Diagnostic accuracy of a lung ultrasound protocol (Vet BLUE) for detection of pleural fluid, pneumothorax and lung pathology in dogs and cats

Retrospective evaluation of the agreement between thoracic point-of-care ultrasound and thoracic radiographs in cats with recent trauma: 111 cats

Introduction

Motor vehicular trauma, bite wounds, high-rise syndrome, and trauma of unknown origin are common reasons cats present to the emergency service. In small animals, thoracic injuries are often associated with trauma. The objective of this retrospective study was to evaluate limits of agreement (LOA) between thoracic point-of-care ultrasound (thoracic POCUS) and thoracic radiography (TXR), and to correlate thoracic POCUS findings to animal trauma triage (ATT) scores and subscores in a population of cats suffering from recent trauma.

Methods

Cats that had thoracic POCUS and TXR performed within 24 h of admission for suspected/witnessed trauma were retrospectively included. Thoracic POCUS and TXR findings were assessed as "positive" or "negative" based on the presence or absence of injuries. Cats positive on thoracic POCUS and TXR were assigned 1 to 5 tentative diagnoses: pulmonary contusions/hemorrhage, pneumothorax, pleural effusion, pericardial effusion, and diaphragmatic hernia. When available ATT scores were calculated. To express LOA between the two imaging modalities a kappa coefficient and 95% CI were calculated. Interpretation of kappa was based on Cohen values.

Results

One hundred and eleven cats were included. 83/111 (74.4%) cats were assessed as positive based on thoracic POCUS and/or TXR. Pulmonary contusion was the most frequent diagnosis. The LOA between thoracic POCUS and TXR were moderate for all combined injuries, moderate for pulmonary contusions/hemorrhage, pneumothorax, diaphragmatic hernia, and fair for pleural effusion. Cats with positive thoracic POCUS had significantly higher median ATT scores and respiratory subscores compared to negative thoracic POCUS cats.

Discussion

The frequency of detecting intrathoracic lesions in cats was similar between thoracic POCUS and TXR with fair to moderate LOA, suggesting thoracic POCUS is useful in cats suffering from trauma. Thoracic POCUS may be more beneficial in cats with higher ATT scores, particularly the respiratory score.

Deep learning models for interpretation of point of care ultrasound in military working dogs

Introduction

Military working dogs (MWDs) are essential for military operations in a wide range of missions. With this pivotal role, MWDs can become casualties requiring specialized veterinary care that may not always be available far forward on the battlefield. Some injuries such as pneumothorax, hemothorax, or abdominal hemorrhage can be diagnosed using point of care ultrasound (POCUS) such as the Global FAST® exam. This presents a unique opportunity for artificial intelligence (AI) to aid in the interpretation of ultrasound images. In this article, deep learning classification neural networks were developed for POCUS assessment in MWDs.

Methods

Images were collected in five MWDs under general anesthesia or deep sedation for all scan points in the Global FAST® exam. For representative injuries, a cadaver model was used from which positive and negative injury images were captured. A total of 327 ultrasound clips were captured and split across scan points for training three different AI network architectures: MobileNetV2, DarkNet-19, and ShrapML. Gradient class activation mapping (GradCAM) overlays were generated for representative images to better explain AI predictions.

Results

Performance of AI models reached over 82% accuracy for all scan points. The model with the highest performance was trained with the MobileNetV2 network for the cystocolic scan point achieving 99.8% accuracy. Across all trained networks the diaphragmatic hepatorenal scan point had the best overall performance. However, GradCAM overlays showed that the models with highest accuracy, like MobileNetV2, were not always identifying relevant features. Conversely, the GradCAM heatmaps for ShrapML show general agreement with regions most indicative of fluid accumulation.

Discussion

Overall, the AI models developed can automate POCUS predictions in MWDs. Preliminarily, ShrapML had the strongest performance and prediction rate paired with accurately tracking fluid accumulation sites, making it the most suitable option for eventual real-time deployment with ultrasound systems. Further integration of this technology with imaging technologies will expand use of POCUS-based triage of MWDs.

Recognition and Diagnosis of Underlying Disease Processes in Bacterial Pneumonia

Bacterial pneumonia can present in both dogs and cats as either acute or chronic disease, and clinical signs may suggest respiratory or systemic impairment. Aspiration pneumonia, a common form of disease, can be caused by esophageal or laryngeal disease, vomiting, or altered mentation, such as with CNS disease or general anesthesia. Bacterial pneumonia can also develop due to inhaled airway foreign bodies, reduced local immune function, or impaired clearance of respiratory secretions. Rapid identification of the etiology and associated risk factors for pneumonia can help veterinarians develop a targeted therapeutic plan and prevent recurrence of disease.

Evaluation of B-lines with 2 point-of-care lung ultrasound protocols in cats with radiographically normal lungs

OBJECTIVE

To compare 2 point-of-care lung ultrasound (LUS) protocols for quantification of B-lines in cats without evidence of respiratory disease based on history, physical examination, and thoracic radiography.

DESIGN

Prospective observational study.

SETTING

Single center, veterinary teaching hospital.

ANIMALS

Fifty-seven cats without respiratory disease based on history, physical examination, and thoracic radiographs.

INTERVENTIONS

All cats had 2 point-of-care LUS protocols performed bilaterally: a regional protocol (veterinary bedside lung ultrasound evaluation [VetBLUE]) and a more comprehensive vertical sweeping (VS) protocol. The total number of B-lines per cat, number of sites with B-lines, and maximal number of B-lines at each site were recorded and compared.

MEASUREMENTS AND MAIN RESULTS

Ten cats (18%) had at least 1 B-line identified with VetBLUE, versus 29 (51%) with VS. Comparing protocols, VS had a statistically higher total number of B-lines per cat, higher number of sites with B-lines, and higher maximal number of B-lines per site. B-lines that were too numerous to count were identified at a single location in 1 cat with VetBLUE and 2 cats with VS. A maximum of 3 B-lines were identified at all other positive sites regardless of the protocol used. On average, it took 1.79 times longer to complete VS bilaterally compared to VetBLUE (median [interquartile range]: 140 [33] and 78 [14] s, respectively) (P = 0.001).

CONCLUSIONS

This study demonstrates it is not uncommon to identify a single or even multiple B-lines in 1 or several sites on LUS in cats deemed to be clinically free of respiratory pathology-essential knowledge when using LUS as a screening test and to monitor intrathoracic lesions. In cats asymptomatic for respiratory disease, VS generally identifies more B-lines than VetBLUE, likely because it assesses a larger lung surface area. The sonographic identification of B-lines should be interpreted considering the LUS protocol used, history, and other diagnostics to determine their clinical significance.

The Rat Thoracic Ultrasound protocol: scanning technique and normal findings

Respiratory diseases (especially pneumonia) are very common disorders in pet rats. The suspected diagnosis is mostly based on the clinical signs, thoracic auscultation, and thoracic radiography. However, auscultation is insensitive in determining the severity of the disease, and radiographs are often unremarkable. Non-cardiac thoracic ultrasonography is increasingly used in veterinary medicine; however, it has not been described in detail in rats. Thoracic ultrasonic examination was conducted on 400 client-owned conscious pet rats. The rats were examined in the period from June 2023 to August 2023 in two veterinary clinics. Due to the small size of the animal, different anatomical considerations, and different evaluation protocols, as well as to meet the optimal outcome of detailed thoracic ultrasound, a standard methodological protocol was developed, and the name RATTUS (Rat Thoracic Ultrasound) was proposed. Typical signs of normal RATTUS were described (bat sign, lung sliding, A-lines, abdominal curtain sign, ski jump sign, lung pulse, seashore sign in M-mode, and bamboo sign). The new evaluation of lung inflation symmetry by substernal access was also described. The methodical approach presented and the normal findings description are proposed to be used for a standard/routine thoracic ultrasound examination in pet rats.

Lung Ultrasonography Does Not Distinguish between Interstitial and Alveolar Pulmonary Edema

For a long time, lung diseases have been considered the "forbidden zone" for ultrasound diagnosis because the lung is filled with gas, and the ultrasound waves are totally reflected when they encounter gas [...].

Spontaneous pneumothorax in a dog with a history of successfully treated heartworm disease: A case report

OBJECTIVE

To describe a dog presented with spontaneous pneumothorax secondary to chronic pulmonary changes associated with a history of resolved canine heartworm disease.

CASE OR SERIES SUMMARY

A 7-year-old 25.2kg female spayed German Shepherd mix was presented for management of spontaneous pneumothorax. The dog had a history of heartworm disease that underwent therapy prior to adoption, and the dog was heartworm antigen negative (SNAP 4Dx) during hospitalization for the pneumothorax. An exploratory thoracotomy was performed due to an unresolving pneumothorax requiring multiple thoracocenteses. Perioperatively, the lungs did not expand with positive pressure ventilation and diffuse, multifocal to coalescing areas of darkened tissue that were grossly consistent with necrosis and/or hemorrhage were noted. The dog was euthanized intraoperatively due to an assumed poor prognosis. Histopathologic examination findings were consistent with chronic reactive changes related to previous heartworm infection. No neoplastic or infectious etiologies were identified.

NEW OR UNIQUE INFORMATION PROVIDED

Spontaneous pneumothorax is a known complication of active heartworm infection. However, this case represents the first report of spontaneous pneumothorax secondary to chronic pulmonary changes caused by resolved heartworm infection.

The Role of Point-of-Care Ultrasound in Managing Cardiac Emergencies

Point-of-care ultrasound (POCUS) is a useful imaging tool for the diagnosis and monitoring of cardiac emergencies. Unlike complete echocardiography, POCUS is a time-sensitive examination involving a subset of targeted thoracic ultrasound views to identify abnormalities of the heart, lungs, pleural space, and caudal vena cava. When integrated with other clinical information, POCUS can be helpful in the diagnosis of left-sided and right-sided congestive heart failure, pericardial effusion and tamponade, and severe pulmonary hypertension and can help clinicians monitor resolution or recurrence of these conditions.

Assessing primary care veterinarians' use of and confidence in performing point-of-care ultrasound

BACKGROUND

Point-of-care ultrasound (POCUS) is gaining popularity in the veterinary field, but there is little information on operator confidence.

METHODS

A survey was distributed to primary care veterinarians (PCVs) via social media between May and July 2020. Details of participants' training in and use of POCUS were recorded. Participants' confidence in using thoracic and abdominal POCUS was also assessed using a five-point Likert scale.

RESULTS

Two hundred and one PCVs used POCUS, of which 32% reported using a non-standardised protocol. Fifty percent of PCVs were self-taught and 17.4% had attended a specific practical course. The median confidence score was 4 out of 5 (interquartile range [IQR] 2-5) for identifying abdominal abnormalities, irrespective of the training method. The median confidence score for thoracic abnormalities was 3 out of 5 (IQR 1-4) for those taught by a colleague or who were self-taught using journal articles or videos.

LIMITATIONS

The survey-based nature of the study relies on self-reporting and is therefore liable to recall bias.

CONCLUSIONS

PCVs' confidence in using POCUS is lacking, particularly with thoracic POCUS. Standardised practical training for PCVs, particularly in thoracic POCUS, would be beneficial. Future studies should explore how best to deliver this training.

A retrospective study on parapneumonic effusion in 130 dogs with a clinical diagnosis of pneumonia

Objective

To screen the occurrence of parapneumonic effusion in dogs.

Methods

Medical records were searched for dogs with a presumptive diagnosis of bacterial pneumonia from 2017 to 2021 at the Liege university teaching hospital. Bacterial pneumonia was presumptively diagnosed based on compatible clinical signs and findings; thoracic radiographs compatible with bacterial bronchopneumonia; and either increased serum C-reactive protein (CRP) levels, a positive bronchoalveolar lavage culture or a positive clinical evolution in response to antibiotic therapy. Patients diagnosed with parasitic or other non-bacterial inflammatory pneumonia or with pulmonary neoplasia were excluded. Signalment, clinical findings, and outcome were recorded.

Results

One hundred and thirty dogs were included in the study, of which 44 dogs (33.8%) developed a parapneumonic effusion. Four of these dogs (4/44; 9%) had thoracocentesis performed, displaying a modified transudate (2) or septic exudate (2).

Conclusions

Although parapneumonic effusion in dogs with a presumptive diagnosis of bacterial pneumonia appears to be rather common (33.8%), thoracocentesis or chest tube placement was rarely performed. Furthermore, the outcome of dogs with and without parapneumonic effusion appears to be similar.

Influence of concurrent lower respiratory tract disease on point-of-care lung ultrasound in small-breed dogs with myxomatous mitral valve disease

Background

Small‐breed dogs commonly have concurrent myxomatous mitral valve disease (MMVD) and lower respiratory tract disease (LRTD).

Hypothesis

Small‐breed dogs with preclinical MMVD and concurrent LRTD have more B‐lines on point‐of‐care lung ultrasound (POC‐LUS) compared to dogs without concurrent LRTD and are prone to misdiagnose as cardiogenic pulmonary edema (CPE).

Animals

A total of 114 small‐breed dogs with preclinical MMVD.

Methods

A prospective study was conducted, in which POC‐LUS was obtained and the number of B‐lines was calculated by a single clinician using the Veterinary Bedside Lung Ultrasound Examination protocol. The presence/absence of LRTD was assessed by clinicians blinded to the POC‐LUS results.

Results

Fifty and 64 dogs were in ACVIM stage B1 and B2, respectively. The presence of LRTD was prevalent in 74.6% (85/114) of small‐breed dogs with preclinical MMVD. When a previously reported criterion for CPE diagnosis (≥2 sites with >3 B‐lines/site) was applied, false‐positive results were observed in 15.8% (18/114) of dogs with preclinical MMVD. The summated number of B‐lines (3 vs. 1, P = .003), as well as the false‐positive rate (20% vs 3%, P = .04), were significantly higher in dogs with LRTD compared with dogs without LRTD. Multivariable logistic regression showed the presence of abnormalities other than B‐line on POC‐LUS (eg, thickened pleura or consolidation) could predict false‐positive results (odds ratio = 3.75, 95% confidence intervals 1.12‐12.54; P = .03) after adjustment for other clinical and echocardiographic factors.

Conclusions and Clinical Importance

Concurrent LRTD and abnormalities other than B‐lines should be considered in the interpretation of POC‐LUS in MMVD dogs.

Comparison of lung ultrasound, chest radiographs, C-reactive protein, and clinical findings in dogs treated for aspiration pneumonia

Background

Comparison of clinical findings, chest radiographs (CXR), lung ultrasound (LUS) findings, and C‐reactive protein (CRP) concentrations at admission and serial follow‐up in dogs with aspiration pneumonia (AP) is lacking.

Hypothesis

Lung ultrasound lesions in dogs with AP are similar to those described in humans with community‐acquired pneumonia (comAP); the severity of CXR and LUS lesions are similar; normalization of CRP concentration precedes resolution of imaging abnormalities and more closely reflects the clinical improvement of dogs.

Animals

Seventeen dogs with AP.

Methods

Prospective observational study. Clinical examination, CXR, LUS, and CRP measurements performed at admission (n = 17), 2 weeks (n = 13), and 1 month after diagnosis (n = 6). All dogs received antimicrobial therapy. Lung ultrasound and CXR canine aspiration scoring systems used to compare abnormalities.

Results

B‐lines and shred signs with or without bronchograms were identified on LUS in 14 of 17 and 16 of 17, at admission. Chest radiographs and LUS scores differed significantly using both canine AP scoring systems at each time point (18 regions per dog, P < .001). Clinical and CRP normalization occurred in all dogs during follow up. Shred signs disappeared on LUS in all but 1 of 6 dogs at 1 month follow‐up, while B‐lines and CXR abnormalities persisted in 4 of 6 and all dogs, respectively.

Conclusion and Clinical Importance

Lung ultrasound findings resemble those of humans with comAP and differ from CXR findings. Shred signs and high CRP concentrations better reflect clinical findings during serial evaluation of dogs.

A Comparative Study of Chest CT With Lung Ultrasound After SARS-CoV-2 Infection in the Assessment of Pulmonary Lesions in Rhesus Monkeys (Macaca Mulatta)

Lung ultrasound (LUS) is a fast and non-invasive modality for the diagnosis of several diseases. In humans, LUS is nowadays of additional value for bedside screening of hospitalized SARS-CoV-2 infected patients. However, the diagnostic value of LUS in SARS-CoV-2 infected rhesus monkeys, with mild-to-moderate disease, is unknown. The aim of this observational study was to explore correlations of the LUS appearance of abnormalities with COVID-19-related lesions detected on computed tomography (CT). There were 28 adult female rhesus monkeys infected with SARS-CoV-2 included in this study. Chest CT and LUS were obtained pre-infection and 2-, 7-, and 14-days post infection. Twenty-five animals were sub-genomic PCR positive in their nose/throat swab for at least 1 day. CT images were scored based on the degree of involvement for lung lobe. LUS was scored based on the aeration and abnormalities for each part of the lungs, blinded to CT findings. Most common lesions observed on CT were ground glass opacities (GGOs) and crazy paving patterns. With LUS, confluent or multiple B-lines with or without pleural abnormalities were observed which is corresponding with GGOs on CT. The agreement between the two modalities was similar over the examination days. Pleural line abnormalities were clearly observed with LUS, but could be easily missed on CT. Nevertheless, due to the air interface LUS was not able to examine the complete volume of the lung. The sensitivity of LUS was high though the diagnostic efficacy for mild-to-moderate disease, as seen in macaques, was relatively low. This leaves CT the imaging modality of choice for diagnosis, monitoring, and longitudinal assessment of a SARS-CoV-2 infection in macaques.

Lung Ultrasonography for Pneumothorax in Dogs and Cats

A sonographic diagnosis of pneumothorax (PTX) traditionally relies on excluding the presence of lung sliding, lung pulse, and/or B lines/lung consolidations, and identifying the lung point. However, these criteria can be difficult to identify, particularly in critically ill patients with respiratory disorders, and the lung point is infrequently used. Newer sonographic findings, such as mirrored ribs, reverse lung sliding, and abnormal curtain signs, have been identified to try to increase the accuracy of diagnosing PTX. This article describes and discusses the lung ultrasonography criteria used to diagnose PTX in both human and small animal patients.