The use of lung ultrasound in evaluation of extravascular lung water in hemodialysis patients: Systematic review and meta-analysis.

RATIONALE AND OBJECTIVES: Determining dry weight is crucial for optimizing hemodialysis, influencing efficacy, cardiovascular outcomes, and overall survival. Traditional clinical assessment methods for dry weight, relying on factors such as blood pressure and edema, frequently lack reliability. Lung ultrasound stands out as a promising tool for assessing volume status, given its non-invasiveness and reproducibility. This study aims to explore the role of Lung ultrasound in evaluating the impact of hemodialysis and ultrafiltration on extravascular lung water, with a specific focus on changes in B-lines post-hemodialysis compared to pre-hemodialysis. MATERIALS AND METHODS: The research encompassed searches across PubMed, WOS, and Scopus databases for studies related to lung ultrasound and hemodialysis. A meta-analysis was then performed to determine the mean differences in various parameters before compared to after, hemodialysis, including the number of B-lines, indexed end-inspiratory and end-expiratory inferior vena cava diameters, inferior vena cava collapsibility index, weight, blood pressure, and serum levels of NT-pro-BNP. RESULTS: Our meta-analysis, included 33 studies with 2301 hemodialysis patients, revealed a significant decrease in the number of B-lines post-hemodialysis (mean difference = 8.30, 95% CI [3.55 to 13.05]). Furthermore, there was a noteworthy reduction in inspiratory and expiratory inferior vena cava diameters post-hemodialysis (mean difference = 2.32, 95% CI [0.31 to 4.33]; mean difference = 4.05, 95% CI [2.44 to 5.65], respectively). Additionally, a significant positive correlation was observed between B-lines and the maximum inferior vena cava diameter both pre- and post-hemodialysis (correlation coefficient = 0.39; correlation coefficient = 0.32, respectively). CONCLUSION: These findings indicate the effectiveness of lung ultrasound in detection of volume overload and assessment of response to ultrafiltration in hemodialysis patients.

Evaluation of extravascular lung water and cardiac function in normal vaginal delivery by intrapartum bedside ultrasound.

BACKGROUND: Healthy parturients may experience pulmonary edema and disturbed cardiac function during labor. We aimed to evaluate the extravascular lung water (EVLW), intravascular volume, and cardiac function of normal parturients during spontaneous vaginal delivery by bedside ultrasound. And to explore the correlation between EVLW and intravascular volume, cardiac function. METHODS: This was a prospective observational study including 30 singleton-term pregnant women undergoing spontaneous vaginal delivery. Bedside ultrasound was performed at the early labor, the end of the second stage of labor, 2 and 24 h postpartum, and 120 scanning results were recorded. EVLW was evaluated by the echo comet score (ECS) obtained by the 28-rib interspaces technique. Inferior vena cava collapsibility index (IVC-CI), left ventricle ejection fraction, right ventricle fractional area change, left and right ventricular E/A ratio, and left and right ventricular index of myocardial performance (LIMP and RIMP) were measured. Measurements among different time points were compared, and the correlations between ECS and other measurements were analyzed. RESULTS: During the spontaneous vaginal delivery of healthy pregnant women, 2 had a mild EVLW increase at the early labor, 8 at the end of the second stage of labor, 13 at 2 h postpartum, and 4 at 24 h postpartum (P < 0.001). From the early labor to 24 h postpartum, ECS first increased and then decreased, reaching its peak at 2 h postpartum (P < 0.001). IVC-CI first decreased and then increased, reaching its minimum at the end of the second stage of labor (P < 0.001). RIMP exceeded the cut-off value of 0.43 at the end of the second stage of labor. ECS was weakly correlated with IVC-CI (r=-0.373, P < 0.001), LIMP (r = 0.298, P = 0.022) and RIMP (r = 0.211, P = 0.021). CONCLUSIONS: During spontaneous vaginal delivery, the most vital period of perinatal care is between the end of the second stage of labor and 2 h postpartum, because the risk of pulmonary edema is higher and the right ventricle function may decline. IVC-CI can be used to evaluate maternal intravascular volume. The increase in EVLW may be related to the increase in intravascular volume and the decrease in ventricular function.

Cardiac arrest related lung edema: examining the role of downtimes in transpulmonary thermodilution analysis.

Pulmonary edema and its association with low flow times has been observed in postcardiac arrest patients. However, diagnosis of distinct types of lung pathology is difficult.The aim of this study was to investigate pulmonary edema by transpulmonary thermodilution (TPTD) after out-of-hospital cardiac arrest (OHCA), and the correlation to downtimes. In this retrospective single-center study consecutive patients with return of spontaneous circulation (ROSC) following OHCA, age ≥ 18, and applied TPTD were enrolled. According to downtimes, patients were divided into a short and a long no-flow-time group, and data of TPTD were analysed. We identified 45 patients (n = 25 short no-flow time; n = 20 long no-flow time) who met the inclusion criteria. 24 h after ROSC, the extra vascular lung water index (EVLWI) was found to be lower in the group with short no-flow time compared to the group with long no-flow time (10.7 ± 3.5 ml/kg vs. 12.8 ± 3.9 ml/kg; p = 0.08) and remained at a similar level 48 h (10.9 ± 4.3 ml/kg vs. 12.9 ± 4.9 ml/kg; p = 0.25) and 72 h (11.1 ± 5.0 ml/kg vs. 13.9 ± 7.7 ml/kg; p = 0.27) post-ROSC. We found a statistically significant and moderate correlation between no-flow duration and EVLWI 48 h (r = 0.51; p = 0.002) and 72 h (r = 0.54; p = 0.004) post-ROSC. Pulmonary vascular permeability index (PVPI) was not correlated with downtimes. Our observation underlines the presence of cardiac arrest-related lung edema by determination of EVLWI. The duration of no-flow times is a relevant factor for increased extravascular lung water index.

Thrombin worsens extravascular lung water and outcomes of septic patients with acute respiratory distress syndrome: A case control study.

Endothelial cell (EC) activation may increase systemic vascular permeability, causing extravascular lung water (EVLW) in sepsis with acute respiratory distress syndrome (ARDS). However, the correlation between thrombin and EVLW in sepsis and ARDS has not yet been addressed. Patients with sepsis and ARDS were prospectively enrolled between 2014 and 2016, and EVLW and serum thrombin levels on days 1 and 3 were measured and compared between surviving and non-surviving patients. Additionally, morphological changes in human umbilical vein endothelial cells (HUVECs) in the serum of patients with high and low EVLW were evaluated. The levels of EVLW, endothelial cells, and thrombin may positively correlate with the survival of patients with severe sepsis and ARDS. Twenty-seven patients were enrolled, and baseline characteristics, including age, sex, Acute Physiology and Chronic Health Evaluation (APACHE) II, prior 24-h fluid balance, body mass index, and shock status, were similar between survivors and non-survivors; however, day 1 EVLW was higher in non-survivors (27.5 ±â€…8.4 vs 22 ±â€…6.5 mL/kg, P = .047). EVLW of survivors improved from day 1 to day 3 (22 ±â€…6.5 vs 11 ±â€…3.8 mL/kg, P < .001), but did not improve in non-survivors (27.5 ±â€…8.4 vs 28 ±â€…6.7 mL/kg, P = .086), which means that patients had significantly lower EVLW on day 3 than on day 1. Thrombin levels of survivors significantly improved (1.03 ±â€…0.55 vs 0.87 ±â€…0.25 U/mL, P = .04) but did not improve in non-survivors (1.97 ±â€…0.75 vs 2.2 ±â€…0.75 U/mL, P = .08) from day 1 to day 3. EVLW and thrombin levels were positively correlated (r2 = 0.71, P < .0001). In vitro, the morphology and junctions of HUVECs changed when the serum from patients with high EVLW was added. The intercellular distances among the control, high EVLW, and low EVLW groups were 5.25 ± 1.22, 21.33 ± 2.15, and 11.17 ± 1.64 µm, respectively (P < .05).

[Evaluation of extravascular lung water index in critically ill patients based on lung ultrasound radiomics analysis combined with machine learning].

OBJECTIVE: To explore lung ultrasound radiomics features which related to extravascular lung water index (EVLWI), and to predict EVLWI in critically ill patients based on lung ultrasound radiomics combined with machine learning and validate its effectiveness. METHODS: A retrospective case-control study was conducted. The lung ultrasound videos and pulse indicated continuous cardiac output (PiCCO) monitoring results of critically ill patients admitted to the department of critical care medicine of the First Affiliated Hospital of Guangxi Medical University from November 2021 to October 2022 were collected, and randomly divided into training set and validation set at 8:2. The corresponding images from lung ultrasound videos were obtained to extract radiomics features. The EVLWI measured by PiCCO was regarded as the "gold standard", and the radiomics features of training set was filtered through statistical analysis and LASSO algorithm. Eight machine learning models were trained using filtered radiomics features including random forest (RF), extreme gradient boost (XGBoost), decision tree (DT), Naive Bayes (NB), multi-layer perceptron (MLP), K-nearest neighbor (KNN), support vector machine (SVM), and Logistic regression (LR). Receiver operator characteristic curve (ROC curve) was plotted to evaluate the predictive performance of models on EVLWI in the validation set. RESULTS: A total of 151 samples from 30 patients were enrolled (including 906 lung ultrasound videos and 151 PiCCO monitoring results), 120 in the training set, and 31 in the validation set. There were no statistically significant differences in main baseline data including gender, age, body mass index (BMI), mean arterial pressure (MAP), central venous pressure (CVP), heart rate (HR), cardiac index (CI), cardiac function index (CFI), stroke volume index (SVI), global end diastolic volume index (GEDVI), systemic vascular resistance index (SVRI), pulmonary vascular permeability index (PVPI) and EVLWI. The overall EVLWI range in 151 PiCCO monitoring results was 3.7-25.6 mL/kg. Layered analysis showed that both datasets had EVLWI in the 7-15 mL/kg interval, and there was no statistically significant difference in EVLWI distribution. Two radiomics features were selected by using LASSO algorithm, namely grayscale non-uniformity (weight was -0.006 464) and complexity (weight was -0.167 583), and they were used for modeling. ROC curve analysis showed that the MLP model had better predictive performance. The area under the ROC curve (AUC) of the prediction validation set EVLWI was higher than that of RF, XGBoost, DT, KNN, LR, SVM, NB models (0.682 vs. 0.658, 0.657, 0.614, 0.608, 0.596, 0.557, 0.472). CONCLUSIONS: The gray level non-uniformity and complexity of lung ultrasound were the most correlated radiomics features with EVLWI monitored by PiCCO. The MLP model based on gray level non-uniformity and complexity of lung ultrasound can be used for semi-quantitative prediction of EVLWI in critically ill patients.

Lung "Comet Tails" in Healthy Individuals: Accumulation or Clearance of Extravascular Lung Water?

Parks, Jordan K, Courtney M. Wheatley-Guy, Glenn M. Stewart, Caitlin C. Fermoyle, Bryan J. Taylor, Jesse Schwartz, Briana Ziegler, Kay Johnson, Alice Gavet, Loïc Chabridon, Paul Robach, and Bruce D. Johnson. Lung "Comet Tails" in healthy individuals: accumulation or clearance of extravascular lung water? High Alt Med Biol. 24:230-233, 2023-Ultrasound lung comet tails (or B-lines) tend to be limited in number (<5) or absent under ultrasound examination, and the appearance of diffuse B-lines with lung sliding has been suggested to identify pulmonary edema. Clinical evaluation of B-lines has been utilized as a bedside test to assess pulmonary congestion in patients with heart failure. Exposure to altitude or prolonged exercise can alter fluid regulation and can lead to pulmonary congestion or edema. As such, B-lines have been utilized in the field to monitor for pathological lung fluid accumulation. However, ultrasound lung comet lines might not be as reliable for identifying extravascular lung water (EVLW) as previously thought in healthy individuals exercising at altitude where an increase in the number of ultrasound lung comets would reflect fluid buildup in the interstitial space of the alveoli and pulmonary capillaries. This report will focus on reviewing the literature and our data from a group of ultraendurance runners that completed the Ultra Trail Mont Blanc race that demonstrates that lung comet tails may not always be evidence of pathological fluid accumulation in healthy individuals and as such should be used to assess EVLW in concert with other diagnostic testing.

Evaluation of the prognostic value of lncRNA UCA1 combined with extravascular lung water index and lung ultrasound score in patients with acute lung injury.

INTRODUCTION: Acute lung injury (ALI) is a common and rapidly developing critical inflammatory lung disease in clinic. This study investigated the predictive value of lncRNA UCA1, extravascular lung water index (EVLWI), and lung ultrasound score (LUS) in predicting the overall outcome of patients with ALI. METHODS: Patients with ALI were recruited for detecting the content of UCA1, EVLWI, and LUS. All patients were cataloged into the survival group and death group according to the prognosis. The discrepancy of UCA1, EVLWI, and LUS was compared in the two groups. The prognostic significance of UCA1, EVLWI, LUS, and their combination was estimated by logistic regression and the receiver operating characteristic (ROC) curve. RESULTS: The levels of UCA1, LUS, and EVLWI were elevated in the death group compared with the survival group. The content of UCA1 was positively correlated with LUS scores and EVLWI scores. UCA1, LUS, and EVLWI were independent indicators of predicting the prognosis of patients with ALI. The ROC curve reflected that UCA1, LUS, and EVLWI could forecast the endpoint events of patients with ALI whereas their combined approach had the highest accuracy. CONCLUSION: Highly expressed UCA1 is a biomarker in forecasting the outcome of patients with ALI. It had high accuracy in predicting the endpoint of patients with ALI when combined with LUS and EVLWI.

Optimizing Fluid Management Guided by Volumetric Parameters in Patients with Sepsis and ARDS.

We compared two de-escalation strategies guided by either extravascular lung water or global end-diastolic volume-oriented algorithms in patients with sepsis and ARDS. Sixty patients with sepsis and ARDS were randomized to receive de-escalation fluid therapy, guided either by the extravascular lung water index (EVLWI, n = 30) or the global end-diastolic volume index (GEDVI, n = 30). In cases of GEDVI > 650 mL/m2 or EVLWI > 10 mL/kg, diuretics and/or controlled ultrafiltration were administered to achieve the cumulative 48-h fluid balance in the range of 0 to -3000 mL. During 48 h of goal-directed de-escalation therapy, we observed a decrease in the SOFA score (p < 0.05). Extravascular lung water decreased only in the EVLWI-oriented group (p < 0.001). In parallel, PaO2/FiO2 increased by 30% in the EVLWI group and by 15% in the GEDVI group (p < 0.05). The patients with direct ARDS demonstrated better responses to dehydration therapy concerning arterial oxygenation and lung fluid balance. In sepsis-induced ARDS, both fluid management strategies, based either on GEDVI or EVLWI, improved arterial oxygenation and attenuated organ dysfunction. The de-escalation therapy was more efficient for direct ARDS.

A deep learning model enables accurate prediction and quantification of pulmonary edema from chest X-rays.

BACKGROUND: A quantitative assessment of pulmonary edema is important because the clinical severity can range from mild impairment to life threatening. A quantitative surrogate measure, although invasive, for pulmonary edema is the extravascular lung water index (EVLWI) extracted from the transpulmonary thermodilution (TPTD). Severity of edema from chest X-rays, to date is based on the subjective classification of radiologists. In this work, we use machine learning to quantitatively predict the severity of pulmonary edema from chest radiography. METHODS: We retrospectively included 471 X-rays from 431 patients who underwent chest radiography and TPTD measurement within 24 h at our intensive care unit. The EVLWI extracted from the TPTD was used as a quantitative measure for pulmonary edema. We used a deep learning approach and binned the data into two, three, four and five classes increasing the resolution of the EVLWI prediction from the X-rays. RESULTS: The accuracy, area under the receiver operating characteristic curve (AUROC) and Mathews correlation coefficient (MCC) in the binary classification models (EVLWI < 15, ≥ 15) were 0.93 (accuracy), 0.98 (AUROC) and 0.86(MCC). In the three multiclass models, the accuracy ranged between 0.90 and 0.95, the AUROC between 0.97 and 0.99 and the MCC between 0.86 and 0.92. CONCLUSION: Deep learning can quantify pulmonary edema as measured by EVLWI with high accuracy.

Transpulmonary thermodilution.

PURPOSE OF REVIEW: The purpose of this article is to systematically review and critically assess the existing data regarding the use of transpulmonary thermodilution (TPTD), by providing a detailed description of technical aspects of TPTD techniques, appraising the use of TPTD-derived parameters in specific clinical settings, and exploring the limits of this technique. RECENT FINDINGS: The aim of hemodynamic monitoring is to optimize cardiac output ( CO ) and therefore improve oxygen delivery to the tissues. Hemodynamic monitoring plays a fundamental role in the management of acutely ill patients. TPTD is a reliable, multiparametric, advanced cardiopulmonary monitoring technique providing not only hemodynamic parameters related to cardiac function, but also to the redistribution of the extravascular water in the thorax. The hemodynamic monitors available in the market usually couple the intermittent measurement of the CO by TPTD with the arterial pulse contour analysis, offering automatic calibration of continuous CO and an accurate assessment of cardiac preload and fluid responsiveness. SUMMARY: The TPTD is an invasive but well tolerated, multiparametric, advanced cardiopulmonary monitoring technique, allowing a comprehensive assessment of cardiopulmonary condition. Beyond the CO estimation, TPTD provides several indices that help answering questions that clinicians ask themselves during hemodynamic management. TPTD-guided algorithm obtained by pulse contour analysis may be useful to optimize fluid resuscitation by titrating fluid therapy according to functional hemodynamic monitoring and to define safety criteria to avoid fluid overload by following the changes in the extravascular lung water (EVLW) and pulmonary vascular permeability index (PVPI).

A deep learning model to identify the fluid overload status in critically ill patients based on chest X-ray images.

INTRODUCTION: Recent studies have highlighted adverse outcomes of fluid overload in critically ill patients. Therefore, its early recognition is essential for the management of these patients. OBJECTIVES: Our aim was to propose a deep learning (DL) model using data from noninvasive chest X­ray (CXR) imaging associated with the fluid overload status. PATIENTS AND METHODS: We collected data from the Medical Information Mart for Intensive Care IV (MIMIC­IV, v. 1.0) and MIMIC Chest X­Ray (v. 2.0.0) databases for modeling, and from our hospital database for testing. The extravascular lung water index (ELWI) greater than 10 ml/kg and the global end-diastolic volume index (GEDI) greater than 700 ml/m2 were used as threshold values for the fluid overload status. A DL model with a transfer learning strategy was proposed to predict the fluid overload status based on CXR images, and compared with clinical and semantic label models. Additionally, a visualization technique was adopted to determine the important areas of features in the input images. RESULTS: The DL model showed a relatively strong performance for predicting the ELWI (area under the curve [AUC] = 0.896; 95% CI, 0.819-0.972 and AUC = 0.718; 95% CI, 0.594-0.822, respectively) and the GEDI status (AUC = 0.814; 95% CI, 0.699-0.930 and AUC = 0.649; 95% CI, 0.510-0.787, respectively) in both the primary and the test cohort. The performance was better than that of the clinical and semantic label models. CONCLUSIONS: As CXR is routinely used in the intensive care unit, a simple, fast, low­cost, and noninvasive DL model based on this modality can be regarded as an effective supplementary tool for identifying fluid overload, and should be widely adopted in the clinical setting, especially when invasive hemodynamic monitoring is not available.

Effects of changes in veno-venous extracorporeal membrane oxygenation blood flow on the measurement of intrathoracic blood volume and extravascular lung water index: a prospective interventional study.

In severe acute respiratory distress syndrome (ARDS), veno-venous extracorporeal membrane oxygenation (V-V ECMO) has been proposed as a therapeutic strategy to possibly reduce mortality. Transpulmonary thermodilution (TPTD) enables monitoring of the extravascular lung water index (EVLWI) and cardiac preload parameters such as intrathoracic blood volume index (ITBVI) in patients with ARDS, but it is not generally recommended during V-V ECMO. We hypothesized that the amount of extracorporeal blood flow (ECBF) influences the calculation of EVLWI and ITBVI due to recirculation of indicator, which affects the measurement of the mean transit time (MTt), the time between injection and passing of half the indicator, as well as downslope time (DSt), the exponential washout of the indicator. EVLWI and ITBVI were measured in 20 patients with severe ARDS managed with V-V ECMO at ECBF rates from 6 to 4 and 2 l/min with TPTD. MTt and DSt significantly decreased when ECBF was reduced, resulting in a decreased EVLWI (26.1 [22.8-33.8] ml/kg at 6 l/min ECBF vs 22.4 [15.3-31.6] ml/kg at 4 l/min ECBF, p < 0.001; and 13.2 [11.8-18.8] ml/kg at 2 l/min ECBF, p < 0.001) and increased ITBVI (840 [753-1062] ml/m2 at 6 l/min ECBF vs 886 [658-979] ml/m2 at 4 l/min ECBF, p < 0.001; and 955 [817-1140] ml/m2 at 2 l/min ECBF, p < 0.001). In patients with severe ARDS managed with V-V ECMO, increasing ECBF alters the thermodilution curve, resulting in unreliable measurements of EVLWI and ITBVI. German Clinical Trials Register (DRKS00021050). Registered 14/08/2018. https://www.drks.de/drks_web/navigate.do?navigationId=trial.HTML&TRIAL_ID=DRKS00021050.

A century of exercise physiology: lung fluid balance during and following exercise.

PURPOSE: This review recalls the principles developed over a century to describe trans-capillary fluid exchanges concerning in particular the lung during exercise, a specific condition where dyspnea is a leading symptom, the question being whether this symptom simply relates to fatigue or also implies some degree of lung edema. METHOD: Data from experimental models of lung edema are recalled aiming to: (1) describe how extravascular lung water is strictly controlled by "safety factors" in physiological conditions, (2) consider how waning of "safety factors" inevitably leads to development of lung edema, (3) correlate data from experimental models with data from exercising humans. RESULTS: Exercise is a strong edemagenic condition as the increase in cardiac output leads to lung capillary recruitment, increase in capillary surface for fluid exchange and potential increase in capillary pressure. The physiological low microvascular permeability may be impaired by conditions causing damage to the interstitial matrix macromolecular assembly leading to alveolar edema and haemorrhage. These conditions include hypoxia, cyclic alveolar unfolding/folding during hyperventilation putting a tensile stress on septa, intensity and duration of exercise as well as inter-individual proneness to develop lung edema. CONCLUSION: Data from exercising humans showed inter-individual differences in the dispersion of the lung ventilation/perfusion ratio and increase in oxygen alveolar-capillary gradient. More recent data in humans support the hypothesis that greater vasoconstriction, pulmonary hypertension and slower kinetics of alveolar-capillary O2 equilibration relate with greater proneness to develop lung edema due higher inborn microvascular permeability possibly reflecting the morpho-functional features of the air-blood barrier.

MONITORING OF PULMONARY EDEMA USING ULTRASOUND RADIOFREQUENCY SIGNAL.

ABSTRACT: Objectives: Excessive accumulation of extravascular lung water impairs respiratory gas exchange and results in respiratory distress. Real-time radiofrequency signals of ultrasound can continuously and quantitatively monitor excessive lung water. This study aims to evaluate the availability of continuous real-time quantitative pulmonary edema monitoring using ultrasound radiofrequency signals and compare it with Pa o2 (partial pressure of arterial oxygen)/F io2 (fraction of inspired oxygen) (PF) ratio, conventional lung ultrasound, and the Hounsfield unit of chest computed tomography. Methods: Male Yorkshire pigs (40.5 ± 0.5 kg) were anesthetized and mechanically ventilated. A balanced crystalloid was administered to induce hydrostatic pulmonary edema. Three different infusion rates of 2, 4, and 6 mL/kg per minute were tested to determine the infusion rate for the appropriate swine model. The chest computed tomography and ultrasonography with radiofrequency signals were taken every 5 min during the full inspiration. The ultrasonography scans with radiofrequency signals were measured at the intercostal space where the line crossing the two armpits and the right anterior axillary line intersected. Results: The infusion rate of fluid for the pulmonary edema model was determined to be 6 mL/kg per minute, and a total of four pigs were tested at an injection rate of 6 mL/kg. The adjusted R2 values of regression analysis between the radiofrequency signal and computer tomography Hounsfield score were 0.990, 0.993, 0.988, and 0.993 (all P values <0.05). All radiofrequency signal changes preceded changes in PF ratio or lung ultrasound changes. The area under the receiver operating characteristic curve of the radiofrequency signal for predicting PF ratio <300 was 0.88 (95% confidence interval, 0.82-0.93). Conclusion: We evaluated ultrasound radiofrequency signals to assess pulmonary edema in a swine model that can worsen gradually and showed that quantitative ultrasound radiofrequency signal analysis could assess pulmonary edema and its progression before PF ratio or lung ultrasound changes.

Evaluation of extravascular lung water index in critically ill patients based on lung ultrasound radiomics analysis combined with machine learning / 中华危重病急救医学

OBJECTIVE@#To explore lung ultrasound radiomics features which related to extravascular lung water index (EVLWI), and to predict EVLWI in critically ill patients based on lung ultrasound radiomics combined with machine learning and validate its effectiveness.@*METHODS@#A retrospective case-control study was conducted. The lung ultrasound videos and pulse indicated continuous cardiac output (PiCCO) monitoring results of critically ill patients admitted to the department of critical care medicine of the First Affiliated Hospital of Guangxi Medical University from November 2021 to October 2022 were collected, and randomly divided into training set and validation set at 8:2. The corresponding images from lung ultrasound videos were obtained to extract radiomics features. The EVLWI measured by PiCCO was regarded as the "gold standard", and the radiomics features of training set was filtered through statistical analysis and LASSO algorithm. Eight machine learning models were trained using filtered radiomics features including random forest (RF), extreme gradient boost (XGBoost), decision tree (DT), Naive Bayes (NB), multi-layer perceptron (MLP), K-nearest neighbor (KNN), support vector machine (SVM), and Logistic regression (LR). Receiver operator characteristic curve (ROC curve) was plotted to evaluate the predictive performance of models on EVLWI in the validation set.@*RESULTS@#A total of 151 samples from 30 patients were enrolled (including 906 lung ultrasound videos and 151 PiCCO monitoring results), 120 in the training set, and 31 in the validation set. There were no statistically significant differences in main baseline data including gender, age, body mass index (BMI), mean arterial pressure (MAP), central venous pressure (CVP), heart rate (HR), cardiac index (CI), cardiac function index (CFI), stroke volume index (SVI), global end diastolic volume index (GEDVI), systemic vascular resistance index (SVRI), pulmonary vascular permeability index (PVPI) and EVLWI. The overall EVLWI range in 151 PiCCO monitoring results was 3.7-25.6 mL/kg. Layered analysis showed that both datasets had EVLWI in the 7-15 mL/kg interval, and there was no statistically significant difference in EVLWI distribution. Two radiomics features were selected by using LASSO algorithm, namely grayscale non-uniformity (weight was -0.006 464) and complexity (weight was -0.167 583), and they were used for modeling. ROC curve analysis showed that the MLP model had better predictive performance. The area under the ROC curve (AUC) of the prediction validation set EVLWI was higher than that of RF, XGBoost, DT, KNN, LR, SVM, NB models (0.682 vs. 0.658, 0.657, 0.614, 0.608, 0.596, 0.557, 0.472).@*CONCLUSIONS@#The gray level non-uniformity and complexity of lung ultrasound were the most correlated radiomics features with EVLWI monitored by PiCCO. The MLP model based on gray level non-uniformity and complexity of lung ultrasound can be used for semi-quantitative prediction of EVLWI in critically ill patients.

Extravascular lung water levels are associated with mortality: a systematic review and meta-analysis.

BACKGROUND: The prognostic value of extravascular lung water (EVLW) measured by transpulmonary thermodilution (TPTD) in critically ill patients is debated. We performed a systematic review and meta-analysis of studies assessing the effects of TPTD-estimated EVLW on mortality in critically ill patients. METHODS: Cohort studies published in English from Embase, MEDLINE, and the Cochrane Database of Systematic Reviews from 1960 to 1 June 2021 were systematically searched. From eligible studies, the values of the odds ratio (OR) of EVLW as a risk factor for mortality, and the value of EVLW in survivors and non-survivors were extracted. Pooled OR were calculated from available studies. Mean differences and standard deviation of the EVLW between survivors and non-survivors were calculated. A random effects model was computed on the weighted mean differences across the two groups to estimate the pooled size effect. Subgroup analyses were performed to explore the possible sources of heterogeneity. RESULTS: Of the 18 studies included (1296 patients), OR could be extracted from 11 studies including 905 patients (464 survivors vs. 441 non-survivors), and 17 studies reported EVLW values of survivors and non-survivors, including 1246 patients (680 survivors vs. 566 non-survivors). The pooled OR of EVLW for mortality from eleven studies was 1.69 (95% confidence interval (CI) [1.22; 2.34], p < 0.0015). EVLW was significantly lower in survivors than non-survivors, with a mean difference of -4.97 mL/kg (95% CI [-6.54; -3.41], p < 0.001). The results regarding OR and mean differences were consistent in subgroup analyses. CONCLUSIONS: The value of EVLW measured by TPTD is associated with mortality in critically ill patients and is significantly higher in non-survivors than in survivors. This finding may also be interpreted as an indirect confirmation of the reliability of TPTD for estimating EVLW at the bedside. Nevertheless, our results should be considered cautiously due to the high risk of bias of many studies included in the meta-analysis and the low rating of certainty of evidence. Trial registration the study protocol was prospectively registered on PROSPERO: CRD42019126985.

[Effects on extravascular lung water of lung protective ventilation strategy applied on piglets with acute respiratory distress syndrome induced by paraquat].

Objective: To study the effects on extravascular lung water of lung protective ventilation strategy applying on piglets with acute respiratory distress syndrome (ARDS) induced by paraquat (PQ) under pulse indicating continuous cardiac output (PiCCO) monitoring. Methods: The piglets models with ARDS induced by PQ were established in June 2020 and all of them were received mechanical ventilation and divided into three groups according to tidal volume (V(T)) : small V(T) group (6 ml/kg) , middle V(T) group (10 ml/kg) and large V(T) group (15 ml/kg) , there were 5 piglets in each group. The positive end expiratory pressure (PEEP) were all setup on 10 cmH(2)O. The indexes such as arterial blood gas analysis, oxygenation index (OI) , extravascular lung water index (ELWI) and pulmonary vascular permeability index (PVPI) were monitored at time of before the model was established (baseline) , time of the model was established (t(0)) and 2 h (t(2)) , 4 h (t(4)) , 6 h (t(6)) after mechanical ventilation. Lung tissue were punctured at time of baseline, t(0) and t(6) to be stained by Hematoxylin-eosin (HE) staining and pulmonary pathology were observed under light microscopy. Results: The heart rate (HR) , mean arterial pressure (MAP) and partial pressure of carbon dioxide (PaCO(2)) of all groups were higher than the base value while the pH values, partial pressure of oxygen (PaO(2)) and OI were lower than the base value when the models were established (P<0.05) . After mechanical ventilation, the HR and MAP values of all groups at t(2), t(4) and t(6) were lower than t(0) while the PaCO(2) of t(4) and t(6) were all higher than t(0), the differences were statistically significant (P<0.05) . The PaO(2) and OI of all groups showed a trend of rising at first and then decreasing after mechanical ventilation. The MAP, PaO(2), PaCO(2) and OI of the middle V(T) group and large V(T) group were apparently lower than that of the small V(T) group at t(2), t(4) and t(6) (P<0.05) . The ELWI and PVPI at t(0) of all groups were higher than that of baseline (P<0.05) . The ELWI of the small V(T) group at t(6) were lower than t(0) of the same group and t(6) of the middle V(T) group and large V(T) group (P<0.05) . HE staining showed congestion and edema of alveolar tissue, swelling of capillaries, exudation of red blood cells and widening of alveolar septum in piglets after successful modeling. And further widening of alveolar septum and rupture of alveolar septum could be seen in the lung tissues of each group at t(6), and the injury was the slightest in the small V(T) group. Conclusion: The lung protective ventilation strategy can alleviate the extravascular lung water and ARDS induced by PQ and improve oxygenation.

Global longitudinal strain assessment of cardiac function and extravascular lung water formation after diving using semi-closed circuit rebreather.

PURPOSE: The aim of the present investigation is to study the relationship of ventricular global longitudinal strain (GLS) and ultrasound lung comets (ULC) formation to establish a link between extravascular pulmonary water formation and cardiac contractile dysfunction. METHODS: This is a prospective observational study including 14 active military divers. The subjects performed two sea dives of 120 min each with a semi-closed SCUBA circuit at 10 m depth. Divers were examined at baseline, 15 min (D1) and 60 min (D2) after diving. The evaluation included pulmonary and cardiac echography (including speckle tracking techniques). Blood samples were drawn at baseline and after diving, assessing hs-TnT and Endothelin-1. RESULTS: ULC were detected in 9 (64.2%) and 8 (57.1%) of the subjects after D1 and D2 respectively. No differences were found in right and left ventricular GLS after both immersions (RV: Baseline: - 17.9 4.9 vs. D1: - 17.2 6.5 and D2: - 16.7 5.8 s-1; p = 0.757 and p = 0.529; LV: Baseline: - 17.0 2.3 vs. D1: - 17.4 2.1 and D2: - 16.9 2.2 s-1; p = 0.546 and p = 0.783). However, a decrease in atrial longitudinal strain parameters have been detected after diving (RA: Baseline: 35.5 9.2 vs. D1: 30.3 12.8 and D2: 30.7 13.0 s-1; p = 0.088 and p = 0.063; LA: Baseline: 39.0 10.0 vs. D1: 31.6 6.1 and D2: 32.4 10.6 s-1; p = 0.019 and p = 0.054). CONCLUSION: In the present study, no ventricular contractile dysfunction was observed. However, increase pulmonary vasoconstriction markers were present after diving.