Radiographic assessment of lung edema (RALE) score is associated with clinical outcomes in patients with refractory cardiogenic shock and refractory cardiac arrest after percutaneous implantation of extracorporeal life support

Predictors of successful weaning from veno-arterial extracorporeal membrane oxygenation (V-A ECMO): A systematic review and meta-analysis

BACKGROUND

Venoarterial extracorporeal membrane oxygenation (V-A ECMO) use to support patients in cardiac failure is increasing. Despite this increased use, predicting successful weaning from ECMO can be challenging, no uniform guidelines on weaning exist. Therefore, we completed a systematic review to evaluate prognostic factors that predict successful weaning from V-A ECMO.

METHODS

Following the PRIMSA guidelines, a systematic literature search of Medline, Embase, SCOPUS and CENTRAL identified original research studies of patients requiring V-A ECMO where weaning was attempted. Data was collected on demographic factors and weaning protocol, biomarkers, haemodynamic, echocardiographic factors for the successfully weaned (SW) and not successfully weaned (NSW) groups. Two investigators reviewed studies for relevance, extracted data, and assessed risk of bias using the ROBINS-I tool. The study was registered on the international prospective register of systematic reviews (PROSPERO ID# CRD42022366153).

RESULTS

1219 records were screened, of which 20 studies were deemed sufficient to be included in the statistical analysis based on pre-specified criteria. Factors associated with successful weaning were higher left ventricular ejection fraction (LVEF) (MD 9.0, 95% CI 4.1-13.8; p < 0.001) and left ventricular outflow tract velocity time integral (LVOT VTI) at time of weaning, (MD 1.35, 95% CI 0.28-2.40 lactate at admission (MD -3.2, 95% CI -4.8 to -1.5, p < 0.001), and CK-MB at admission (MD -4.11, 95%CI -6.6 to -1.6, p = 0.001). Critical appraisal demonstrated moderate-high risk of bias owing to confounding and low sample sizes.

CONCLUSION

In patients on V-A ECMO support being assessed for weaning multi-parametric assessment is required. Moderate-high heterogeneity and low sample sizes warrant higher-quality studies to help guide decisions to wean patients from V-A ECMO.

Chest Radiography Scores for Predicting the Severity of Respiratory Status in Newborns Weighing More Than 1,500 g

Background Acute respiratory failure (ARF) may occur in neonates. Chest radiography is commonly used to evaluate the severity of ARF; however, the application of quantitative scales in clinical practice in neonatal intensive care units is uncommon. This study aimed to assess the usefulness of two semi-quantitative radiographical scales, the Brixia and radiographic assessment of lung edema (RALE) scores, in newborns weighing more than 1,500 g. Methods Newborns weighing > 1,500 g who received invasive respiratory support with arterial lines between January 2010 and October 2023 were enrolled in this study (n = 98; gestational age, 35.6 ± 3.1 weeks; birthweight, 2,321 ± 600 g). We investigated the correlation between the Brixia or RALE scores and the oxygen index (OI), alveolar-arterial oxygen gradient (A-aDO2), and ventilation index (VI). Furthermore, the cut-off points of the two radiographic scores for the prediction of these respiratory indices were determined. Results All respiratory indices correlated with the Brixia (OI: r = 0.71, p < 0.001; A-aDO2: r = 0.74, p < 0.001; VI: r = 0.56, p < 0.001) and RALE scores (OI: r = 0.78, p < 0.001; A-aDO2: r = 0.82, p < 0.001; VI: r = 0.60, p < 0.001). Additionally, the receiver operating characteristic curve showed that the radiographical scores had a statistically significant ability to predict respiratory index values. Conclusion Brixia and RALE scores are useful predictive markers of acute respiratory failure in infants weighing >1,500 g. These chest radiography scores may be good predictors of respiratory status and have wider clinical applications in neonatal care.

Automated quantification of SARS-CoV-2 pneumonia with large vision model knowledge adaptation

Background

Large vision models (LVM) pretrained by large datasets have demonstrated their enormous capacity to understand visual patterns and capture semantic information from images. We proposed a novel method of knowledge domain adaptation with pretrained LVM for a low-cost artificial intelligence (AI) model to quantify the severity of SARS-CoV-2 pneumonia based on frontal chest X-ray (CXR) images.

Methods

Our method used the pretrained LVMs as the primary feature extractor and self-supervised contrastive learning for domain adaptation. An encoder with a 2048-dimensional feature vector output was first trained by self-supervised learning for knowledge domain adaptation. Then a multi-layer perceptron (MLP) was trained for the final severity prediction. A dataset with 2599 CXR images was used for model training and evaluation.

Results

The model based on the pretrained vision transformer (ViT) and self-supervised learning achieved the best performance in cross validation, with mean squared error (MSE) of 23.83 (95 % CI 22.67-25.00) and mean absolute error (MAE) of 3.64 (95 % CI 3.54-3.73). Its prediction correlation has theR 2 of 0.81 (95 % CI 0.79-0.82) and Spearman ρ of 0.80 (95 % CI 0.77-0.81), which are comparable to the current state-of-the-art (SOTA) methods trained by much larger CXR datasets.

Conclusion

The proposed new method has achieved the SOTA performance to quantify the severity of SARS-CoV-2 pneumonia at a significantly lower cost. The method can be extended to other infectious disease detection or quantification to expedite the application of AI in medical research.

Factors Influencing Successful Weaning From Venoarterial Extracorporeal Membrane Oxygenation: A Systematic Review

With advancements in extracorporeal life support (ECLS) technologies, venoarterial extracorporeal membrane oxygenation (VA-ECMO) has emerged as a crucial cardiopulmonary support mechanism. This review explores the significance of VA-ECMO system configuration, cannulation strategies, and timing of initiation. Through an analysis of medication management strategies, complication management, and comprehensive preweaning assessments, it aims to establish a multidimensional evaluation framework to assist clinicians in making informed decisions regarding weaning from VA-ECMO, thereby ensuring the safe and effective transition of patients.

Study on the Predictive Value of a Pulmonary Edema Imaging Score for Delayed Extubation in Patients after Heart Valve Surgery on Cardiopulmonary Bypass

Background

Delayed extubation with mechanical ventilation after cardiac valve surgery is an important clinical challenge. Early extubation can improve the survival rate and prognosis of patients. The study aims to explore the predictive value of a chest X-ray pulmonary edema imaging score on the first day after surgery for delayed extubation in patients after cardiac valve surgery on cardiopulmonary bypass.

Method

Retrospective analysis of the clinical data of patients undergoing cardiac valve surgery under extracorporeal circulation admitted to the intensive care unit of Zhongshan Hospital Affiliated with Fudan University (Xiamen) from January 2020 to October 2023. The patients were divided into an early extubation group according to the postoperative mechanical ventilation time (time: <24 h) and a delayed extubation group (time: ≥24 h). The radiographic assessment of lung edema (RALE) score was performed on the chest X-ray of the patient on the first day after surgery to analyze the correlation between delayed extubation of mechanical ventilation and the chest radiograph RALE score on the first day after surgery and to verify its predictive performance.

Results

Significant differences in age, the incidence of hypertension, body mass index (BMI), left ventricular ejection fraction (LVEF), pump time, RALE score, ventilation time, oxygenation index, PaCO2, and brain natriuretic peptide (BNP) levels after the first 24 h were seen between patients who were extubated before and 24 h post operation (p = 0.013, 0.001, 0.034, <0.001, <0.001, <0.001, <0.001, <0.001, 0.014, and <0.001, respectively). No significant differences were observed in the proportion of males and the lactate level after the first 24 h between the two groups (p = 0.792 and 0.191, respectively). The time of mechanical ventilation was positively correlated with the RALE score in all patients, and the correlation coefficient was 0.419; the difference was statistically significant (p < 0.001). Multivariate binary logistic regression analysis with stepwise regression was performed on each research factor, and it was found that RALE score, pump time, oxygenation index, and postoperative BNP were independent risk factors for predicting delayed extubation in patients undergoing cardiac surgery on cardiopulmonary bypass. A 10-fold cross-validation revealed that the mean accuracy, sensitivity, specificity, and area under the curve (AUC) of the regression model were 0.737, 0.749, 0.741, and 0.825, respectively.

Conclusions

The RALE score on the chest radiograph on the first day after surgery is an independent risk factor for predicting delayed extubation in patients after cardiac valve surgery on cardiopulmonary bypass and has good predictive value.

Radiographic scores as a predictor of oxygenation index in very low-birthweight infants

BACKGROUND

Very low birthweight infants (VLBWIs) often undergo chest radiographic examinations without standardization or objectivity. This study aimed to assess the association of two radiographic scores, the Brixia and radiographic assessment of lung edema (RALE), with oxygenation index (OI) in ventilated VLBWIs and to determine the optimal cutoff values to predict hypoxic respiratory severity.

METHODS

VLBWIs who received invasive respiratory support with arterial lines between January 2010 and October 2023 were enrolled in this study (n = 144). The correlation between the Brixia or RALE scores and OI was investigated. Receiver operating characteristic curve analysis was performed to determine the optimal cutoff points of the two radiographic scores for predicting OI values (OI ≥5, ≥10, and ≥15).

RESULTS

The enrolled infants had a median gestational age of 27 weeks (interquartile range [IQR], 25-28 weeks) and a median birthweight of 855 g (IQR, 684-1003 g). Radiographic scoring methods correlated with the OI (Brixia score: r = 0.79, p < 0.001; RALE score: r = 0.72, p < 0.001). The optimal cutoff points for predicting OI values were as follows: Brixia score: OI ≥5, 10; OI ≥10, 13; OI ≥15, 15; RALE score: OI ≥5, 22; OI ≥10, 31; and OI ≥15, 40.

CONCLUSIONS

Brixia and RALE scores are useful predictive markers of the oxygenation status in intubated VLBWIs with stable hemodynamics. These scores are easy to use and promising tools for clinicians to identify patients with a higher risk of hypoxic respiratory failure.

Association of Late Radiographic Assessment of Lung Edema Score with Clinical Outcome in Patients with Influenza-Associated Acute Respiratory Distress Syndrome

Background: Influenza virus infection leads to acute pulmonary injury and acute respiratory distress syndrome (ARDS). The Radiographic Assessment of Lung Edema (RALE) score has been proposed as a reliable tool for the evaluation of the opacity of chest X-rays (CXRs). This study aimed to examine the RALE scores and outcomes in patients with influenza-associated ARDS. Methods: Patients who were newly diagnosed with influenza-associated ARDS from December 2015 to March 2016 were enrolled. Two independent reviewers scored the CXRs obtained on the day of ICU admission and on days 2 and 7 after intensive care unit (ICU) admission. Results: During the study, 47 patients had influenza-associated ARDS. Five died within 7 days of ICU admission. Of the remaining 42, non-survivors (N = 12) had higher Sequential Organ Failure Assessment scores (SOFA) at ICU admission and higher day 7 RALE scores than survivors (N = 30). The day 7 RALE score independently related to late in-hospital mortality (aOR = 1.121, 95% CI: 1.014-1.240, p = 0.025). Conclusions: The RALE score for the evaluation of opacity on CXRs is a highly reproducible tool. Moreover, RALE score on day 7 was an independent predictor of late in-hospital mortality in patients with influenza-associated ARDS.

Automatic Quantification of COVID-19 Pulmonary Edema by Self-supervised Contrastive Learning

We proposed a self-supervised machine learning method to automatically rate the severity of pulmonary edema in the frontal chest X-ray radiographs (CXR) which could be potentially related to COVID-19 viral pneumonia. For this we use the modified radiographic assessment of lung edema (mRALE) scoring system. The new model was first optimized with the simple Siamese network (SimSiam) architecture where a ResNet-50 pretrained by ImageNet database was used as the backbone. The encoder projected a 2048-dimension embedding as representation features to a downstream fully connected deep neural network for mRALE score prediction. A 5-fold cross-validation with 2,599 frontal CXRs was used to examine the new model's performance with comparison to a non-pretrained SimSiam encoder and a ResNet-50 trained from scratch. The mean absolute error (MAE) of the new model is 5.05 (95%CI 5.03-5.08), the mean squared error (MSE) is 66.67 (95%CI 66.29-67.06), and the Spearman's correlation coefficient (Spearman ρ) to the expert-annotated scores is 0.77 (95%CI 0.75-0.79). All the performance metrics of the new model are superior to the two comparators (P<0.01), and the scores of MSE and Spearman ρ of the two comparators have no statistical difference (P>0.05). The model also achieved a prediction probability concordance of 0.811 and a quadratic weighted kappa of 0.739 with the medical expert annotations in external validation. We conclude that the self-supervised contrastive learning method is an effective strategy for mRALE automated scoring. It provides a new approach to improve machine learning performance and minimize the expert knowledge involvement in quantitative medical image pattern learning.

Relationship between hyponatremia at hospital admission and cardiopulmonary profile at follow-up in patients with SARS-CoV-2 (COVID-19) infection

PURPOSE

Hyponatremia occurs in about 30% of patients with pneumonia, including those with SARS-CoV-2 (COVID-19) infection. Hyponatremia predicts a worse outcome in several pathologic conditions and in COVID-19 has been associated with a higher risk of non-invasive ventilation, ICU transfer and death. The main objective of this study was to determine whether early hyponatremia is also a predictor of long-term sequelae at follow-up.

METHODS

In this observational study, we collected 6-month follow-up data from 189 laboratory-confirmed COVID-19 patients previously admitted to a University Hospital. About 25% of the patients (n = 47) had hyponatremia at the time of hospital admission.

RESULTS

Serum [Na⁺] was significantly increased in the whole group of 189 patients at 6 months, compared to the value at hospital admission (141.4 ± 2.2 vs 137 ± 3.5 mEq/L, p < 0.001). In addition, IL-6 levels decreased and the PaO₂/FiO₂ increased. Accordingly, pulmonary involvement, evaluated at the chest X-ray by the RALE score, decreased. However, in patients with hyponatremia at hospital admission, higher levels of LDH, fibrinogen, troponin T and NT-ProBNP were detected at follow-up, compared to patients with normonatremia at admission. In addition, hyponatremia at admission was associated with worse echocardiography parameters related to right ventricular function, together with a higher RALE score.

CONCLUSION

These results suggest that early hyponatremia in COVID-19 patients is associated with the presence of laboratory and imaging parameters indicating a greater pulmonary and right-sided heart involvement at follow-up.

A radiographic, deep transfer learning framework, adapted to estimate lung opacities from chest x-rays

Chest radiographs (CXRs) are the most widely available radiographic imaging modality used to detect respiratory diseases that result in lung opacities. CXR reports often use non-standardized language that result in subjective, qualitative, and non-reproducible opacity estimates. Our goal was to develop a robust deep transfer learning framework and adapt it to estimate the degree of lung opacity from CXRs. Following CXR data selection based on exclusion criteria, segmentation schemes were used for ROI (Region Of Interest) extraction, and all combinations of segmentation, data balancing, and classification methods were tested to pick the top performing models. Multifold cross validation was used to determine the best model from the initial selected top models, based on appropriate performance metrics, as well as a novel Macro-Averaged Heatmap Concordance Score (MA HCS). Performance of the best model is compared against that of expert physician annotators, and heatmaps were produced. Finally, model performance sensitivity analysis across patient populations of interest was performed. The proposed framework was adapted to the specific use case of estimation of degree of CXR lung opacity using ordinal multiclass classification. Acquired between March 24, 2020, and May 22, 2020, 38,365 prospectively annotated CXRs from 17,418 patients were used. We tested three neural network architectures (ResNet-50, VGG-16, and ChexNet), three segmentation schemes (no segmentation, lung segmentation, and lateral segmentation based on spine detection), and three data balancing strategies (undersampling, double-stage sampling, and synthetic minority oversampling) using 38,079 CXR images for training, and validation with 286 images as the out-of-the-box dataset that underwent expert radiologist adjudication. Based on the results of these experiments, the ResNet-50 model with undersampling and no ROI segmentation is recommended for lung opacity classification, based on optimal values for the MAE metric and HCS (Heatmap Concordance Score). The degree of agreement between the opacity scores predicted by this model with respect to the two sets of radiologist scores (OR or Original Reader and OOBTR or Out Of Box Reader) in terms of performance metrics is superior to the inter-radiologist opacity score agreement.