Introduction
Point-of-care LUS is an emerging imaging modality found to have diagnostic utility in human and veterinary patients with respiratory clinical signs. The presence of alveolar or interstitial pulmonary abnormalities can be suggested by the presence of ultra-sound artifacts known as B lines, which are created by the high impendence gradient between interstitial or alveolar fluid and adjacent pulmonary air.1,2,3 These discrete, narrow-based, vertical hyperechoic artifacts extend from the pulmonary-pleural interface in the near field to the far aspect of the ultrasound screen without fading, and they move synchronously with respiration. Identification of B lines in ultrasonographic images of the lungs allows differentiation between L-CHF and noncardiac causes of dyspnea or cough with high sensitivity and specificity in people, dogs, and cats.2,4,5,6,7,8,9,10 In this context, ultrasonographic findings correspond well with results for other diagnostic tests such as N-terminal pro-B-type natriuretic peptide3,5,9,11,12 or thoracic radiography,4,13,14,15,16 although false-positive results can occur with primary respiratory diseases that cause diffuse interstitial or alveolar infiltrates.
As a point-of-care test, LUS is designed to be performed in an emergency setting with a portable ultrasound machine, but the ideal transducer type for LUS has not been established in either human or veterinary medicine. Whereas most human LUS studies use curvilinear-array (convex or microconvex) transducers,13,17,18 other investigators report similar results with phased-array transducers.9,19,20 In veterinary medicine, most published LUS studies have been performed with curvilinear-array (microconvex) transducers,4,5,6,21,22 although linear-array23,24 or phased-array25 transducers have also been used.
Veterinary textbooks typically recommend curvilinear-array (microconvex) transducers for thoracic ultrasonography26,27 to ensure that the transducer footprint fits within the patient's intercostal spaces; these transducers are also advocated as all-purpose transducers for point-of-care ultrasonography28,29,30 and are generally the only transducer available when a portable ultrasound machine is equipped with only 1 transducer. In contrast, phased-array transducers are preferred for cardiac ultrasonography owing to their superior temporal resolution26,28 and are the most common transducer type available to cardiologists when performing echocardiography. Because diagnostic LUS is practiced routinely by a variety of general and specialty practitioners with access to different transducer types, operators would benefit from knowing about differences in diagnostic value and image quality between transducers.
The purpose of the study reported here was to compare the use of curvilinear-array (microconvex) and phased-array transducers for performing LUS in dogs with acute and resolving L-CHF. Specific aims were to compare the number of B lines detected, number of sites strongly positive for B lines (ie, those with > 3 B lines) detected, and subjectively assessed image quality between the 2 transducer types, as well as interobserver agreement for each of these variables. We hypothesized that both transducers would allow detection of B lines for diagnostic purposes and that the curvilinear-array transducer would produce higher-quality images with lower interobserver variability.
Materials and Methods
Dogs
The present study was a performed in the context of a larger prospective investigation31 of the diagnostic utility of LUS for monitoring resolution of pulmonary edema in dogs with L-CHF. Procedures were approved by the Institutional Animal Care and Use Committee of Iowa State University (protocol 10-15-8106-K). Informed owner consent was obtained for each dog enrolled in the study.
Client-owned dogs presented to the Small Animal Emergency Service or Cardiology Service at Iowa State University Lloyd Veterinary Medical Center were prospectively enrolled after diagnosis of first-onset L-CHF. Congestive heart failure diagnosis was made on the basis of a combination of clinical and diagnostic findings, including results of echocardiography performed by a board-certified veterinary cardiologist or supervised cardiology resident-in-training (showing myxomatous mitral valve disease or dilated cardiomyopathy with severe left atrial enlargement), thoracic radiography (revealing an interstitial or alveolar pulmonary pattern [or both] consistent with pulmonary edema), and a positive response to furosemide-induced diuresis. Dogs that had more than a trace amount of pleural effusion that could cause pressure atelectasis and create B lines independent of pathological pulmonary changes, had received furosemide prior to evaluation at our facility, or had mild L-CHF for which hospitalization and inpatient management for ≥ 6 hours were not clinically required were excluded.
Procedures
Dogs received emergency inpatient and subsequent outpatient treatment for L-CHF according to standard medical practices and clinical judgment of the attending clinician. The LUS was performed at the following clinically relevant time points: ≤ 6 hours after presentation at the time of initial L-CHF diagnosis (time point 1), ≤ 2 hours before hospital discharge (time point 2), and at first outpatient recheck examination approximately 7 to 14 days after hospital discharge (time point 3). Although the overarching research project followed study participants through future recheck examinations and assessed clinical outcomes, long-term follow-up was not part of the methods comparison study reported here.
All LUS procedures were performed by board-certified veterinary cardiologists or supervised cardiology residents (the same cardiologist or resident who performed echocardiography) using a compact portable ultrasound machine.a Examinations consisted of B-mode images performed with 2 electronic sector-array transducers (1 curvilinear array [micro-convex]b and 1 phased arrayc) and standardized settings (ultrasound frequency, 4 to 8 MHz; depth, 4 to 8 cm). Because the manufacturer's presets for lung imaging were not available for either transducer, image quality was optimized for each examination by adjusting the focal zone, overall gain, and near-field gain and by engaging harmonics.
A previously described protocol for LUS was followed.4,5,10,21,31,32 Briefly, dogs underwent imaging in a standing or sternally recumbent position. The fur was parted, and isopropyl alcohol was used to facilitate transducer contact, but fur was not shaved and no coupling gel was used. Images were obtained from 4 standardized acoustic windows over each hemithorax for a total of 8 LUS sites (caudal, perihilar, middle, and cranial sites on the left side, followed by the same sites in the same order on the right side). A complete ultrasonographic examination with all sites imaged in the indicated order was performed first with the phased-array transducer and then with the curvilinear-array (microconvex) transducer. Three-second cineloops were captured and archived from a single intercostal space at each site with each transducer.
Post hoc analysis of cineloops recorded during LUS (with both transducer types) was performed by 2 independent observers (a board-certified veterinary cardiologist (JLW; observer 1) and a cardiology resident (SDM; observer 2). All images of all dogs were evaluated in a single session, with the observer blinded to the dog's identity and the time point; blinding to the transducer type used was not possible for technical reasons. Each site was evaluated for the presence and number of B lines per intercostal space, recorded as 0, 1, 2, 3, > 3, or infinite (coalescing B lines that could not be individually enumerated).4,21 Sites with > 3 or infinite B lines were further designated as strongly positive for B lines (termed strong-positive sites) on the basis of recommendations for point-of-care LUS in people.2 A total B-line score for each examined dog per time point was determined by summing the number of B lines from the 8 individual sites, with > 3 B lines/intercostal space recorded as 4 and infinite B lines recorded as 10, respectively. This scoring system, with a maximum possible total B-line score of 80, was used as described in previous studies5,8 of human and veterinary patients. The total number of strong-positive sites was also summed, with a maximum possible number of 8 sites/dog/time point.
Observers also assessed the quality of each ultrasonographic image according to a scoring system with the following criteria: pulmonary-pleural interface is visible and identifiable (yes = 1; no = 0), both rib shadows are visible and identifiable (yes = 1; no = 0), A lines are visible and identifiable or B lines are clearly present and obscuring A lines (yes = 1; unclear = 0), and B lines are clearly present and identifiable or clearly absent (yes = 1; unclear = 0). Quality scores thus ranged from 0 to 4/site (Figure 1). The total quality scores for images obtained with each transducer (sum for the 8 sites) ranged from 0 to 32/dog/time point.
Examples of 2-D still images obtained with curvilinear-array (microconvex; A) and phased-array (B) transducers during LUS of a dog with L-CHF in a study to compare use of the 2 transducer types for this purpose. For all dogs, images were obtained according to a previously described protocol4,5,10,21,31,32; 4 standardized acoustic windows over each hemithorax were used for a total of 8 LUS sites (caudal, perihilar, middle, and cranial sites on the left side, followed by the same sites in the same order on the right side). Images for the right caudal site are shown. A—In the curvilinear-array transducer image, the pulmonary-pleural interface (arrow) and rib shadows (asterisks) are easily visible, and B lines (number classified as infinite; bracket) are clearly present and obscuring A lines (quality score, 4/4). B—In the phased-array transducer image, neither the pulmonary-pleural interface nor rib shadows can be clearly delineated, and differentiation between B lines and A lines is not clear (quality score, 0/4). RCd = Right caudal site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained with curvilinear-array (microconvex; A) and phased-array (B) transducers during LUS of a dog with L-CHF in a study to compare use of the 2 transducer types for this purpose. For all dogs, images were obtained according to a previously described protocol4,5,10,21,31,32; 4 standardized acoustic windows over each hemithorax were used for a total of 8 LUS sites (caudal, perihilar, middle, and cranial sites on the left side, followed by the same sites in the same order on the right side). Images for the right caudal site are shown. A—In the curvilinear-array transducer image, the pulmonary-pleural interface (arrow) and rib shadows (asterisks) are easily visible, and B lines (number classified as infinite; bracket) are clearly present and obscuring A lines (quality score, 4/4). B—In the phased-array transducer image, neither the pulmonary-pleural interface nor rib shadows can be clearly delineated, and differentiation between B lines and A lines is not clear (quality score, 0/4). RCd = Right caudal site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained with curvilinear-array (microconvex; A) and phased-array (B) transducers during LUS of a dog with L-CHF in a study to compare use of the 2 transducer types for this purpose. For all dogs, images were obtained according to a previously described protocol4,5,10,21,31,32; 4 standardized acoustic windows over each hemithorax were used for a total of 8 LUS sites (caudal, perihilar, middle, and cranial sites on the left side, followed by the same sites in the same order on the right side). Images for the right caudal site are shown. A—In the curvilinear-array transducer image, the pulmonary-pleural interface (arrow) and rib shadows (asterisks) are easily visible, and B lines (number classified as infinite; bracket) are clearly present and obscuring A lines (quality score, 4/4). B—In the phased-array transducer image, neither the pulmonary-pleural interface nor rib shadows can be clearly delineated, and differentiation between B lines and A lines is not clear (quality score, 0/4). RCd = Right caudal site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained with curvilinear-array (microconvex; A) and phased-array (B) transducers during LUS of a dog with L-CHF in a study to compare use of the 2 transducer types for this purpose. For all dogs, images were obtained according to a previously described protocol4,5,10,21,31,32; 4 standardized acoustic windows over each hemithorax were used for a total of 8 LUS sites (caudal, perihilar, middle, and cranial sites on the left side, followed by the same sites in the same order on the right side). Images for the right caudal site are shown. A—In the curvilinear-array transducer image, the pulmonary-pleural interface (arrow) and rib shadows (asterisks) are easily visible, and B lines (number classified as infinite; bracket) are clearly present and obscuring A lines (quality score, 4/4). B—In the phased-array transducer image, neither the pulmonary-pleural interface nor rib shadows can be clearly delineated, and differentiation between B lines and A lines is not clear (quality score, 0/4). RCd = Right caudal site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained with curvilinear-array (microconvex; A) and phased-array (B) transducers during LUS of a dog with L-CHF in a study to compare use of the 2 transducer types for this purpose. For all dogs, images were obtained according to a previously described protocol4,5,10,21,31,32; 4 standardized acoustic windows over each hemithorax were used for a total of 8 LUS sites (caudal, perihilar, middle, and cranial sites on the left side, followed by the same sites in the same order on the right side). Images for the right caudal site are shown. A—In the curvilinear-array transducer image, the pulmonary-pleural interface (arrow) and rib shadows (asterisks) are easily visible, and B lines (number classified as infinite; bracket) are clearly present and obscuring A lines (quality score, 4/4). B—In the phased-array transducer image, neither the pulmonary-pleural interface nor rib shadows can be clearly delineated, and differentiation between B lines and A lines is not clear (quality score, 0/4). RCd = Right caudal site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Statistical analysis
Statistical analyses were selected and performed by 1 investigator (AKV) using 2 commercially available software packages.d,e Interobserver agreement for LUS results (B-line scores and number of strong-positive sites) and image-quality scores were compared between observers with quadratic-weighted analysis; nonweighted κ analysis was used for binary variables. Agreement was classified as poor (< 0.20), fair (κ = 0.21 to 0.40), moderate (κ = 0.41 to 0.60), good (κ = 0.61 to 0.80), and excellent (κ = 0.81 to 1.00).33 Agreement and analytic bias for LUS results and quality scores were assessed between transducers (curvilinear array [microconvex] vs phased array) with Passing-Bablok regression analysis; results for all dogs at time points 1 and 2 were independently evaluated. Regression analysis was not performed for time point 3 results because a very small number of dogs underwent evaluation at this time point. For Passing-Bablok analysis, constant bias was considered present if the 95% CI of the y-intercept for the best-fit line did not include 0, and proportional bias was considered present if the 95% CI of the slope of the best-fit line did not include 1. To compare the frequency of strong-positive designations at individual anatomic sites between transducers, Fisher exact tests were performed separately for time points 1 and 2. Quality scores were compared between transducers by use of a 3-variable repeated-measures mixed-effects model, with the transducer type, anatomic site, and time point as model factors; time point 3 results were not included in this analysis because of the small number of dogs examined. Values of P < 0.05 were considered significant.
Results
Dogs and descriptive data
Thirteen dogs that underwent a total of 24 ultra-sonographic examinations during the first 6 months of study enrollment for the larger investigation31 were enrolled in the present study. Images obtained during LUS were available for all 13 dogs at time point 1 (day of initial L-CHF diagnosis), 9 dogs at time point 2 (day of hospital discharge), and 2 dogs at time point 3 (first recheck examination, approx 7 to 14 days after discharge). The study sample included 2 Chihuahuas, 2 mixed-breed dogs, and 1 each of the following breeds: Toy Poodle, Dachshund, Lhasa Apso, Cocker Spaniel, Weimaraner, Golden Retriever, Coonhound, Labrador Retriever, and Great Dane. Nine dogs were male (8 castrated and 1 sexually intact) and 4 were spayed females. The median age was 9.8 years (range, 3 to 12.9 years). Eight dogs had degenerative mitral valve disease diagnosed as the cause of L-CHF, and 5 dogs had a diagnosis of dilated cardiomyopathy.
The median total B-line score for all dogs, regardless of the transducer used, was 36 (range, 8 to 68; maximum possible score, 80) at time point 1 and 6 (range, 2 to 32) at time point 2; the absolute value was 0 for both dogs evaluated at time point 3. The median total number of strong-positive sites was 5 (range, 1 to 8; maximum possible score, 8) at time point 1 and 0 (range, 0 to 5) at time point 2; the absolute value was 0 for both dogs at time point 3.
Interobserver agreement for B-line data
Agreement between observers 1 and 2 for the total B-line scores and total number of strong-positive sites was excellent for both transducers. For the curvilinear (microconvex) transducer, the weighted κ values were 0.990 (95% CI, 0.981 to 0.998) for the total B-line score and 0.995 (95% CI, 0.987 to 1.0) for the total number of strong-positive sites. Results were similar for individual LUS sites. For total B-line scores, site-specific κ values ranged from 0.861 (95% CI, 0.678 to 1.0) for the right perihilar site to 0.989 (95% CI, 0.974 to 1.0) for the left perihilar site. For the frequency of strong-positive designations per LUS site, κ values (nonweighted because of the binary nature of this variable) ranged from 0.895 (95% CI, 0.694 to 1.0) for the left caudal and right perihilar sites to 1.0 (95% CI, 1.0 to 1.0) for all other sites.
For the phased-array transducer, the weighted κ values for the total B-line score (0.977 [95% CI, 0.961 to 0.993]) and for the total number of strong-positive sites (0.979 [95% CI, 0.963 to 0.995]) both suggested excellent agreement. Interobserver agreement was similar for individual LUS sites; for B-line scores, site-specific values ranged from 0.806 (95% CI, 0.608 to 1.0) for the right middle site to 0.983 (95% CI, 0.950 to 1.0) for the left perihilar site. For the frequency of strong-positive designations per site, the site-specific nonweighted κ values ranged from 0.747 (95% CI, 0.416 to 1.0) for the right caudal site to 1.0 (95% CI, 1.0 to 1.0) for the left cranial site. Although interobserver agreement for the total B-line scores and total number of strong-positive sites determined with the phased-array transducer was (numerically) slightly lower than those for the curvilinear-array (microconvex) transducer, these differences were deemed nonsignificant on the basis of overlap in 95% CIs for κ values.
Comparison of B-line data between transducers
Because interobserver agreement for the total B-line scores and total number of strong-positive sites was excellent, only results from observer 1 were used to compare results of LUS between transducers. For total B-lines scores determined with each transducer, Passing-Bablok regression revealed no evidence of constant or proportional bias at either time point 1 (y-intercept, 11.2 [95% CI, –2.0 to 15.3]; slope, 0.840 [95% CI, 0.684 to 1.500]; Figure 2) or time point 2 (y-intercept, 3.0 [95% CI, –4.0 to 6.0]; slope, 0.500 [95% CI, 0.500 to 2.000]). Comparison of the total number of strong-positive sites detected with each transducer similarly revealed no evidence of constant or proportional bias for measurements at either time point 1 (y-intercept, 1.1 [95% CI, –2.0 to 2.5]; slope, 0.857 [95% CI, 0.500 to 1.500]) or time point 2 (y-intercept, 0.6 [95% CI, –0.2 to 1.5]; slope, 0.681 [95% CI, 0.207 to 1.155]). For individual LUS sites, the Fisher exact test revealed no difference in the likelihood of identifying a site as strongly positive between transducer types at any site for time point 1 or 2 (P > 0.58 for all sites). Examples of ultrasonographic images obtained with each transducer showing various numbers of B lines are provided (Figure 3).
Results of Passing-Bablok analysis to compare curvilinear-array (microconvex) and phased-array transducer results for total B-line scores (A; range of possible scores, 0 to 80) and the number of strong-positive sites (B; range of possible scores, 0 to 8) during LUS of dogs with L-CHF at 1 of 3 clinical time points (time point 1 [time of L-CHF diagnosis]; n = 13). Strong-positive sites were those with > 3 B lines. Overlapping values result in 12 visible data points in panel B. Regression equations are shown. The solid blue line indicates the line of best fit, dashed blue lines indicate 95% CI for the line of best fit, and the thin black line represents an x = y reference line.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare curvilinear-array (microconvex) and phased-array transducer results for total B-line scores (A; range of possible scores, 0 to 80) and the number of strong-positive sites (B; range of possible scores, 0 to 8) during LUS of dogs with L-CHF at 1 of 3 clinical time points (time point 1 [time of L-CHF diagnosis]; n = 13). Strong-positive sites were those with > 3 B lines. Overlapping values result in 12 visible data points in panel B. Regression equations are shown. The solid blue line indicates the line of best fit, dashed blue lines indicate 95% CI for the line of best fit, and the thin black line represents an x = y reference line.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare curvilinear-array (microconvex) and phased-array transducer results for total B-line scores (A; range of possible scores, 0 to 80) and the number of strong-positive sites (B; range of possible scores, 0 to 8) during LUS of dogs with L-CHF at 1 of 3 clinical time points (time point 1 [time of L-CHF diagnosis]; n = 13). Strong-positive sites were those with > 3 B lines. Overlapping values result in 12 visible data points in panel B. Regression equations are shown. The solid blue line indicates the line of best fit, dashed blue lines indicate 95% CI for the line of best fit, and the thin black line represents an x = y reference line.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare curvilinear-array (microconvex) and phased-array transducer results for total B-line scores (A; range of possible scores, 0 to 80) and the number of strong-positive sites (B; range of possible scores, 0 to 8) during LUS of dogs with L-CHF at 1 of 3 clinical time points (time point 1 [time of L-CHF diagnosis]; n = 13). Strong-positive sites were those with > 3 B lines. Overlapping values result in 12 visible data points in panel B. Regression equations are shown. The solid blue line indicates the line of best fit, dashed blue lines indicate 95% CI for the line of best fit, and the thin black line represents an x = y reference line.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare curvilinear-array (microconvex) and phased-array transducer results for total B-line scores (A; range of possible scores, 0 to 80) and the number of strong-positive sites (B; range of possible scores, 0 to 8) during LUS of dogs with L-CHF at 1 of 3 clinical time points (time point 1 [time of L-CHF diagnosis]; n = 13). Strong-positive sites were those with > 3 B lines. Overlapping values result in 12 visible data points in panel B. Regression equations are shown. The solid blue line indicates the line of best fit, dashed blue lines indicate 95% CI for the line of best fit, and the thin black line represents an x = y reference line.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 3 dogs with L-CHF by use of the curvilinear-array (microconvex) transducer (A, C, and E) and corresponding images of the same LUS sites obtained with the phased-array transducer (B, D, and F). A and B—Both transducers reveal absence of B lines in the left middle site of 1 dog. C and D—A single B line is apparent with both transducers in the left middle site of another dog. E and F—Infinite B lines (representing a strong-positive site) are seen with both transducers in the right middle site of a third dog. LMd = Left middle site. RMd = Right middle site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Image-quality evaluations
The median overall image-quality score for the curvilinear-array (microconvex) transducer was 30 (maximum possible score, 32) for both observers (range, 26 for 32 for observer 1 and 24 to 32 for observer 2). For the phased-array transducer, the median overall image-quality score was 22 (range, 13 to 28) for observer 1 and 17.5 (range, 7 to 27) for observer 2. The mixed-effects model results indicated that the curvilinear-array transducer images received higher overall quality scores than the phased-array transducer from both observers (P < 0.001), independent of the LUS site and time point. Examples of images obtained during LUS that had various quality scores are shown (Figure 4).
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Examples of 2-D still images obtained during LUS of 1 dog with L-CHF by use of the curvilinear-array (microconvex) transducer (A and C) and corresponding images of the same LUS sites obtained with the phased-array transducer (B and D). A and B—The curvilinear-array transducer image of the right caudal site shows a distinct pulmonary-pleural interface, presence of A lines and absence of B lines, and presence of both rib shadows (quality score, 4/4); in the corresponding phased-array image, A lines are distinctly seen, but rib shadows cannot be identified and it is difficult to determine the depth of the pulmonary-pleural interface in the near field (quality score, 2/4). C and D—The curvilinear-array transducer image shows a distinct pulmonary-pleural interface, > 3 B lines (representing a strong-positive result), and presence of both rib shadows at the perihilar site (quality score, 4/4); in the corresponding phased-array image, the pulmonary-pleural interface is visible and a strong-positive result is similarly seen, but only 1 rib shadow can be identified (quality score, 3/4). The pulmonary-pleural interface (arrows) and rib shadows (asterisks) are indicated. RCd = Right caudal site. RPh = Right perihilar site.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Interobserver agreement for total quality scores was moderate for both transducers. For the curvilinear-array (microconvex) transducer, the weighted κ value for total quality scores was 0.498 (95% CI, 0.211 to 0.786), and the values for individual sites ranged from 0.250 (95% CI, –0.260 to 0.761) for the right perihilar site to 0.875 (95% CI, 0.739 to 1.000) for the right caudal site. For the phased-array transducer, the weighted κ value for total quality scores was 0.517 (95% CI, 0.261 to 0.773), and the values for individual sites ranged from 0.194 (95% CI, –0.039 to 0.428) for the right caudal site to 0.773 (95% CI, 0.164 to 1.000) for the left cranial site.
Because interobserver agreement for image-quality scores was only moderate, scores from each observer were analyzed separately to assess for systematic bias between transducers. For observer 1, Passing-Bablok analysis revealed a significant positive constant bias toward higher total quality scores for the curvilinear-array (microconvex) transducer images (y-intercept, 32.0; 95% CI, 23.1 to 32.0), accompanied by a significant negative proportional bias (slope, 0.000; 95% CI, 0.000 to 0.375) for time point 1 observations (Figure 5). Similar biases were present for time point 2 observations (y-intercept, 30.7 [95% CI, 24.2 to 37.1]; slope, 0.000 [95% CI, –0.350 to 0.263). Total quality score comparisons for observer 2 followed similar patterns, with a significant positive constant bias and negative proportional bias for both time point 1 (y-intercept, 28.3 [95% CI, 19.5 to 33.2]; slope, 0.100 [95% CI, –0.002 to 0.500) and time point 2 (y-intercept, 35.9 [95% CI, 30.0 to 41.8]; slope, 0.390 [95% CI, –0.035 to 0.743]).
Results of Passing-Bablok analysis to compare total image-quality scores at time point 1 between the curvilinear-array (microconvex) and phased-array transducers for observer 1 (A) and observer 2 (B). Overlapping values result in 11 visible data points in panel A and 12 in panel B. The upper 95% confidence limit in both panels exceeds the limit of the y-axis. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare total image-quality scores at time point 1 between the curvilinear-array (microconvex) and phased-array transducers for observer 1 (A) and observer 2 (B). Overlapping values result in 11 visible data points in panel A and 12 in panel B. The upper 95% confidence limit in both panels exceeds the limit of the y-axis. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare total image-quality scores at time point 1 between the curvilinear-array (microconvex) and phased-array transducers for observer 1 (A) and observer 2 (B). Overlapping values result in 11 visible data points in panel A and 12 in panel B. The upper 95% confidence limit in both panels exceeds the limit of the y-axis. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare total image-quality scores at time point 1 between the curvilinear-array (microconvex) and phased-array transducers for observer 1 (A) and observer 2 (B). Overlapping values result in 11 visible data points in panel A and 12 in panel B. The upper 95% confidence limit in both panels exceeds the limit of the y-axis. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Results of Passing-Bablok analysis to compare total image-quality scores at time point 1 between the curvilinear-array (microconvex) and phased-array transducers for observer 1 (A) and observer 2 (B). Overlapping values result in 11 visible data points in panel A and 12 in panel B. The upper 95% confidence limit in both panels exceeds the limit of the y-axis. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 82, 8; 10.2460/ajvr.82.8.619
Discussion
Results of the present study indicated that both curvilinear-array (microconvex) and phased-array transducers can be used by experienced ultrasonographers in a clinical setting to quantify B lines during LUS in dogs with acute or resolving L-CHF. There were no significant differences between transducers in terms of the total B-line scores or number of strong-positive sites that were detected for the examination overall or for individual LUS sites. However, image-quality scores were consistently and significantly higher for the curvilinear-array transducer, suggesting that this transducer may be preferred for LUS if multiple transducers are available.
Previous studies of LUS in human and veterinary patients have not provided consistent recommendations regarding transducer type. The physician who described the association between B lines and alveolar-interstitial disease in 19971 and developed the Bedside Lung Ultrasound Examination protocol for LUS in human patients18 advocated the use of a curvilinear-array (microconvex) transducer for LUS, describing it as universal, ideal for LUS, and useful for multipurpose point-of-care evaluation of critically ill patients. In a meta-analysis17 of LUS for the diagnosis of cardiogenic pulmonary edema in people with acute dyspnea, most included studies (4/7) were performed with curvilinear-array (convex or microconvex) transducers, whereas 2 of 7 studies relied on phased-array (cardiac) transducers and 1 study did not have the transducer type reported. A meta-analysis34 that investigated the use of LUS for the diagnosis of pneumonia in people included 10 studies, all of which used curvilinear-array (microconvex [n = 9] or convex [1]) transducers. Other studies of LUS in people that were not included in these meta-analyses vary, with some investigative teams consistently using curvilinear-array transducers13,35 while others use phased-array19 or linear-array11 transducers. An international consensus statement on thoracic ultra-sonography in people2 recommends the use of curvilinear-array transducers for the detection of pleural effusion and pneumothorax, but does not make a specific recommendation regarding transducer type for detection of B lines.
In the veterinary literature, the most widely used LUS protocol, termed the Veterinary Brief Lung Ultra-sound Examination and modeled after the protocol used in human medicine, involves the use of a curvilinear-array (microconvex) transducer.4,5,6,21 However, other protocols to detect B lines in dogs with linear-array23,24 or phased-array25 transducers have been described. The latter study25 describes the use of a phased-array transducer from a subcostal window to detect B lines extending from the pericardial surface. The authors of that study25 mentioned that a potential benefit of this technique includes the ability of veterinary cardiologists to detect B lines during echocardiography without needing to change the ultrasound transducer or the patient's position.
The choice of transducer to perform LUS is obviously constrained by the transducer types available in a given clinical setting. Curvilinear-array (micro-convex) transducers are considered multipurpose transducers for general ultrasonography28,29 and are likely to be the first (and potentially only) transducer purchased in a general practice or emergency setting. In contrast, phased-array transducers may be the only transducer type available to veterinary cardiologists performing LUS as an extension of echocardiography. Results of the present study suggested that both transducers can be used to detect and quantify B lines. However, given the consistent discrepancy in image-quality scores between transducers, the authors recommend that practitioners preferentially use curvilinear-array (microconvex) transducers for LUS when available. Although not investigated in the present study, we consider that these transducers may be particularly useful for novice sonographers; by allowing better visualization of anatomic landmarks such as the pulmonary-pleural interface and rib shadows, these transducers may help facilitate proper transducer positioning and orientation for beginners, which is critical for obtaining diagnostic images. Phased-array transducers, in contrast, may be equally useful for LUS by more experienced sonographers (eg, cardiologists), who would also represent the population of sonographers most likely to have exclusive access to this transducer type. Overall, these recommendations are consistent with human LUS protocols for which curvilinear-array transducers appear to be most prevalent and preferred.
Interobserver agreement with respect to total B-line scores was excellent for both transducer types in the present study, with weighted κ values of 0.990 and 0.977 for curvilinear-array (microconvex) and phased-array transducer images, respectively. Inter-observer agreement for this variable at individual LUS sites was also good to excellent, with all κ values > 0.80 for both transducers. This level of agreement for quantifying B lines was similar to the findings of a previous report4 by our group on LUS in dogs and cats with respiratory distress, in which κ values were > 0.850 for each LUS site. These results were also consistent with the findings of studies12,36 that indicate a high degree of interobserver agreement for LUS results for human patients.
The study reported here had several limitations. We assessed interobserver agreement for images obtained with LUS by comparing interpretations from 2 independent observers without attempting to assess intraobserver agreement. Both observers in this study had extensive prior experience performing LUS; furthermore, these individuals were a cardiologist and a resident supervised by that cardiolo-gist, introducing the possibility of conformity bias. Results for interobserver agreement may have been different if the studied observers were less experienced, although the investigators of a previous study4 reported excellent interobserver agreement for results of LUS even when comparing interpretations of an experienced observer with those of an inexperienced observer. The ultrasonographic examinations in the present study were always performed in the same order and were later interpreted in the same order (complete examination with the phased-array transducer followed by complete examination with the curvilinear-array [microconvex] transducer); lack of randomization may have introduced bias during image acquisition or interpretation. Only isopropyl alcohol was used to facilitate transducer contact, and fur was not shaved for any dog; it is possible that the presence of fur and lack of coupling gel could have reduced the image quality disproportionately for one transducer type versus the other, although this does not reflect the authors' experience in other clinical situations when the transducer contact site is shaved of fur or coupling gel is used. Blinding in regard to transducer type was not possible when B lines and image quality were scored because the ultrasonographic characteristics of the images were distinctive to each transducer type. Although ultra-sound settings were standardized within a certain range for all examinations (frequency, 4 to 8 MHz; depth, 4 to 8 cm), there was occasionally mild variation between the curvilinear-array and phased-array settings within these ranges that could have affected image quality for one transducer versus the other. In addition, we evaluated only the 2 described transducer types; further comparison with linear-array transducers may have produced different results, as this transducer type may allow superior resolution of the near-field aspect of ultrasonographic images such as the pulmonary-pleural interface. Additionally, most examinations in this study represented findings for dogs with L-CHF that had either many B lines and many strong-positive sites (time point 1) or very few B lines and few or no strong-positive sites (time points 2 and 3) on LUS, which could have masked subtle differences between transducers for detecting rare or isolated B lines.
Another limitation to the present study was that the quality-scoring system for ultrasonographic images was subjective and, as our data indicated, observer dependent. To the best of our knowledge, there are no previously published or validated tools for objective quality assessment of images obtained by LUS; therefore, we developed a 5-point (0 to 4) scoring system specifically for use in this study. The goal was to focus on binary scoring (presence or absence) of anatomic features or ultrasound artifacts expected to be present in each image, rather than relying on subjective characterizations (such as an image appearing grainy or lacking visual contrast). However, interobserver agreement for total quality scores was only moderate, regardless of transducer type. Nonetheless, each observer consistently scored images from the curvilinear-array (microconvex) transducer as having greater quality than phased-array images. This internal consistency suggested that the curvilinear-array transducer used produced images of greater quality, despite an imperfect scoring method.
The study reported here took place as part of a larger investigation31 of the diagnostic utility of LUS to monitor resolution of L-CHF in dogs. The overarching study eventually enrolled 24 dogs and involved a total of 104 ultrasonographic examinations. The subset of data presented here represents examinations performed within the first 6 months after study enrollment, for which all LUS was performed with both curvilinear-array (microconvex) and phased-array transducers. As a result, the present study included variable numbers of ultrasonographic examinations (1 to 3) per dog. Although the number of times LUS was performed and the intervals between these procedures were not standardized, approximately half (13/24) were completed at time point 1 (the time of initial diagnosis of L-CHF), with the remainder occurring at time points 2 and 3 (the times of hospital discharge and the first recheck examination, respectively). This distribution of time points provided a clinically relevant range of findings with which to compare LUS transducer performance (many B lines and strong-positive sites at time point 1; few B lines or strong-positive sites at time points 2 and 3). After this initial 6-month period, all subsequent LUS examinations for the remainder of the larger study were performed with the curvilinear-array transducer only on the basis of the authors' subjective clinical impression of greater image quality when this transducer was used.
In dogs with active or resolving L-CHF in the present study, LUS performed with curvilinear-array (microconvex) or phased-array transducers provided images suitable for diagnostic purposes. Further research is warranted to confirm the findings that suggested curvilinear-array transducers may be preferable to phased-array transducers for performing LUS in dogs and to investigate the clinical impression that images obtained with the former transducer type may be more easily interpreted by inexperienced ultrasonographers.
Acknowledgments
Supported by an Iowa State University College of Veterinary Medicine Seed Grant.
Dr. Lisciandro is the owner of FASTVet.com, a private corporation that provides veterinary ultrasound training to practicing veterinarians. He has received ultrasound equipment on loan from SonoSite, Sound, EI Medical, and Scil Animal Care.
The present study was performed as part of a larger study,31 for which results were presented in abstract form at the American College of Veterinary Internal Medicine Summit, Phoenix, June 2020.
The authors thank Lori Moran and Allison Klein for technical assistance.
Footnotes
CX50 CompactXtreme Model 101855, Philips Healthcare, Andover, Mass.
C8-5, Philips Healthcare, Andover, Mass.
S8-3, Philips Healthcare, Andover, Mass.
Prism, version 8, Graphpad Software, La Jolla, Calif.
MedCalc, version 17.6, MedCalc Software, Seoul, South Korea.
Abbreviations
| L-CHF | Left-sided congestive heart failure |
| LUS | Lung ultrasonography |
References
- 1. ↑
Lichtenstein D, Meziere G, Biderman P, et al. The comet-tail artifact: an ultrasound sign of alveolar-interstitial syndrome. Am J Respir Crit Care Med 1997;156:1640–1646.
- 2. ↑
Volpicelli G, Elbarbary M, Blaivas M, et al. International evidence-based recommendations for point-of-care lung ultra-sound. Intensive Care Med 2012;38:577–591.
- 3. ↑
Prosen G, Klemen P, Štrnad M, et al. Combination of lung ultrasound (a comet-tail sign) and N-terminal pro-brain natriuretic peptide in differentiating acute heart failure from chronic obstructive pulmonary disease and asthma as cause of acute dyspnea in prehospital emergency setting. Crit Care 2011;15:R114.
- 4. ↑
Ward JL, Lisciandro GR, Keene BW, et al. Accuracy of point-of-care lung ultrasonography for the diagnosis of cardiogenic pulmonary edema in dogs and cats with acute dyspnea. J Am Vet Med Assoc 2017;250:666–675.
- 5. ↑
Ward JL, Lisciandro GR, Ware WA, et al. Evaluation of point-of-care thoracic ultrasound and NT-proBNP for the diagnosis of congestive heart failure in cats with respiratory distress. J Vet Intern Med 2018;32:1530–1540.
- 6. ↑
Ward JL, Lisciandro GR, Ware WA, et al. Lung ultrasonography findings in dogs with various underlying causes of cough. J Am Vet Med Assoc 2019;255:574–583.
- 7. ↑
Vitturi N, Soattin M, Allemand E, et al. Thoracic ultrasonography: a new method for the work-up of patients with dyspnea. J Ultrasound 2011;14:147–151.
- 9. ↑
Gargani L, Frassi F, Soldati G, et al. Ultrasound lung comets for the differential diagnosis of acute cardiogenic dyspnoea: a comparison with natriuretic peptides. Eur J Heart Fail 2008;10:70–77.
- 10. ↑
Lisciandro GR, Fulton RM, Fosgate GT, et al. Frequency and number of B-lines using a regionally based lung ultrasound examination in cats with radiographically normal lungs compared to cats with left-sided congestive heart failure. J Vet Emerg Crit Care (San Antonio) 2017;27:499–505.
- 11. ↑
Manson WC, Bonz JW, Carmody K, et al. Identification of sonographic B-lines with linear transducer predicts elevated B-type natriuretic peptide level. West J Emerg Med 2011;12:102–106.
- 12.
Liteplo AS, Marill KA, Villen T, et al. Emergency thoracic ultra-sound in the differentiation of the etiology of shortness of breath (ETUDES): sonographic B-lines and N-terminal pro-brain-type natriuretic peptide in diagnosing congestive heart failure. Acad Emerg Med 2009;16:201–210.
- 13. ↑
Volpicelli G, Mussa A, Garofalo G, et al. Bedside lung ultra-sound in the assessment of alveolar-interstitial syndrome. Am J Emerg Med 2006;24:689–696.
- 14. ↑
Peris A, Tutino L, Zagli G, et al. The use of point-of-care bedside lung ultrasound significantly reduces the number of radiographs and computed tomography scans in critically ill patients. Anesth Analg 2010;111:687–692.
- 15. ↑
Gargani L, Lionetti V, DiCristofano C, et al. Early detection of acute lung injury uncoupled to hypoxemia in pigs using ultrasound lung comets. Crit Care Med 2007;35:2769–2774.
- 16. ↑
Ward JL, Lisciandro GR, DeFrancesco TC. Distribution of alveolar-interstitial syndrome in dogs and cats with respiratory distress as assessed by lung ultrasound versus thoracic radio-graphs. J Vet Emerg Crit Care (San Antonio) 2018;28:415– 428.
- 17. ↑
Al Deeb M, Barbic S, Featherstone R, et al. Point-of-care ultra-sonography for the diagnosis of acute cardiogenic pulmonary edema in patients presenting with acute dyspnea: a systematic review and meta-analysis. Acad Emerg Med 2014;21:843–852.
- 18. ↑
Lichtenstein D. Lung ultrasound in acute respiratory failure an introduction to the BLUE-protocol. Minerva Anestesiol 2009;75:313–317.
- 19. ↑
Jambrik Z, Monti S, Coppola V, et al. Usefulness of ultrasound lung comets as a nonradiologic sign of extravascular lung water. Am J Cardiol 2004;93:1265–1270.
- 20.
Platz E, Pivetta E, Merz AA, et al. Impact of device selection and clip duration on lung ultrasound assessment in patients with heart failure. Am J Emerg Med 2015;33:1552–1556.
- 21. ↑
Lisciandro GR, Fosgate GT, Fulton RM. Frequency and number of ultrasound lung rockets (B-lines) using a regionally based lung ultrasound examination named Vet BLUE (Veterinary Bedside Lung Ultrasound Exam) in dogs with radio-graphically normal lung findings. Vet Radiol Ultrasound 2014;55:315–322.
- 22. ↑
Armenise A, Boysen RS, Rudloff E, et al. Veterinary-focused assessment with sonography for trauma—airway, breathing, circulation, disability and exposure: a prospective observational study in 64 canine trauma patients. J Small Anim Pract 2019;60:173–182.
- 23. ↑
Rademacher N, Pariaut R, Pate J, et al. Transthoracic lung ultrasound in normal dogs and dogs with cardiogenic pulmonary edema: a pilot study. Vet Radiol Ultrasound 2014;55:447–452.
- 24. ↑
Vezzosi T, Mannucci T, Pistoresi A, et al. Assessment of lung ultrasound B-lines in dogs with different stages of chronic valvular heart disease. J Vet Intern Med 2017;31:700–704.
- 25. ↑
Hori Y, Yamashita Y, Sakakibara K, et al. Usefulness of pericardial lung ultrasonography for the diagnosis of cardiogenic pulmonary edema in dogs. Am J Vet Res 2020;81:227–232.
- 26.
Neelis D, Mattoon J, Nyland T. Thorax. In: Mattoon J, Nyland T, eds. Small animal diagnostic ultrasound. 3rd ed. St Louis: Elsevier, 2015;188–216.
- 27.
Lisciandro GR. The Vet BLUE lung scan. In: Lisciandro GR, ed. Focused ultrasound techniques for the small animal practitioner. Ames, Iowa: Wiley-Blackwell, 2014;166–187.
- 28. ↑
Mattoon J, Nyland T. Fundamentals of diagnostic ultrasound. In: Mattoon J, Nyland T, ed. Small animal diagnostic ultra-sound. 3rd ed. St Louis: Elsevier, 2015;1–49.
- 29. ↑
Kremkau F. Transducers. In: Kremkau F, ed. Sonography: principles and instruments. 9th ed. St Louis: Elsevier, 2015;42–61.
- 30. ↑
Lisciandro G, Armenise A. Focused or COAST – Cardiopulmo-nary Resuscitation (CPR), Global FAST (GFAST) & the FAST-ABCDE Exam. In: Focused ultrasound for the small animal practitioner. Ames, Iowa: Wiley-Blackwell, 2014;269–285.
- 31. ↑
Murphy SD, Ward JL, Viall AK, et al. Utility of point-of-care lung ultrasound for monitoring cardiogenic pulmonary edema in dogs. J Vet Intern Med 2021;35:68–77.
- 32.
Lichtenstein DA, Mezière GA. Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest 2008;134:117–125.
- 34. ↑
Chavez MA, Shams N, Ellington LE, et al. Lung ultrasound for the diagnosis of pneumonia in adults: a systematic review and meta-analysis. Respir Res 2014;15:50.
- 35.
Volpicelli G, Caramello V, Cardinale L, et al. Bedside ultra-sound of the lung for the monitoring of acute decompensated heart failure. Am J Emerg Med 2008;26:585–591.
- 36.
Cibinel GA, Casoli G, Elia F, et al. Diagnostic accuracy and reproducibility of pleural and lung ultrasound in discriminating cardiogenic causes of acute dyspnea in the emergency department. Intern Emerg Med 2012;7:65–70.
