Thoracic Ultrasonography in Calves: A Narrative Review of Techniques and Reporting Practices

Abstract

Ultrasonography of the bovine lung is a noninvasive technique allowing recognition of lower respiratory tract lesions and differentiation from disease limited to the upper respiratory tract. Techniques for scanning the thorax have evolved to facilitate examination of cohorts of calves quickly, while maintaining accuracy. Classification systems for the interpretation of images, their assignment as normal or abnormal, and grading of their severity are varied. Without a reporting consensus, comparison of short-and long-term outcomes attributable to ultrasonographic findings is challenging. Differences in operator agreement might complicate interpretation further. The objective of this review was to gather methods for screening calf lungs using thoracic ultrasonography and describe the heterogeneity in scanning techniques and methods of image interpretation, including available scoring methods.

Keywords: cattle; pneumonia; thoracic ultrasonography; youngstock.

FIGURE 1

A PRISMA flowchart describing the literature search and study collection process.

FIGURE 2

(a) Image of a comet tail artifact, with shadowing arising from the pleural surface; (b) image of a superficial alveologram. A focal hypoechoic region with acoustic shadowing radiating medially from the unaerated region is visible with disruption of the normal reverberation artifacts (A lines).

FIGURE 3

Illustration of the different assumptions of scoring systems using a threshold for determining negative outcomes. A linear relationship between maximal depth of consolidation (probability of negative outcome increase linearly for each additional centimeter of maximal depth) is presented (blue dotted line). For most dichotomous outcomes (survival, relapse (Y/N)), the increased risk is generally represented as a logistic curve (black line) since assuming a linear relation of the predictor of interest (here consolidation depth) and the log‐odds of the probability of the negative outcome. Therefore, the increased probability is not linearly associated with the depth of consolidation. When using a “threshold approach” (here deciding that 3 cm or more of consolidation are associated with negative outcomes), we assume that the probability outcome of is the “mean” risk of the strata with the consolidation range. In this particular case, we can obviously see, for example, that a calf with 8 cm depth of consolidation would have a predicted probability of negative event at 0.95, whereas it would be estimated at only 0.5 when using the dichotomous approach. On the other hand, a calf with 3 cm consolidation depth will still be estimated at a 0.5 probability of negative outcome with the dichotomous approach versus 0.12 using the continuous logistic curve approach.