identical samples from asymptomatic dogs in an animal shelter (n=295) were similarly tested for selected pathogens. Risk differences were calculated between client-owned dogs and shelter dogs for each of the respiratory pathogens included in the analyses.

RESULTS

A total of 15 of 133 (11.3%) asymptomatic client-owned dogs were positive for at least one pathogen in the complex. Seven dogs (6.1%) were positive for M. cynos, six (5.2%) were positive for B. bronchiseptica, two (1.7%) were positive for canine herpesvirus type 1 and two (1.7%) were positive for canine respiratory coronavirus. For all eight pathogens tested in both groups, the proportion of positive cases was higher among shelter dogs than among client-owned dogs. Shelter dogs had a higher risk for M. cynos (0.18, 95% confidence interval: 0.12 to 0.25), canine respiratory coronavirus (0.15, 95% confidence interval: 0.10 to 0.19), canine distemper virus (0.06, 95% confidence interval: 0.03 to 0.09), and canine pneumovirus (0.05, 95% confidence interval: 0.03 to 0.08) than client-owned dogs. Odds ratios for M. cynos (0.31, 95% confidence interval: 0.08 to 0.92) and canine respiratory coronavirus (0.05, 95% confidence interval: 0.01 to 0.18) were significantly different between client-owned and shelter dogs. In all cases except for canine herpesvirus type 1, dogs within the shelter population were observed to be at higher risk of exhibiting asymptomatic carriage of a respiratory pathogen as compared to client-owned dogs. The strength of this association was strongest for M. cynos and canine respiratory coronavirus.

CLINICAL SIGNIFICANCE

The risk of canine infectious respiratory disease pathogen exposure posed by asymptomatic client-owned dogs is poorly defined. This study also corroborates previous reports of high canine infectious respiratory disease prevalence among clinically healthy shelter dogs, and further determined that the overall prevalence of canine infectious respiratory disease pathogen carriage among clinically healthy client-owned dogs is low but is highest for the traditional pathogen B. bronchiseptica and the emerging pathogen M. cynos.

Diagnostic accuracy of computer aided electrocardiogram analysis in dogs

OBJECTIVES

Evaluation of a computerised electrocardiogram algorithm compared to the interpretation of a team of board-certified veterinary cardiologists.

MATERIALS AND METHODS

This was a cross-sectional retrospective cohort study. A total of 399 electronic canine electrocardiogram recordings screened from 1391 electrocardiograms were enrolled in the study. A panel of seven cardiologists, masked to patient information, evaluated electrocardiograms for the following: P-wave amplitude and duration; PR-interval; R-wave amplitude; QRS duration; heart rate; mean electrical axis; and final overall diagnosis for the detection of arrhythmia and any abnormal electrocardiogram anomaly.

RESULTS

The sensitivity of the electrocardiogram algorithm for detecting arrhythmias was 99.7% (95% confidence intervals, CI: 98.5 to 99.9) and the specificity was 99.5% (95% CI: 98.0 to 99.9) compared to the consensus result created by panel of cardiologists. The sensitivity of the algorithm for the detection of any electrocardiogram anomaly, including abnormal measurements, was 71.3% (95% CI: 65.5 to 76.7) and the specificity was 35.1% (95% CI: 27.0 to 43.8) compared to the panel of cardiologists.

CLINICAL SIGNIFICANCE

The electrocardiogram algorithm was shown to have high sensitivity for the detection of arrhythmias, but not all electrocardiogram anomalies. The results support the use of this algorithm as a tool to aid in the triage of the electrocardiogram workflow.

Urinary clusterin and cystatin B as biomarkers of tubular injury in dogs following envenomation by the European adder

Diagnosing acute kidney injury remains a challenge since the established renal biomarkers, serum creatinine (sCr) and symmetric dimethylarginine (SDMA) reflect glomerular function and not tubular injury. Sensitive tubular markers such as urinary clusterin (uClust) and cystatin B (uCysB) have been proposed to detect AKI at an earlier stage. Since envenomation by the European adder (Vipera berus berus) could serve as a spontaneous disease model of AKI we investigated these new biomarkers in affected dogs. Concentrations of uClust and uCysB as well as sCr and SDMA were analyzed retrospectively in stored samples from 26 dogs with snake envenomation and 13 healthy controls. Higher concentrations of uClust (P < 0.012) and uCysB (P < 0.001) were observed in the snake-envenomed group. Normalization of uClust and uCysB to urinary creatinine did not alter the results. No differences were observed in sCr and SDMA between the snake-envenomed group and the healthy control group. Spearman rank correlation analysis revealed a strong association of uClust with uCysB in the snake-envenomed dogs (r = 0.75 P < 0.001) but not in the healthy controls. The high percentage of snake-envenomed dogs with increased uClust and uCysB concentrations in the absence of increased sCr and SDMA suggests renal tubular injury in the affected dogs. Larger prospective case-controlled studies are warranted to evaluate the clinical utility and prognostic value of these biomarkers.