A New Lung Ultrasound Protocol Able to Predict Worsening in Patients Affected by Severe Acute Respiratory Syndrome Coronavirus 2 Pneumonia

Deep learning for real-time multi-class segmentation of artefacts in lung ultrasound

Lung ultrasound (LUS) has emerged as a safe and cost-effective modality for assessing lung health, particularly during the COVID-19 pandemic. However, interpreting LUS images remains challenging due to its reliance on artefacts, leading to operator variability and limiting its practical uptake. To address this, we propose a deep learning pipeline for multi-class segmentation of objects (ribs, pleural line) and artefacts (A-lines, B-lines, B-line confluence) in ultrasound images of a lung training phantom. Lightweight models achieved a mean Dice Similarity Coefficient (DSC) of 0.74, requiring fewer than 500 training images. Applying this method in real-time, at up to 33.4 frames per second in inference, allows enhanced visualisation of these features in LUS images. This could be useful in providing LUS training and helping to address the skill gap. Moreover, the segmentation masks obtained from this model enable the development of explainable measures of disease severity, which have the potential to assist in the triage and management of patients. We suggest one such semi-quantitative measure called the B-line Artefact Score, which is related to the percentage of an intercostal space occupied by B-lines and in turn may be associated with the severity of a number of lung conditions. Moreover, we show how transfer learning could be used to train models for small datasets of clinical LUS images, identifying pathologies such as simple pleural effusions and lung consolidation with DSC values of 0.48 and 0.32 respectively. Finally, we demonstrate how such DL models could be translated into clinical practice, implementing the phantom model alongside a portable point-of-care ultrasound system, facilitating bedside assessment and improving the accessibility of LUS.

Machine Learning-based Voice Assessment for the Detection of Positive and Recovered COVID-19 Patients

Many virological tests have been implemented during the Coronavirus Disease 2019 (COVID-19) pandemic for diagnostic purposes, but they appear unsuitable for screening purposes. Furthermore, current screening strategies are not accurate enough to effectively curb the spread of the disease. Therefore, the present study was conducted within a controlled clinical environment to determine eventual detectable variations in the voice of COVID-19 patients, recovered and healthy subjects, and also to determine whether machine learning-based voice assessment (MLVA) can accurately discriminate between them, thus potentially serving as a more effective mass-screening tool. Three different subpopulations were consecutively recruited: positive COVID-19 patients, recovered COVID-19 patients and healthy individuals as controls. Positive patients were recruited within 10 days from nasal swab positivity. Recovery from COVID-19 was established clinically, virologically and radiologically. Healthy individuals reported no COVID-19 symptoms and yielded negative results at serological testing. All study participants provided three trials for multiple vocal tasks (sustained vowel phonation, speech, cough). All recordings were initially divided into three different binary classifications with a feature selection, ranking and cross-validated RBF-SVM pipeline. This brough a mean accuracy of 90.24%, a mean sensitivity of 91.15%, a mean specificity of 89.13% and a mean AUC of 0.94 across all tasks and all comparisons, and outlined the sustained vowel as the most effective vocal task for COVID discrimination. Moreover, a three-way classification was carried out on an external test set comprised of 30 subjects, 10 per class, with a mean accuracy of 80% and an accuracy of 100% for the detection of positive subjects. Within this assessment, recovered individuals proved to be the most difficult class to identify, and all the misclassified subjects were declared positive; this might be related to mid and short-term vocal traces of COVID-19, even after the clinical resolution of the infection. In conclusion, MLVA may accurately discriminate between positive COVID-19 patients, recovered COVID-19 patients and healthy individuals. Further studies should test MLVA among larger populations and asymptomatic positive COVID-19 patients to validate this novel screening technology and test its potential application as a potentially more effective surveillance strategy for COVID-19.

Low-complexity lung ultrasound video scoring by means of intensity projection-based video compression

Since the outbreak of COVID-19, efforts have been made towards semi-quantitative analysis of lung ultrasound (LUS) data to assess the patient's condition. Several methods have been proposed in this regard, with a focus on frame-level analysis, which was then used to assess the condition at the video and prognostic levels. However, no extensive work has been done to analyze lung conditions directly at the video level. This study proposes a novel method for video-level scoring based on compression of LUS video data into a single image and automatic classification to assess patient's condition. The method utilizes maximum, mean, and minimum intensity projection-based compression of LUS video data over time. This enables to preserve hyper- and hypo-echoic data regions, while compressing the video down to a maximum of three images. The resulting images are then classified using a convolutional neural network (CNN). Finally, the worst predicted score given among the images is assigned to the corresponding video. The results show that this compression technique can achieve a promising agreement at the prognostic level (81.62%), while the video-level agreement remains comparable with the state-of-the-art (46.19%). Conclusively, the suggested method lays down the foundation for LUS video compression, shifting from frame-level to direct video-level analysis of LUS data.

Can Artificial Intelligence Aid Diagnosis by Teleguided Point-of-Care Ultrasound? A Pilot Study for Evaluating a Novel Computer Algorithm for COVID-19 Diagnosis Using Lung Ultrasound

With the 2019 coronavirus disease (COVID-19) pandemic, there is an increasing demand for remote monitoring technologies to reduce patient and provider exposure. One field that has an increasing potential is teleguided ultrasound, where telemedicine and point-of-care ultrasound (POCUS) merge to create this new scope. Teleguided POCUS can minimize staff exposure while preserving patient safety and oversight during bedside procedures. In this paper, we propose the use of teleguided POCUS supported by AI technologies for the remote monitoring of COVID-19 patients by non-experienced personnel including self-monitoring by the patients themselves. Our hypothesis is that AI technologies can facilitate the remote monitoring of COVID-19 patients through the utilization of POCUS devices, even when operated by individuals without formal medical training. In pursuit of this goal, we performed a pilot analysis to evaluate the performance of users with different clinical backgrounds using a computer-based system for COVID-19 detection using lung ultrasound. The purpose of the analysis was to emphasize the potential of the proposed AI technology for improving diagnostic performance, especially for users with less experience.

Prevalence of exercise-induced oxygen desaturation after recovery from SARS-CoV-2 pneumonia and use of lung ultrasound to predict need for pulmonary rehabilitation

BACKGROUND

Persistence of breathlessness after recovery from SARS-CoV-2 pneumonia is frequent. Recovery from acute respiratory failure (ARF) is usually determined by normalized arterial blood gases (ABGs), but the prevalence of persistent exercise-induced desaturation (EID) and dyspnea is still unknown.

METHODS

We investigated the prevalence of EID in 70 patients with normal arterial oxygen at rest after recovery from ARF due to COVID-19 pneumonia. Patients underwent a 6-min walking test (6MWT) before discharge from hospital. We recorded dyspnea score and heart rate during 6MWT. We also investigated the possible role of lung ultrasound (LU) in predicting EID. Patients underwent a LU scan and scores for each explored area were summed to give a total LU score.

RESULTS

In 30 patients (43%), oxygen desaturation was >4% during 6MWT. These patients had significantly higher dyspnea and heart rate compared to non-desaturators. LU score >8.5 was significantly able to discriminate patients with EID.

CONCLUSION

In SARS-CoV-2 pneumonia, ABGs at discharge cannot predict the persistence of EID, which is frequent. LU may be useful to identify patients at risk who could benefit from a rehabilitation program.

Lung ultrasound in the COVID-19 era: a lesson to be learned for the future

Lung Ultrasound (LUS) is a reliable, radiation free and bedside imaging technique to assess several pulmonary diseases. Although the diagnosis of COVID-19 is made with the nasopharyngeal swab, detection of pulmonary involvement is key for a safe patient management. LUS is a valid alternative to explore, in paucisymptomatic self-presenting patients, the presence and extension of pneumonia compared to High Resolution Computed Tomography (HRCT) that represent the gold standard. This is a single-centre prospective study with 131 patients enrolled. Twelve lung areas were explored reporting a semiquantitative assessment to obtain the LUS score. Each patient performed reverse-transcription polymerase chain reaction test (rRT-PCR), hemogasanalysis and HRCT. We observed an inverse correlation between LUSs and pO2, P/F, SpO2, AaDO2 (p value < 0.01), a direct correlation with LUSs and AaDO2 (p value < 0.01). Compared with HRCT, LUS showed sensitivity and specificity of 81.8% and 55.4%, respectively, and VPN 75%, VPP 65%. Therefore, LUS can represent an effective alternative tool to detect pulmonary involvement in COVID-19 compared to HRCT.

Serial lung ultrasound in monitoring viral pneumonia: the lesson learned from COVID-19

Background

Lung ultrasound (LUS) has proven to be useful in the evaluation of lung involvement in COVID-19. However, its effectiveness for predicting the risk of severe disease is still up for debate. The aim of the study was to establish the prognostic accuracy of serial LUS examinations in the prediction of clinical deterioration in hospitalised patients with COVID-19.

Methods

Prospective single-centre cohort study of patients hospitalised for COVID-19. The study protocol consisted of a LUS examination within 24 h from admission and a follow-up examination on day 3 of hospitalisation. Lung involvement was evaluated by a 14-area LUS score. The primary end-point was the ability of LUS to predict clinical deterioration defined as need for intensive respiratory support with high-flow oxygen or invasive mechanical ventilation.

Results

200 patients were included and 35 (17.5%) of them reached the primary end-point and were transferred to the intensive care unit (ICU). The LUS score at admission had been significantly higher in the ICU group than in the non-ICU group (22 (interquartile range (IQR) 20-26) versus 12 (IQR 8-15)). A LUS score at admission ≥17 was shown to be the best cut-off point to discriminate patients at risk of deterioration (area under the curve (AUC) 0.95). The absence of progression in LUS score on day 3 significantly increased the prediction accuracy by ruling out deterioration with a negative predictive value of 99.29%.

Conclusion

Serial LUS is a reliable tool in predicting the risk of respiratory deterioration in patients hospitalised due to COVID-19 pneumonia. LUS could be further implemented in the future for risk stratification of viral pneumonia.

Benchmark methodological approach for the application of artificial intelligence to lung ultrasound data from COVID-19 patients: From frame to prognostic-level

Automated ultrasound imaging assessment of the effect of CoronaVirus disease 2019 (COVID-19) on lungs has been investigated in various studies using artificial intelligence-based (AI) methods. However, an extensive analysis of state-of-the-art Convolutional Neural Network-based (CNN) models for frame-level scoring, a comparative analysis of aggregation techniques for video-level scoring, together with a thorough evaluation of the capability of these methodologies to provide a clinically valuable prognostic-level score is yet missing within the literature. In addition to that, the impact on the analysis of the posterior probability assigned by the network to the predicted frames as well as the impact of temporal downsampling of LUS data are topics not yet extensively investigated. This paper takes on these challenges by providing a benchmark analysis of methods from frame to prognostic level. For frame-level scoring, state-of-the-art deep learning models are evaluated with additional analysis of best performing model in transfer-learning settings. A novel cross-correlation based aggregation technique is proposed for video and exam-level scoring. Results showed that ResNet-18, when trained from scratch, outperformed the existing methods with an F1-Score of 0.659. The proposed aggregation method resulted in 59.51%, 63.29%, and 84.90% agreement with clinicians at the video, exam, and prognostic levels, respectively; thus, demonstrating improved performances over the state of the art. It was also found that filtering frames based on the posterior probability shows higher impact on the LUS analysis in comparison to temporal downsampling. All of these analysis were conducted over the largest standardized and clinically validated LUS dataset from COVID-19 patients.

Lung ultrasound score severity cut-off points in COVID-19 pneumonia. A systematic review and validating cohort

Objectives

Our purpose was to establish different cut-off points based on the lung ultrasound score (LUS) to classify COVID-19 pneumonia severity.

Methods

Initially, we conducted a systematic review among previously proposed LUS cut-off points. Then, these results were validated by a single-centre prospective cohort study of adult patients with confirmed SARS-CoV-2 infection. Studied variables were poor outcome (ventilation support, intensive care unit admission or 28-days mortality) and 28-days mortality.

Results

From 510 articles, 11 articles were included. Among the cut-off points proposed in the articles included, only the LUS > 15 cut-off point could be validated for its original endpoint, demonstrating also the strongest relation with poor outcome (odds ratio [OR] = 3.636, confidence interval [CI] 1.411-9.374). Regarding our cohort, 127 patients were admitted. In these patients, LUS was statistically associated with poor outcome (OR = 1.303, CI 1.137-1.493), and with 28-days mortality (OR = 1.024, CI 1.006-1.042). LUS > 15 showed the best diagnostic performance when choosing a single cut-off point in our cohort (area under the curve 0.650). LUS ≤ 7 showed high sensitivity to rule out poor outcome (0.89, CI 0.695-0.955), while LUS > 20 revealed high specificity to predict poor outcome (0.86, CI 0.776-0.917).

Conclusions

LUS is a good predictor of poor outcome and 28-days mortality in COVID-19. LUS ≤ 7 cut-off point is associated with mild pneumonia, LUS 8-20 with moderate pneumonia and ≥20 with severe pneumonia. If a single cut-off point were used, LUS > 15 would be the point which better discriminates mild from severe disease.

COVID-19 Lung Ultrasound Scores and Lessons from the Pandemic: A Narrative Review

The WHO recently declared that COVID-19 no longer constitutes a public health emergency of international concern; however, lessons learned through the pandemic should not be left behind. Lung ultrasound was largely utilized as a diagnostic tool thanks to its feasibility, easy application, and the possibility to reduce the source of infection for health personnel. Lung ultrasound scores consist of grading systems used to guide diagnosis and medical decisions, owning a good prognostic value. In the emergency context of the pandemic, several lung ultrasound scores emerged either as new scores or as modifications of pre-existing ones. Our aim is to clarify the key aspects of lung ultrasound and lung ultrasound scores to standardize their clinical use in a non-pandemic context. The authors searched on PubMed for articles related to "COVID-19", "ultrasound", and "Score" until 5 May 2023; other keywords were "thoracic", "lung", "echography", and "diaphragm". A narrative summary of the results was made. Lung ultrasound scores are demonstrated to be an important tool for triage, prediction of severity, and aid in medical decisions. Ultimately, the existence of numerous scores leads to a lack of clarity, confusion, and an absence of standardization.

Human-to-AI Interrater Agreement for Lung Ultrasound Scoring in COVID-19 Patients

OBJECTIVES

Lung ultrasound (LUS) has sparked significant interest during COVID-19. LUS is based on the detection and analysis of imaging patterns. Vertical artifacts and consolidations are some of the recognized patterns in COVID-19. However, the interrater reliability (IRR) of these findings has not been yet thoroughly investigated. The goal of this study is to assess IRR in LUS COVID-19 data and determine how many LUS videos and operators are required to obtain a reliable result.

METHODS

A total of 1035 LUS videos from 59 COVID-19 patients were included. Videos were randomly selected from a dataset of 1807 videos and scored by six human operators (HOs). The videos were also analyzed by artificial intelligence (AI) algorithms. Fleiss' kappa coefficient results are presented, evaluated at both the video and prognostic levels.

RESULTS

Findings show a stable agreement when evaluating a minimum of 500 videos. The statistical analysis illustrates that, at a video level, a Fleiss' kappa coefficient of 0.464 (95% confidence interval [CI] = 0.455-0.473) and 0.404 (95% CI = 0.396-0.412) is obtained for pairs of HOs and for AI versus HOs, respectively. At prognostic level, a Fleiss' kappa coefficient of 0.505 (95% CI = 0.448-0.562) and 0.506 (95% CI = 0.458-0.555) is obtained for pairs of HOs and for AI versus HOs, respectively.

CONCLUSIONS

To examine IRR and obtain a reliable evaluation, a minimum of 500 videos are recommended. Moreover, the employed AI algorithms achieve results that are comparable with HOs. This research further provides a methodology that can be useful to benchmark future LUS studies.

Lung ultrasound monitoring: impact on economics and outcomes

PURPOSE OF REVIEW

This review aims to summarize the impact of lung ultrasonography (LUS) on economics and possible impact on patients' outcomes, proven its diagnostic accuracy in patients with acute respiratory failure.

RECENT FINDINGS

Despite some previous ethical concerns on LUS examination, today this technique has showed several advantages. First, it is now clear that the daily use of LUS can provide a relevant cost reduction in healthcare of patients with acute respiratory failure, while reducing the risk of transport of patients to radiological departments for chest CT scan. In addition, LUS reduces the exposition to x-rays since can replace the bedside chest X-ray examination in many cases. Indeed, LUS is characterized by a diagnostic accuracy that is even superior to portable chest X-ray when performed by well trained personnel. Finally, LUS examination is a useful tool to predict the course of patients with pneumonia, including the need for hospitalization and ICU admission, noninvasive ventilation failure and orotracheal intubation, weaning success, and mortality.

SUMMARY

LUS should be implemented not only in Intensive Care Units, but also in other setting like emergency departments. Since most data comes from the recent coronavirus disease 2019 pandemic, further investigations are required in Acute Respiratory Failure of different etiologies.

New International Guidelines and Consensus on the Use of Lung Ultrasound

Following the innovations and new discoveries of the last 10 years in the field of lung ultrasound (LUS), a multidisciplinary panel of international LUS experts from six countries and from different fields (clinical and technical) reviewed and updated the original international consensus for point-of-care LUS, dated 2012. As a result, a total of 20 statements have been produced. Each statement is complemented by guidelines and future developments proposals. The statements are furthermore classified based on their nature as technical (5), clinical (11), educational (3), and safety (1) statements.

Vertical Artifacts as Lung Ultrasound Signs: Trick or Trap? Part 2- An Accademia di Ecografia Toracica Position Paper on B-Lines and Sonographic Interstitial Syndrome

Although during the last few years the lung ultrasound (LUS) technique has progressed substantially, several artifacts, which are currently observed in clinical practice, still need a solid explanation of the physical phenomena involved in their origin. This is particularly true for vertical artifacts, conventionally known as B-lines, and for their use in clinical practice. A wider consensus and a deeper understanding of the nature of these artifactual phenomena will lead to a better classification and a shared nomenclature, and, ultimately, result in a more objective correlation between anatomo-pathological data and clinical scenarios. The objective of this review is to collect and document the different signs and artifacts described in the history of chest ultrasound, with a particular focus on vertical artifacts (B-lines) and sonographic interstitial syndrome (SIS). By reviewing the possible physical and anatomical interpretation of the signs and artifacts proposed in the literature, this work also aims to bring order to the available studies and to present the AdET (Accademia di Ecografia Toracica) viewpoint in terms of nomenclature and clinical approach to the SIS.

Lung ultrasound in the diagnosis of COVID-19-associated pneumonia

Over the past decades, lung ultrasound in the diagnosis of lung diseases has become widespread. Ultrasound examination has a number of advantages (no radiation exposure, real-time imaging, clear visualization of the subpleural lung regions and costophrenic angles), which make it possible to use ultrasound to monitor the dynamics of pneumonia in children and pregnant women. Currently, in the context of the COVID-19 pandemic, lung ultrasound is widely used due to its high diagnostic efficiency, which is comparable with classical radiography and X-ray computed tomography (CT) by a number of parameters.

The article describes the method of lung ultrasound and the radiographic pattern of COVID-19-associated pneumonia. It also provides a review of the literature, according to which the severity of pneumonia was determined, depending on the radiographic pattern, and the need for a lung ultrasound was identified.

The article indicates that information on assessment of the radiographic pattern of the lungs at runtime in different variants of the course of coronavirus infection, as well as many methodological issues, including the frequency of second-look lung ultrasound, has not been sufficiently studied.

Multi-objective automatic analysis of lung ultrasound data from COVID-19 patients by means of deep learning and decision trees

COVID-19 raised the need for automatic medical diagnosis, to increase the physicians' efficiency in managing the pandemic. Among all the techniques for evaluating the status of the lungs of a patient with COVID-19, lung ultrasound (LUS) offers several advantages: portability, cost-effectiveness, safety. Several works approached the automatic detection of LUS imaging patterns related COVID-19 by using deep neural networks (DNNs). However, the decision processes based on DNNs are not fully explainable, which generally results in a lack of trust from physicians. This, in turn, slows down the adoption of such systems. In this work, we use two previously built DNNs as feature extractors at the frame level, and automatically synthesize, by means of an evolutionary algorithm, a decision tree (DT) that aggregates in an interpretable way the predictions made by the DNNs, returning the severity of the patients' conditions according to a LUS score of prognostic value. Our results show that our approach performs comparably or better than previously reported aggregation techniques based on an empiric combination of frame-level predictions made by DNNs. Furthermore, when we analyze the evolved DTs, we discover properties about the DNNs used as feature extractors. We make our data publicly available for further development and reproducibility.

Lack of Association Between a Quantified Lung Ultrasound Score and Illness Severity in Pediatric Emergency Department Patients With Acute Lower Respiratory Infections

OBJECTIVES

Lung ultrasound (LUS) may help determine illness severity in children with acute lower respiratory tract infections (LRTI) but limited pediatric studies exist. Our objective was to determine the association between LUS findings and illness severity in children with LRTI.

METHODS

We conducted a prospective study of patients <20 years with LRTI. Trained investigators performed standardized LUS examinations of 12 regions. Blinded sonologists reviewed examinations for individual pathologic features and also calculated a Quantified Lung Ultrasound Score (QLUS). We defined focal severity as QLUS of ≥2 in ≥1 region, and diffuse severity as QLUS of ≥1 in ≥3 regions. The primary outcome was the Respiratory component of the Pediatric Early Warning Score (RPEWS), a 14-item scale measuring respiratory illness severity. Secondary outcomes included hospital admission, length of stay, supplemental oxygen, and antibiotic use.

RESULTS

We enrolled 85 patients with LRTIs, 46 (54%) whom were hospitalized (5.4% intensive care). Median RPEWS was 1 (interquartile range 2). Neither individual features on ultrasound nor total QLUS were associated with RPEWS, hospitalization, length of stay, or oxygen use. Mean RPEWS was similar for participants regardless of focal (1.46 versus 1.26, P = .57) or diffuse (1.47 versus 1.21, P = .47) severity findings, but those with focal or diffuse severity, or isolated consolidation, had greater antibiotic administration (P < .001).

CONCLUSIONS

In children with LRTI, neither individual features nor QLUS were associated with illness severity. Antibiotics were more likely in patients with either focal or diffuse severity or presence of consolidation on ultrasound.

State of the Art in Lung Ultrasound, Shifting from Qualitative to Quantitative Analyses

Lung ultrasound (LUS) has been increasingly expanding since the 1990s, when the clinical relevance of vertical artifacts was first reported. However, the massive spread of LUS is only recent and is associated with the coronavirus disease 2019 (COVID-19) pandemic, during which semi-quantitative computer-aided techniques were proposed to automatically classify LUS data. In this review, we discuss the state of the art in LUS, from semi-quantitative image analysis approaches to quantitative techniques involving the analysis of radiofrequency data. We also discuss recent in vitro and in silico studies, as well as research on LUS safety. Finally, conclusions are drawn highlighting the potential future of LUS.

Automatic deep learning-based consolidation/collapse classification in lung ultrasound images for COVID-19 induced pneumonia

Our automated deep learning-based approach identifies consolidation/collapse in LUS images to aid in the identification of late stages of COVID-19 induced pneumonia, where consolidation/collapse is one of the possible associated pathologies. A common challenge in training such models is that annotating each frame of an ultrasound video requires high labelling effort. This effort in practice becomes prohibitive for large ultrasound datasets. To understand the impact of various degrees of labelling precision, we compare labelling strategies to train fully supervised models (frame-based method, higher labelling effort) and inaccurately supervised models (video-based methods, lower labelling effort), both of which yield binary predictions for LUS videos on a frame-by-frame level. We moreover introduce a novel sampled quaternary method which randomly samples only 10% of the LUS video frames and subsequently assigns (ordinal) categorical labels to all frames in the video based on the fraction of positively annotated samples. This method outperformed the inaccurately supervised video-based method and more surprisingly, the supervised frame-based approach with respect to metrics such as precision-recall area under curve (PR-AUC) and F1 score, despite being a form of inaccurate learning. We argue that our video-based method is more robust with respect to label noise and mitigates overfitting in a manner similar to label smoothing. The algorithm was trained using a ten-fold cross validation, which resulted in a PR-AUC score of 73% and an accuracy of 89%. While the efficacy of our classifier using the sampled quaternary method significantly lowers the labelling effort, it must be verified on a larger consolidation/collapse dataset, our proposed classifier using the sampled quaternary video-based method is clinically comparable with trained experts' performance.

Reliability and clinical correlations of semi-quantitative lung ultrasound on BLUE points in COVID-19 mechanically ventilated patients: The 'BLUE-LUSS'-A feasibility clinical study

INTRODUCTION

Bedside lung ultrasound has gained a key role in each segment of the treatment chain during the COVID-19 pandemic. During the diagnostic assessment of the critically ill patients in ICUs, it is highly important to maximize the amount and quality of gathered information while minimizing unnecessary interventions (e.g. moving/rotating the patient). Another major factor is to reduce the risk of infection and the workload of the staff.

OBJECTIVES

To serve these significant issues we constructed a feasibility study, in which we used a single-operator technique without moving the patient, only assessing the easily achievable lung regions at conventional BLUE points. We hypothesized that calculating this 'BLUE lung ultrasound score' (BLUE-LUSS) is a reasonable clinical tool. Furthermore, we used both longitudinal and transverse scans to measure their reliability and assessed the interobserver variability as well.

METHODS

University Intensive Care Unit based, single-center, prospective, observational study was performed on 24 consecutive SARS-CoV2 RT-PCR positive, mechanically ventilated critically ill patients. Altogether 400 loops were recorded, rated and assessed off-line by 4 independent intensive care specialists (each 7+ years of LUS experience).

RESULTS

Intraclass correlation values indicated good reliability for transversal and longitudinal qLUSS scores, while we detected excellent interrater agreement of both cLUSS calculation methods. All of our LUS scores correlated inversely and significantly to the P/F values. Best correlation was achieved in the case of longitudinal qLUSS (r = -0.55, p = 0.0119).

CONCLUSION

Summarized score of BLUE-LUSS can be an important, easy-to-perform adjunct tool for assessing and quantifying lung pathology in critically ill ventilated patients at bedside, especially for the P/F ratio. The best agreement for the P/F ratio can be achieved with the longitudinal scans. Regarding these findings, assessing BLUE-points can be extended with the BLUE-LUSS for daily routine using both transverse and longitudinal views.

Lung Ultrasound in COVID-19 and Post-COVID-19 Patients, an Evidence-Based Approach

OBJECTIVES

Worldwide, lung ultrasound (LUS) was utilized to assess coronavirus disease 2019 (COVID-19) patients. Often, imaging protocols were however defined arbitrarily and not following an evidence-based approach. Moreover, extensive studies on LUS in post-COVID-19 patients are currently lacking. This study analyses the impact of different LUS imaging protocols on the evaluation of COVID-19 and post-COVID-19 LUS data.

METHODS

LUS data from 220 patients were collected, 100 COVID-19 positive and 120 post-COVID-19. A validated and standardized imaging protocol based on 14 scanning areas and a 4-level scoring system was implemented. We utilized this dataset to compare the capability of 5 imaging protocols, respectively based on 4, 8, 10, 12, and 14 scanning areas, to intercept the most important LUS findings. This to evaluate the optimal trade-off between a time-efficient imaging protocol and an accurate LUS examination. We also performed a longitudinal study, aimed at investigating how to eventually simplify the protocol during follow-up. Additionally, we present results on the agreement between AI models and LUS experts with respect to LUS data evaluation.

RESULTS

A 12-areas protocol emerges as the optimal trade-off, for both COVID-19 and post-COVID-19 patients. For what concerns follow-up studies, it appears not to be possible to reduce the number of scanning areas. Finally, COVID-19 and post-COVID-19 LUS data seem to show differences capable to confuse AI models that were not trained on post-COVID-19 data, supporting the hypothesis of the existence of LUS patterns specific to post-COVID-19 patients.

CONCLUSIONS

A 12-areas acquisition protocol is recommended for both COVID-19 and post-COVID-19 patients, also during follow-up.

Current Advances in Lung Ultrasound in COVID-19 Critically Ill Patients: A Narrative Review

Lung ultrasound (LUS) has a relatively recent democratization due to the better availability and training of physicians, especially in intensive care units. LUS is a relatively cheap and easy-to-learn and -use bedside technique that evaluates pulmonary morphology when using simple algorithms. During the global COVID-19 pandemic, LUS was found to be an accurate tool to quickly diagnose, triage and monitor patients with COVID-19 pneumonia. This paper aims to provide a comprehensive review of LUS use during the COVID-19 pandemic. The first section of our work defines the technique, the practical approach and the semeiotic signs of LUS examination. The second section exposed the COVID-19 pattern in LUS examination and the difference between the differential diagnosis patterns and the well-correlation found with computer tomography scan findings. In the third section, we described the utility of LUS in the management of COVID-19 patients, allowing an early diagnosis and triage in the emergency department, as the monitoring of pneumonia course (pneumonia progression, alveolar recruitment, mechanical ventilation weaning) and detection of secondary complications (pneumothorax, superinfection). Moreover, we describe the usefulness of LUS as a marker of the prognosis of COVID-19 pneumonia in the fourth section. Finally, the 5th part is focused on describing the interest of the LUS, as a non-ionized technique, in the management of pregnant COVID-19 women.

The Role of Lung Ultrasound in SARS-CoV-19 Pneumonia Management

Purpose: We aimed to assess the role of lung ultrasound (LUS) in the diagnosis and prognosis of SARS-CoV-2 pneumonia, by comparing it with High Resolution Computed Tomography (HRCT). Patients and methods: All consecutive patients with laboratory-confirmed SARS-CoV-2 infection and hospitalized in COVID Centers were enrolled. LUS and HRCT were carried out on all patients by expert operators within 48−72 h of admission. A four-level scoring system computed in 12 regions of the chest was used to categorize the ultrasound imaging, from 0 (absence of visible alterations with ultrasound) to 3 (large consolidation and cobbled pleural line). Likewise, a semi-quantitative scoring system was used for HRCT to estimate pulmonary involvement, from 0 (no involvement) to 5 (>75% involvement for each lobe). The total CT score was the sum of the individual lobar scores and ranged from 0 to 25. LUS scans were evaluated according to a dedicated scoring system. CT scans were assessed for typical findings of COVID-19 pneumonia (bilateral, multi-lobar lung infiltration, posterior peripheral ground glass opacities). Oxygen requirement and mortality were also recorded. Results: Ninety-nine patients were included in the study (male 68.7%, median age 71). 40.4% of patients required a Venturi mask and 25.3% required non-invasive ventilation (C-PAP/Bi-level). The overall mortality rate was 21.2% (median hospitalization 30 days). The median ultrasound thoracic score was 28 (IQR 20−36). For the CT evaluation, the mean score was 12.63 (SD 5.72), with most of the patients having LUS scores of 2 (59.6%). The bivariate correlation analysis displayed statistically significant and high positive correlations between both the CT and composite LUS scores and ventilation, lactates, COVID-19 phenotype, tachycardia, dyspnea, and mortality. Moreover, the most relevant and clinically important inverse proportionality in terms of P/F, i.e., a decrease in P/F levels, was indicative of higher LUS/CT scores. Inverse proportionality P/F levels and LUS and TC scores were evaluated by univariate analysis, with a P/F−TC score correlation coefficient of −0.762, p < 0.001, and a P/F−LUS score correlation coefficient of −0.689, p < 0.001. Conclusions: LUS and HRCT show a synergistic role in the diagnosis and disease severity evaluation of COVID-19.

Deep Learning-Based Classification of Reduced Lung Ultrasound Data From COVID-19 Patients

The application of lung ultrasound (LUS) imaging for the diagnosis of lung diseases has recently captured significant interest within the research community. With the ongoing COVID-19 pandemic, many efforts have been made to evaluate LUS data. A four-level scoring system has been introduced to semiquantitatively assess the state of the lung, classifying the patients. Various deep learning (DL) algorithms supported with clinical validations have been proposed to automate the stratification process. However, no work has been done to evaluate the impact on the automated decision by varying pixel resolution and bit depth, leading to the reduction in size of overall data. This article evaluates the performance of DL algorithm over LUS data with varying pixel and gray-level resolution. The algorithm is evaluated over a dataset of 448 LUS videos captured from 34 examinations of 20 patients. All videos are resampled by a factor of 2, 3, and 4 of original resolution, and quantized to 128, 64, and 32 levels, followed by score prediction. The results indicate that the automated scoring shows negligible variation in accuracy when it comes to the quantization of intensity levels only. Combined effect of intensity quantization with spatial down-sampling resulted in a prognostic agreement ranging from 73.5% to 82.3%.These results also suggest that such level of prognostic agreement can be achieved over evaluation of data reduced to 32 times of its original size. Thus, laying foundation to efficient processing of data in resource constrained environments.

Risk Stratification in COVID-19 Pneumonia - Determining the Role of Lung Ultrasound

LUS patterns of COVID-19 pneumonia have been described and shown to be characteristic. The aim of the study was to predict the prognosis of patients with COVID-19 pneumonia, using a score based on LUS findings.

MATERIALS AND METHODS

 An observational, retrospective study was conducted on patients admitted to Niguarda hospital with a diagnosis of COVID-19 pneumonia during the period of a month, from March 2^nd to April 3^rd 2020. Demographics, clinical, laboratory, and radiological findings were collected. LUS was performed in all patients. The chest was divided into 12 areas. The LUS report was drafted using a score from 0 to 3 with 0 corresponding to A pattern, 1 corresponding to well separated vertical artifacts (B lines), 2 corresponding to white lung and small consolidations, 3 corresponding to wide consolidations. The total score results from the sum of the scores for each area. The primary outcome was endotracheal intubation, no active further management, or death. The secondary outcome was discharge from the emergency room (ER).

RESULTS

 255 patients were enrolled. 93.7 % had a positive LUS. ETI was performed in 43 patients, and 24 received a DNI order. The general mortality rate was 15.7 %. Male sex (OR 3.04, p = 0.014), cardiovascular disease and hypertension (OR 2.75, p = 0.006), P/F (OR 0.99, p < 0.001) and an LUS score > 20 (OR 2.52, p = 0.046) were independent risk factors associated with the primary outcome. Receiver operating characteristic (ROC) curve analysis for an LUS score > 20 was performed with an AUC of 0.837. Independent risk factors associated with the secondary outcome were age (OR 0.96, p = 0.073), BMI (OR 0.87, p = 0,13), P/F (OR 1.03, p < 0.001), and LUS score < 10 (OR 20.9, p = 0.006). ROC curve analysis was performed using an LUS score < 10 with an AUC 0.967.

CONCLUSION

 The extent of lung abnormalities evaluated by LUS score is a predictor of a worse outcome, ETI, or death. Moreover, the LUS score could be an additional tool for the safe discharge of patient from the ER.

Lung ultrasound findings following COVID-19 hospitalization: A prospective longitudinal cohort study

Background

Lung ultrasound (LUS) is a useful tool for diagnosis and monitoring in patients with active COVID-19-infection. However, less is known about the changes in LUS findings after a hospitalization for COVID-19.

Methods

In a prospective, longitudinal study in patients with COVID-19 enrolled from non-ICU hospital units, adult patients underwent 8-zone LUS and blood sampling both during the hospitalization and 2–3 months after discharge. LUS images were analyzed blinded to clinical variables and outcomes.

Results

A total of 71 patients with interpretable LUS at baseline and follow up (mean age 64 years, 61% male, 24% with acute respiratory distress syndrome (ARDS)) were included. The follow-up LUS was performed a median of 72 days after the initial LUS performed during hospitalization. At baseline, 87% had pathologic LUS findings in ≥1 zone (e.g. ≥3 B-lines, confluent B-lines or subpleural or lobar consolidation), whereas 30% had pathologic findings at follow-up (p < 0.001). The total number of B-lines and LUS score decreased significantly from hospitalization to follow-up (median 17 vs. 4, p < 0.001 and 4 vs. 0, p < 0.001, respectively). On the follow-up LUS, 28% of all patients had ≥3 B-lines in ≥1 zone, whereas in those with ARDS during the baseline hospitalization (n = 17), 47% had ≥3 B-lines in ≥1 zone.

Conclusion

LUS findings improved significantly from hospitalization to follow-up 2–3 months after discharge in COVID-19 survivors. However, persistent B-lines were frequent at follow-up, especially among those who initially had ARDS. LUS seems to be a promising method to monitor COVID-19 lung changes over time.

Clinicaltrials.gov ID

NCT04377035.

What Is COVID 19 Teaching Us about Pulmonary Ultrasound?

In lung ultrasound (LUS), the interactions between the acoustic pulse and the lung surface (including the pleura and a small subpleural layer of tissue) are crucial. Variations of the peripheral lung density and the subpleural alveolar shape and its configuration are typically connected to the presence of ultrasound artifacts and consolidations. COVID-19 pneumonia can give rise to a variety of pathological pulmonary changes ranging from mild diffuse alveolar damage (DAD) to severe acute respiratory distress syndrome (ARDS), characterized by peripheral bilateral patchy lung involvement. These findings are well described in CT imaging and in anatomopathological cases. Ultrasound artifacts and consolidations are therefore expected signs in COVID-19 pneumonia because edema, DAD, lung hemorrhage, interstitial thickening, hyaline membranes, and infiltrative lung diseases when they arise in a subpleural position, generate ultrasound findings. This review analyzes the structure of the ultrasound images in the normal and pathological lung given our current knowledge, and the role of LUS in the diagnosis and monitoring of patients with COVID-19 lung involvement.

Lung Ultrasound Findings in COVID-19: A Descriptive Retrospective Study

Background

Point-of-care ultrasound (POCUS) is an indispensable tool in emergency medicine. With the emergence of the coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a need for improved diagnostic capabilities and prognostic indicators for patients who are symptomatic for COVID-19 has become apparent. POCUS has been demonstrated to be a useful diagnostic and prognostic tool in the emergency department (ED) in assessing other lung complications. Still, limited data regarding its utility in assessing COVID-19 are available. This study sought to evaluate whether POCUS findings in the ED were correlated with vital signs or laboratory abnormalities typically seen among patients with COVID-19.

Methods

A retrospective study was conducted that included 39 patients who presented with COVID-19 and systemic inflammatory response syndrome (SIRS) to a large, urban tertiary care ED. The study population was limited to adults aged 18 and above who came to the ED with the primary complaint of respiratory symptoms, met SIRS criteria on admission, and had images of at least one anterior and one posterior intercostal space per lung and a minimum of four intercostal spaces. POCUS images were obtained by trained operators in the ED using portable ultrasound machines, recorded in an image database, and reviewed by ultrasound fellowship-trained emergency physicians. Clinical data (e.g., acute phase reactants and vital signs) were obtained through a chart review of patients’ electronic medical records.

Results 

Both the percentage of intercostal spaces with B-lines and the percentage of merging B-lines were correlated with decreased oxygen saturation on presentation. No other statistically significant correlations were observed between these sonographic findings and other vital signs or acute phase reactants, nor between these clinical data and the percentage of intercostal spaces that were positive for the shred sign.

Conclusions

With the emergence of the COVID-19 pandemic, emergency medicine physicians are on the frontline of identifying and caring for patients affected by the virus. This study found that sonographic findings associated with interstitial pneumonitis, notably merging B-lines, and the overall percentage of intercostal spaces with B-lines, were clearly associated with worsening oxygen saturation, now thought to be one of the driving causes of morbidity and mortality in COVID-19. As ultrasound has become a ubiquitous and indispensable tool in the ED, this study demonstrated its utility in assessing and managing patients with COVID-19. Bedside ultrasound is a cheap, fast, and non-invasive tool that healthcare providers can use as an essential adjunct in addition to laboratory markers and other imaging modalities for the diagnosis and prognosis of COVID-19.

Integrating Domain Knowledge Into Deep Networks for Lung Ultrasound With Applications to COVID-19

Lung ultrasound (LUS) is a cheap, safe and non-invasive imaging modality that can be performed at patient bed-side. However, to date LUS is not widely adopted due to lack of trained personnel required for interpreting the acquired LUS frames. In this work we propose a framework for training deep artificial neural networks for interpreting LUS, which may promote broader use of LUS. When using LUS to evaluate a patient's condition, both anatomical phenomena (e.g., the pleural line, presence of consolidations), as well as sonographic artifacts (such as A- and B-lines) are of importance. In our framework, we integrate domain knowledge into deep neural networks by inputting anatomical features and LUS artifacts in the form of additional channels containing pleural and vertical artifacts masks along with the raw LUS frames. By explicitly supplying this domain knowledge, standard off-the-shelf neural networks can be rapidly and efficiently finetuned to accomplish various tasks on LUS data, such as frame classification or semantic segmentation. Our framework allows for a unified treatment of LUS frames captured by either convex or linear probes. We evaluated our proposed framework on the task of COVID-19 severity assessment using the ICLUS dataset. In particular, we finetuned simple image classification models to predict per-frame COVID-19 severity score. We also trained a semantic segmentation model to predict per-pixel COVID-19 severity annotations. Using the combined raw LUS frames and the detected lines for both tasks, our off-the-shelf models performed better than complicated models specifically designed for these tasks, exemplifying the efficacy of our framework.

Ultrasound findings of lung ultrasonography in COVID-19: A systematic review

PURPOSE

To identify the defining lung ultrasound (LUS) findings of COVID-19, and establish its association to the initial severity of the disease and prognostic outcomes.

METHOD

Systematic review was conducted according to the PRISMA guidelines. We queried PubMed, Embase, Web of Science, Cochrane Database and Scopus using the terms ((coronavirus) OR (covid-19) OR (sars AND cov AND 2) OR (2019-nCoV)) AND (("lung ultrasound") OR (LUS)), from 31st of December 2019 to 31st of January 2021. PCR-confirmed cases of SARS-CoV-2 infection, obtained from original studies with at least 10 participants 18 years old or older, were included. Risk of bias and applicability was evaluated with QUADAS-2.

RESULTS

We found 1333 articles, from which 66 articles were included, with a pooled population of 4687 patients. The most examined findings were at least 3 B-lines, confluent B-lines, subpleural consolidation, pleural effusion and bilateral or unilateral distribution. B-lines, its confluent presentation and pleural abnormalities are the most frequent findings. LUS score was higher in intensive care unit (ICU) patients and emergency department (ED), and it was associated with a higher risk of developing unfavorable outcomes (death, ICU admission or need for mechanical ventilation). LUS findings and/or the LUS score had a good negative predictive value in the diagnosis of COVID-19 compared to RT-PCR.

CONCLUSIONS

The most frequent ultrasound findings of COVID-19 are B-lines and pleural abnormalities. High LUS score is associated with developing unfavorable outcomes. The inclusion of pleural effusion in the LUS score and the standardisation of the imaging protocol in COVID-19 LUS remains to be defined.

Executing Lung Sonography to Assess Acute and Chronic Disease: Can These Techniques Be Used to Monitor Adults and Children Surviving COVID-19?

Lung ultrasonography (LUS) has been used as a clinical diagnostic tool for the location of pleural fluid and marking patients for a thoracentesis, for decades, but has not been deemed as appropriate for other diagnostic uses. In the context of the COVID-19 pandemic, the necessity of a portable, low-cost, and non-ionizing diagnostic choice is needed to assess patient lungs. LUS has become a strong candidate to fill this diagnostic gap. With the use of Lichtenstein’s bedside LUS in emergency (BLUE) protocol, LUS may have potential to diagnose lung disease and assist with treatment decisions. While evidence of LUS as a COVID-19 diagnostic tool is not conclusive, early diagnostic results are promising. Further research on the use of LUS and the clinical implementation of the technique have a true potential to improve patient outcomes.

ECHOPAEDIA: Echography in Paediatric Patients in the Age of Coronavirus Disease 2019: Utility of Lung Ultrasound and Chest X-Ray in Diagnosis of Community-Acquired Pneumonia and Severe Acute Respiratory Syndrome Coronavirus 2 Pneumonia

BACKGROUND

In recent years, lung ultrasound (LUS) has spread to emergency departments and clinical practise gaining great support, especially in time of pandemic, but only a few studies have been done on children. The aim of the present study is to compare the diagnostic accuracy of LUS (using Soldati LUS score) and that of chest X-ray (CXR) in CAP and COVID-19 pneumonia in paediatric patients. Secondary objective of the study is to examine the association between LUS score and disease severity. Finally, we describe the local epidemiology of paediatric CAP during the study period in the era of COVID-19 by comparing it with the previous 2 years.

METHODS

This is an observational retrospective single-centre study carried out on patients aged 18 or younger and over the month of age admitted to the Paediatric Unit of our Foundation for suspected community-acquired pneumonia or SARS-CoV-2 pneumonia during the third pandemic wave of COVID-19. Quantitative variables were elaborated with Shapiro-Wilks test or median and interquartile range (IQR). Student's t-test was used for independent data. Association between quantitative data was evaluated with Pearson correlation. ROC curve analysis was used to calculate best cut-off of LUS score in paediatric patients. Area under the ROC curve (AUC), sensibility, and specificity are also reported with 95% confidence interval (CI).

RESULTS

The diagnostic accuracy of the LUS score in pneumonia, the area underlying the ROC curve (AUC) was 0.67 (95% CI: 0.27-1) thus showing a discrete discriminatory power, with a sensitivity of 89.66% and specificity 50% setting a LUS score greater than or equal to 1 as the best cut-off. Nine patients required oxygen support and a significant statistical correlation (p = 0.0033) emerged between LUS score and oxygen therapy. The mean LUS score in patients requiring oxygen therapy was 12. RCP was positively correlated to the patient's LUS score (p = 0.0024).

CONCLUSIONS

Our study has shown that LUS is a valid alternative to CXR. Our results show how LUS score can be applied effectively for the diagnosis and stratification of paediatric pneumonia.

Coronavirus Disease 2019 Phenotypes, Lung Ultrasound, Chest Computed Tomography and Clinical Features in Critically Ill Mechanically Ventilated Patients

Chest computed tomography (CT) may provide insights into the pathophysiology of coronavirus disease 2019 (COVID-19), although it is not suitable for a timely bedside dynamic assessment of patients admitted to intensive care unit (ICU); therefore, lung ultrasound (LUS) has been proposed as a complementary diagnostic tool. The aims of this study were to investigate different lungs phenotypes in patients with COVID-19 and to assess the differences in CT and LUS scores between ICU survivors and non-survivors. We also explored the association between CT and LUS, and oxygenation (arterial partial pressure of oxygen [PaO₂]/fraction of inspired oxygen [FiO_2]) and clinical parameters. The study included 39 patients with COVID-19. CT scans revealed types 1, 2 and 3 phenotypes in 62%, 28% and 10% of patients, respectively. Among survivors, pattern 1 was prevalent (p < 0.005). Chest CT and LUS scores differed between survivors and non-survivors both at ICU admission and 10 days after and were associated with ICU mortality. Chest CT score was positively correlated with LUS findings at ICU admission (r = 0.953, p < 0.0001) and was inversely correlated with PaO₂/FiO₂ (r = -0.375, p = 0.019) and C-reactive protein (r = 0.329, p = 0.041). LUS score was inversely correlated with PaO₂/FiO₂ (r = -0.345, p = 0.031). COVID-19 presents distinct phenotypes with differences between survivors and non-survivors. LUS is a valuable monitoring tool in an ICU setting because it may correlate with CT findings and mortality, although it cannot predict oxygenation changes.

Italian Society of Anesthesia, Analgesia, Resuscitation, and Intensive Care expert consensus statement on the use of lung ultrasound in critically ill patients with coronavirus disease 2019 (ITACO)

Background

To produce statements based on the available evidence and an expert consensus (as members of the Lung Ultrasound Working Group of the Italian Society of Analgesia, Anesthesia, Resuscitation, and Intensive Care, SIAARTI) on the use of lung ultrasound for the management of patients with COVID-19 admitted to the intensive care unit.

Methods

A modified Delphi method was applied by a panel of anesthesiologists and intensive care physicians expert in the use of lung ultrasound in COVID-19 intensive critically ill patients to reach a consensus on ten clinical questions concerning the role of lung ultrasound in the following: COVID-19 diagnosis and monitoring (with and without invasive mechanical ventilation), positive end expiratory pressure titration, the use of prone position, the early diagnosis of pneumothorax- or ventilator-associated pneumonia, the process of weaning from invasive mechanical ventilation, and the need for radiologic chest imaging.

Results

A total of 20 statements were produced by the panel. Agreement was reached on 18 out of 20 statements (scoring 7–9; “appropriate”) in the first round of voting, while 2 statements required a second round for agreement to be reached. At the end of the two Delphi rounds, the median score for the 20 statements was 8.5 [IQR 8.9], and the agreement percentage was 100%.

Conclusion

The Lung Ultrasound Working Group of the Italian Society of Analgesia, Anesthesia, Resuscitation, and Intensive Care produced 20 consensus statements on the use of lung ultrasound in COVID-19 patients admitted to the ICU. This expert consensus strongly suggests integrating lung ultrasound findings in the clinical management of critically ill COVID-19 patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s44158-021-00015-6.

Investigating training-test data splitting strategies for automated segmentation and scoring of COVID-19 lung ultrasound images

Ultrasound in point-of-care lung assessment is becoming increasingly relevant. This is further reinforced in the context of the COVID-19 pandemic, where rapid decisions on the lung state must be made for staging and monitoring purposes. The lung structural changes due to severe COVID-19 modify the way ultrasound propagates in the parenchyma. This is reflected by changes in the appearance of the lung ultrasound images. In abnormal lungs, vertical artifacts known as B-lines appear and can evolve into white lung patterns in the more severe cases. Currently, these artifacts are assessed by trained physicians, and the diagnosis is qualitative and operator dependent. In this article, an automatic segmentation method using a convolutional neural network is proposed to automatically stage the progression of the disease. 1863 B-mode images from 203 videos obtained from 14 asymptomatic individual,14 confirmed COVID-19 cases, and 4 suspected COVID-19 cases were used. Signs of lung damage, such as the presence and extent of B-lines and white lung areas, are manually segmented and scored from zero to three (most severe). These manually scored images are considered as ground truth. Different test-training strategies are evaluated in this study. The results shed light on the efficient approaches and common challenges associated with automatic segmentation methods.

Diagnostic Imaging in Newborns, Children and Adolescents Infected with Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2): Is There a Realistic Alternative to Lung High-Resolution Computed Tomography (HRCT) and Chest X-Rays? A Systematic Review of the Literature

Chest computed tomography has been frequently used to evaluate patients with potential coronavirus disease 2019 (COVID-19) infection. However, this may be particularly risky for pediatric patients owing to high doses of ionizing radiation. We sought to evaluate COVID-19 imaging options in pediatric patients based on the published literature. We performed an exhaustive literature review focusing on COVID-19 imaging in pediatric patients. We used the search terms "COVID-19," "SARS-CoV2," "coronavirus," "2019-nCoV," "Wuhan virus," "lung ultrasound (LUS)," "sonography," "lung HRCT," "children," "childhood" and "newborn" to query the online databases PubMed, Medical Subject Headings (MeSH), Embase, LitCovid, the World Health Organization COVID-19 database and Medline Bireme. Articles meeting the inclusion criteria were included in the analysis and review. We identified only seven studies using lung ultrasound (LUS) to diagnose severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in newborns and children. The studies evaluated small numbers of patients, and only 6% had severe or critical illness associated with COVID-19. LUS showed the presence of B-lines in 50% of patients, sub-pleural consolidation in 43.18%, pleural irregularities in 34.09%, coalescent B-lines and white lung in 25%, pleural effusion in 6.82% and thickening of the pleural line in 4.55%. We found 117 studies describing the use of chest X-ray or chest computed tomography in pediatric patients with COVID-19. The proportion of those who were severely or critically ill was similar to that in the LUS study population. Our review indicates that use of LUS should be encouraged in pediatric patients, who are at highest risk of complications from medical ionizing radiation. Increased use of LUS may be of particularly high impact in under-resourced areas, where access to chest computed tomography may be limited.

Diagnostic ultrasound imaging of the lung: A simulation approach based on propagation and reverberation in the human body

Although ultrasound cannot penetrate a tissue/air interface, it images the lung with high diagnostic accuracy. Lung ultrasound imaging relies on the interpretation of "artifacts," which arise from the complex reverberation physics occurring at the lung surface but appear deep inside the lung. This physics is more complex and less understood than conventional B-mode imaging in which the signal directly reflected by the target is used to generate an image. Here, to establish a more direct relationship between the underlying acoustics and lung imaging, simulations are used. The simulations model ultrasound propagation and reverberation in the human abdomen and at the tissue/air interfaces of the lung in a way that allows for direct measurements of acoustic pressure inside the human body and various anatomical structures, something that is not feasible clinically or experimentally. It is shown that the B-mode images beamformed from these acoustical simulations reproduce primary clinical features that are used in diagnostic lung imaging, i.e., A-lines and B-lines, with a clear relationship to known underlying anatomical structures. Both the oblique and parasagittal views are successfully modeled with the latter producing the characteristic "bat sign," arising from the ribs and intercostal part of the pleura. These simulations also establish a quantitative link between the percentage of fluid in exudative regions and the appearance of B-lines, suggesting that the B-mode may be used as a quantitative imaging modality.

Prediction Accuracy of Serial Lung Ultrasound in COVID-19 Hospitalized Patients (Pred-Echovid Study)

The value of serial lung ultrasound (LUS) in patients with COVID-19 is not well defined. In this multicenter prospective observational study, we aimed to assess the prognostic accuracy of serial LUS in patients admitted to hospital due to COVID-19. The serial LUS protocol included two examinations (0–48 h and 72–96 h after admission) using a 10-zones sequence, and a 0 to 5 severity score. Primary combined endpoint was death or the need for invasive mechanical ventilation. Calibration (Hosmer–Lemeshow test and calibration curves), and discrimination power (area under the ROC curve) of both ultrasound exams (SCORE1 and 2), and their difference (DIFFERENTIAL-SCORE) were performed. A total of 469 patients (54.2% women, median age 60 years) were included. The primary endpoint occurred in 51 patients (10.9%). Probability risk tertiles of SCORE1 and SCORE2 (0–11 points, 12–24 points, and ≥25 points) obtained a high calibration. SCORE-2 showed a higher discrimination power than SCORE-1 (AUC 0.72 (0.58–0.85) vs. 0.61 (0.52–0.7)). The DIFFERENTIAL-SCORE showed a higher discrimination power than SCORE-1 and SCORE-2 (AUC 0.78 (0.66–0.9)). An algorithm for clinical decision-making is proposed. Serial lung ultrasound performing two examinations during the first days of hospitalization is an accurate strategy for predicting clinical deterioration of patients with COVID-19.

Lung Ultrasound Can Predict the Clinical Course and Severity of COVID-19 Disease

Coronavirus disease 2019 (COVID-19) compromises the lung in large numbers of people. The development of minimally invasive methods to determine the severity of pulmonary extension is desired. This study aimed to describe the characteristics of sequential lung ultrasound and to test the prognostic usefulness of this exam in a group of patients admitted to the hospital with COVID-19. We prospectively evaluated patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection admitted to our hospital between April and August 2020. Bedside lung ultrasound exams were performed at three time points: at inclusion in the study, after 48 h and on the seventh day of follow-up. Lung ultrasound scores were quantified according to the aeration loss in each of eight zones scanned. Sixty-six participants were included: 42 (63.6%) in the intensive care unit and 24 (36.3%) in the ward. Lung ultrasound scores were higher in participants admitted to the intensive care unit than in those admitted to the ward at the time of inclusion (16 [13-17] vs. 10 [4-14], p < 0.001), after 48 h (15.5 [13-17] vs. 12.5 [8.2-14.7], p = 0.001) and on the seventh day (16 [14-17] vs. 7 [4.5-13.7], p < 0.001) respectively. Lung ultrasound score measured at the time of inclusion in the study was independently associated with the need for admission to the intensive care unit (odds ratio = 1.480; 95% confidence interval, 1.093-2.004; p = 0.011) adjusted by the Sequential Organ Failure Assessment score.

Comparison of two different lung ultrasound imaging protocols in COVID-19 pneumonia

Purpose

The aim of this study was to determine the effectiveness of two different lung ultrasonography (LUS) methods that can be used in the diagnosis of coronavirus disease 2019 (COVID-19) and to investigate their correlations with computed tomography (CT).

Methods

In this prospective, randomized, and single-blind study, 60 patients with COVID-19 were included. The patients were randomized to either the 12-zone LUS group (n=30) or the 14-zone LUS group (n=30). The correlation between LUS and thoracic CT scores was evaluated. As a secondary outcome measure, the characteristic features of the findings of thoracic CT and LUS were examined.

Results

The study was completed with a total of 59 patients. Moderate and high correlations were found between the total CT and LUS scores in the 12-zone and 14-zone study groups. There were no statistically significant differences in the lesion types detected in patients using LUS and CT (P>0.05). The left lung lower lobe CT scores were statistically significantly lower in the 14-zone study group than in the 12-zone group (P=0.019). The left lower lobe CT and LUS scores were highly correlated in the 14-zone group (P<0.001, r=0.902).

Conclusion

The results of our study indicated that the two different LUS examination methods performed in different patients had similar findings in terms of the diagnosis and their correlations with CT results.

Quantitative Analysis and Automated Lung Ultrasound Scoring for Evaluating COVID-19 Pneumonia With Neural Networks

As being radiation-free, portable, and capable of repetitive use, ultrasonography is playing an important role in diagnosing and evaluating the COVID-19 Pneumonia (PN) in this epidemic. By virtue of lung ultrasound scores (LUSS), lung ultrasound (LUS) was used to estimate the excessive lung fluid that is an important clinical manifestation of COVID-19 PN, with high sensitivity and specificity. However, as a qualitative method, LUSS suffered from large interobserver variations and requirement for experienced clinicians. Considering this limitation, we developed a quantitative and automatic lung ultrasound scoring system for evaluating the COVID-19 PN. A total of 1527 ultrasound images prospectively collected from 31 COVID-19 PN patients with different clinical conditions were evaluated and scored with LUSS by experienced clinicians. All images were processed via a series of computer-aided analysis, including curve-to-linear conversion, pleural line detection, region-of-interest (ROI) selection, and feature extraction. A collection of 28 features extracted from the ROI was specifically defined for mimicking the LUSS. Multilayer fully connected neural networks, support vector machines, and decision trees were developed for scoring LUS images using the fivefold cross validation. The model with 128×256 two fully connected layers gave the best accuracy of 87%. It is concluded that the proposed method could assess the ultrasound images by assigning LUSS automatically with high accuracy, potentially applicable to the clinics.

Association of Lung Ultrasound Score with Mortality and Severity of COVID-19: A Meta-Analysis and Trial Sequential Analysis

OBJECTIVES

The coronavirus disease 2019 (COVID-19) pandemic has rapidly spread all over the world. Lung ultrasound (LUS) has emerged as a useful tool for diagnosing many respiratory diseases. The prognostic role of LUS in COVID-19 patients has not yet been established.

METHODS

Several databases were searched on 09 April 2021. The difference in LUS score between the death and survival groups, and the relationship between LUS score and COVID-19 severity were both assessed.

RESULTS

The LUS score was significantly higher in the death group compared with the survival group (weighted mean difference (WMD) = 8.21, 95% CI: 4.74-11.67, P < 0.001), which was confirmed by trial sequential analysis. Those with mild/moderate, severe and critical COVID-19 had a progressively higher LUS score (critical vs. severe: WMD = 8.78, 95% CI: 4.17-13.38; P < 0.001; critical vs. mild/moderate/severe: WMD = 10.00, 95% CI: 6.83-13.17, P < 0.001; severe vs. moderate: WMD = 5.96, 95% CI: 3.48-8.44, P < 0.001; severe vs. mild/moderate: WMD = 7.31, 95% CI: 4.45-10.17, P < 0.001).

CONCLUSIONS

The LUS score was associated with mortality and severity of COVID-19. The LUS score might be a risk stratification tool for COVID-19 patients.

Is lung ultrasound score a useful tool to monitoring and handling moderate and severe COVID-19 patients in the general ward? An observational pilot study

Lung ultrasound is a well-established diagnostic tool in acute respiratory failure, and it has been shown to be particularly suited for the management of COVID-19-associated respiratory failure. We present exploratory analyses on the diagnostic and prognostic performance of lung ultrasound score (LUS) in general ward patients with moderate-to-severe COVID-19 pneumonia receiving O₂ supplementation and/or noninvasive ventilation. From March 10 through May 1, 2020, 103 lung ultrasound exams were performed by our Forward Intensive Care Team (FICT) on 26 patients (18 males and 8 females), aged 62 (54 - 76) and with a Body Mass Index (BMI) of 30.9 (28.7 - 31.5), a median 6 (5 - 9) days after admission to the COVID-19 medical unit of the University Hospital of Parma, Italy. All patients underwent chest computed tomography (CT) the day of admission. The initial LUS was 16 (11 - 21), which did not significantly correlate with initial CT scans, probably due to rapid progression of the disease and time between CT scan on admission and first FICT evaluation; conversely, LUS was significantly correlated with PaO₂/FiO₂ ratio throughout patient follow-up [R = - 4.82 (- 6.84 to - 2.80; p < 0.001)]. The area under the receiving operating characteristics curve of LUS for the diagnosis of moderate-severe disease (PaO₂/FiO₂ ratio ≤ 200 mmHg) was 0.73, with an optimal cutoff value of 11 (positive predictive value: 0.98; negative predictive value: 0.29). Patients who eventually needed invasive ventilation and/or died during admission had significantly higher LUS throughout their stay.

Deep learning applied to lung ultrasound videos for scoring COVID-19 patients: A multicenter study

In the current pandemic, lung ultrasound (LUS) played a useful role in evaluating patients affected by COVID-19. However, LUS remains limited to the visual inspection of ultrasound data, thus negatively affecting the reliability and reproducibility of the findings. Moreover, many different imaging protocols have been proposed, most of which lacked proper clinical validation. To address these problems, we were the first to propose a standardized imaging protocol and scoring system. Next, we developed the first deep learning (DL) algorithms capable of evaluating LUS videos providing, for each video-frame, the score as well as semantic segmentation. Moreover, we have analyzed the impact of different imaging protocols and demonstrated the prognostic value of our approach. In this work, we report on the level of agreement between the DL and LUS experts, when evaluating LUS data. The results show a percentage of agreement between DL and LUS experts of 85.96% in the stratification between patients at high risk of clinical worsening and patients at low risk. These encouraging results demonstrate the potential of DL models for the automatic scoring of LUS data, when applied to high quality data acquired accordingly to a standardized imaging protocol.