Machine learning-based hemodynamics quantitative assessment of pulmonary circulation using computed tomographic pulmonary angiography.

Obstructive sleep apnea and pulmonary hypertension: the pendulum swings again.

Right Heart Catheterization—To Do or Not To Do? Introducing a New Diagnostic Algorithm for Pulmonary Hypertension

The septal angle, an angle between the interventricular septum and the line connecting the sternum midpoint and thoracic vertebral spinous process, as measured by computed tomographic pulmonary angiography (CTPA), has been observed to be increased in patients with pulmonary hypertension (PH), but its meaning remains unclear. The aim of this study was to investigate the potential role of the septal angle in evaluating hemodynamics and its association with N-terminal pro-B-type natriuretic peptide (NT-proBNP) in patients with PH. Patients with PH (n=106), including 76 with chronic thromboembolic pulmonary hypertension (CTEPH) and 30 with pulmonary artery hypertension (PAH), were retrospectively reviewed. The patients underwent CTPA prior to right-heart catheterization. The septal angle was measured on transversal CTPA images. Hemodynamic parameters were evaluated by right-heart catheterization. The level of plasma NT-proBNP was measured by enzyme-linked sandwich immunoassay. The septal angle had a moderate correlation with cardiac output (CO; r=−0.535, P=0.000) and a high correlation with pulmonary vascular resistance (PVR; r=0.642, P=0.000). The mean level of NT-proBNP in PH was 1,716.09±1,498.30 pg/ml, which correlated with the septal angle (r=0.693, P=0.000). In a stepwise forward regression analysis, the Septal angle was entered into the final equation for predicting PVR, leading to the following equation: PVR = 28.256 × Septal angle - 728.72. In CTEPH, the Septal angle strongly correlated with NT-proBNP (r=0.668, P=0.000) and PVR (r=0.676, P=0.000). In PAH, the Septal angle strongly correlated with NT-proBNP (r=0.616, P=0.003) and PVR (r=0.623, P=0.000). The CTPA-derived Septal angle is a superior predictor for evaluating and monitoring the level of NT-proBNP and PVR in patients with PH.

BackgroundTo compare the performance between one-slice two-dimensional (2D) and whole-volume three-dimensional (3D) computed tomography (CT)-based radiomics models in the prediction of lymphovascular invasion (LVI) status in esophageal squamous cell carcinoma (ESCC).MethodsTwo hundred twenty-four patients with ESCC (158 LVI-absent and 66 LVI-present) were enrolled in this retrospective study. The enrolled patients were randomly split into the training and testing sets with a 7:3 ratio. The 2D and 3D radiomics features were derived from the primary tumors’ 2D and 3D regions of interest (ROIs) using 1.0 mm thickness contrast-enhanced CT (CECT) images. The 2D and 3D radiomics features were screened using inter-/intra-class correlation coefficient (ICC) analysis, Wilcoxon rank-sum test, Spearman correlation test, and the least absolute shrinkage and selection operator, and the radiomics models were built by multivariate logistic stepwise regression. The performance of 2D and 3D radiomics models was assessed by the area under the receiver operating characteristic (ROC) curve. The actual clinical utility of the 2D and 3D radiomics models was evaluated by decision curve analysis (DCA).ResultsThere were 753 radiomics features from 2D ROIs and 1130 radiomics features from 3D ROIs, and finally, 7 features were retained to construct 2D and 3D radiomics models, respectively. ROC analysis revealed that in both the training and testing sets, the 3D radiomics model exhibited higher AUC values than the 2D radiomics model (0.930 versus 0.852 and 0.897 versus 0.851, respectively). The 3D radiomics model showed higher accuracy than the 2D radiomics model in the training and testing sets (0.899 versus 0.728 and 0.788 versus 0.758, respectively). In addition, the 3D radiomics model has higher specificity and positive predictive value, while the 2D radiomics model has higher sensitivity and negative predictive value. The DCA indicated that the 3D radiomics model provided higher actual clinical utility regarding overall net benefit than the 2D radiomics model.ConclusionsBoth 2D and 3D radiomics features can be employed as potential biomarkers to predict the LVI in ESCC. The performance of the 3D radiomics model is better than that of the 2D radiomics model for the prediction of the LVI in ESCC.

Ventilation-Perfusion Scan Is Superior to CT Pulmonary Angiography in Detection of Chronic Thromboembolic Pulmonary Hypertension in SCD Adults

Pulmonary hypertension (PH) in adults with sickle cell disease (SCD) is associated with early mortality. Chronic thromboembolic PH (CTEPH) is an important complication and contributor to PH in SCD but is likely underappreciated. Guidelines recommend ventilation-perfusion (V/Q) scintigraphy as the imaging modality of choice to exclude CTEPH. Data on V/Q scanning are limited in SCD. Our objective was to compare the performance of V/Q scanning with that of CT pulmonary angiography (CTPA) and to report clinical outcomes associated with abnormal V/Q findings. Methods: Laboratory data, echocardiography, 6-min-walk testing, V/Q scanning, CTPA, and right heart catheterization (RHC) were prospectively obtained. High-probability and intermediate-probability V/Q findings were considered to be abnormal. Included for analysis were 142 SCD adults (aged 40.1 ± 13.7 y, 83 women, 87% hemoglobin SS) in a stable state enrolled consecutively between March 13, 2002, and June 8, 2017. Results: V/Q results were abnormal in 65 of 142 patients (45.8%). CTPA was positive for pulmonary embolism in 16 of 60 (26.7%). RHC confirmed PH (mean pulmonary artery pressure ≥ 25 mmHg) in 46 of 64 (71.9%), of whom 34 (73.9%) had abnormal V/Q findings. Among those without PH by RHC (n = 18), 2 of 18 patients had abnormal V/Q findings. Thirty-three patients had a complete dataset (V/Q scanning, CTPA, and RHC); 29 of 33 had abnormal RHC findings, of whom 26 had abnormal V/Q findings, compared with 11 who had abnormal CTPA findings. There was greater concordance between V/Q findings and RHC (κ-value = 0.53; P < 0.001) than between CTPA and RHC (κ-value = 0.13; P = 0.065). The sensitivity and specificity for V/Q scanning was 89.7% and 75.0%, respectively, whereas CTPA had sensitivity of 37.3% and specificity of 100%. Abnormal V/Q finding swere associated with hemodynamic severity (mean pulmonary artery pressure, 35.2 ± 9.6 vs. 26.9 ± 10.5 mm Hg, P = 0.002; transpulmonary gradient, 21.5 ± 9.7 vs. 12.16 ± 11 mmHg, P = 0.005; and pulmonary vascular resistance, 226.5 ± 135 vs. 140.7 ± 123.7 dynes⋅s⋅cm-5, P = 0.013) and exercise capacity (6-min-walk distance, 382.8 ± 122.3 vs. 442.3 ± 110.6 m, P < 0.010). Thirty-four deaths were observed over 15 y. All-cause mortality was higher in the abnormal-V/Q group (21 [61.8%]) than in the normal-V/Q group (13 [38.2%]) (log-rank test, P = 0.006; hazard ratio, 2.54). Conclusion: V/Q scanning is superior to CTPA in detecting thrombotic events in SCD. Abnormal V/Q findings are associated with PH, worse hemodynamics, lower functional capacity, and higher mortality. Despite high sensitivity in detecting CTEPH, V/Q scanning is underutilized. We recommend the use of V/Q scanning in the evaluation of dyspnea in adult SCD patients given the important implications toward management.

BACKGROUND. Noninvasive tests for pulmonary hypertension (PH) are needed to help select patients for diagnostic right heart catheterization (RHC). CT pulmonary angiography (CTPA) is commonly performed for suspected PH. OBJECTIVE. The purpose of this study was to assess the utility of CTPA-based cardiac chamber volumetric measurements for the diagnosis of PH in comparison with echocardiographic and conventional CTPA parameters, with the 2018 updated hemodynamic definition used as reference. METHODS. This retrospective study included 109 patients (72 women and 37 men; median age, 68 years) who underwent nongated CTPA, transthoracic echocardiography, and RHC for the workup of suspected PH between August 2013 and February 2016. Two radiologists independently used automated 3D segmentation software to determine the volumes of the right ventricle (RV), right atrium (RA), left ventricle (LV), and left atrium (LA) and also measured the axial diameters of the cardiac chambers, main pulmonary artery, and ascending aorta. Interobserver agreement was assessed, and mean values were obtained; one observer repeated volumetric measurements to assess intraobserver agreement. ROC analysis was used to assess diagnostic performance for the detection of PH. A multivariable binary logistic regression model was established. RESULTS. A total of 60 of 109 patients had PH. Intra- and interobserver agreements were excellent for all volume measurements (intraclass correlation coefficients, 0.935-0.999). In patients with PH versus those without PH, RV volume was 172.6 versus 118.1 mL, and RA volume was 130.2 versus 77.0 mL (both p < .05). Cardiac chamber measurements with the highest AUC for PH were the RV/LV volume ratio and RA volume (both 0.791). Significant predictors of PH20 (as defined using the 2018 hemodynamic definition from the Sixth World Symposium on Pulmonary Hypertension) after adjustment for age, sex, and body surface area included RV volume per 10 mL (odds ratio [OR], 1.21), RA volume per 10 mL (OR, 1.27), RV/LV volume ratio (OR, 2.91), and RA/LA volume ratio (OR, 11.22). Regression analysis yielded a predictive model for PH that contained two independent predictors: echocardiographic pulmonary arterial systolic pressure and CTPA-based RA volume; the model had an AUC of 0.898, sensitivity of 83.3%, and specificity of 85.7%. CONCLUSION. Automated cardiac chamber volumetry using nongated CTPA, particularly of the RA, provides incremental utility relative to echocardiographic and conventional CTPA parameters for diagnosis of PH. CLINICAL IMPACT. Automated volumetry of cardiac chambers based on nongated CTPA may facilitate early noninvasive detection of PH, identifying patients who warrant further evaluation by RHC.

Computed tomography pulmonary angiography (CTPA) is frequently performed in patients with pulmonary hypertension (PH) and may aid non-invasive estimation of pulmonary hemodynamics. We, therefore, investigated automated volumetry of intrapulmonary vasculature on CTPA, separated into core and peel fractions of the lung volume and its potential to differentially reflect pulmonary hemodynamics in patients with pre- and postcapillary PH. A retrospective case-control study of 72 consecutive patients with PH according to the 2022 joint guidelines of the European Society of Cardiology and the European Respiratory Society who underwent right heart catheterization (RHC) and CTPA within 7 days between August 2013 and February 2016 at Thoraxklinik at Heidelberg University Hospital (Heidelberg, Germany) was conducted. Vessel segmentation was performed using the in-house software YACTA. Vascular volumes in different core and peel fractions of the lung were corrected for body surface area. Spearman correlation coefficients with mean pulmonary arterial pressure (mPAP), pulmonary arterial wedge pressure (PAWP) and pulmonary vascular resistance (PVR) were calculated, and a linear regression analysis was done to account for potential confounders. Median age of the study sample was 71.5 years [interquartile range (IQR), 60.0-77.0 years], 48 (66.67%) were female. Median mPAP was 35.5 mmHg (IQR, 27.0-47.2 mmHg). Postcapillary PH was present in 24/72 (33.3%) patients and precapillary PH in 48/72 (66.7%) patients. Moderate to strong correlations between core intrapulmonary vessel volumes and mPAP were observed in postcapillary PH patients with a maximum at 50% core lung volume (r=0.71, P<0.001). No significant influence of age or sex on this relationship was identified. Correlation with RHC measurements was weak or negligible in patients with precapillary PH. Automated volumetry of vessels in the core lung strongly correlated with mPAP in patients with postcapillary PH and has potential for non-invasive assessment of postcapillary PH in patients undergoing CTPA.

Objective: Pulmonary vascular resistance (PVR) is a measurement obtained with invasive right heart catheterization (RHC) that is commonly used for management of patients with pulmonary arterial hypertension (PAH). Computed tomography pulmonary angiography (CTPA) is also done as part of the workup for PAH in some cases. The aim of our study was to assess the correlation of contrast dynamic changes in the main pulmonary artery (MPA) on CTPA with PVR obtained with RHC.Methods: This is an IRB-approved retrospective study performed in two separate institutions (Medical College of Wisconsin and University of Alabama) between January 2010 and December 2013. During CTPA done as test bolus, serial images are acquired at the level of MPA after intravenous injection of contrast to determine timing of the CT acquisition. Since the PVR changes with the degree of PAH, we hypothesize that will be reflected in the contrast kinetics in MPA. A correlation of standard CT metrics (MPA diameter, right pulmonary artery [PA] diameter, left PA diameter, MPA/aorta ratio, and right ventricle/left ventricle [RV/LV] ratio) and dynamic (full width at half maximum) CTPA parameters in patients with known PAH was performed with PVR obtained from RHC done within 30 days. Statistical analysis was performed by Pearson correlation coefficient.Results: Among 221 patients in our database, 37 patients fulfilled the selection criteria. There was a strong correlation between full width half maximum (FWHM) and mean pulmonary artery pressure (mPAP) (r=0.69, p value<0.00001), PVR (r=0.8, p value<0.00001) and indexed PVR (PVRI) (r=0.75, p value<0.00001).Conclusion: FWHM obtained from CTPA strongly correlates with RHC parameters and is potentially more helpful than static measurements for follow-up of patients with known PAH to assess response to treatment or progression.

The Cardio-Obstetrics Patient and the Cardiothoracic Anesthesiologist

ObjectivesThe objective of this study was to compare the predictive performance of 2D and 3D radiomics features in meningioma grade based on enhanced T1 WI images.MethodsThere were 170 high grade meningioma and 170 low grade meningioma were selected randomly. The 2D and 3D features were extracted from 2D and 3D ROI of each meningioma. The Spearman correlation analysis and least absolute shrinkage and selection operator (LASSO) regression were used to select the valuable features. The 2D and 3D predictive models were constructed by naive Bayes (NB), gradient boosting decision tree (GBDT), and support vector machine (SVM). The ROC curve was drawn and AUC was calculated. The 2D and 3D models were compared by Delong test of AUCs and decision curve analysis (DCA) curve.ResultsThere were 1143 features extracted from each ROI. Six and seven features were selected. The AUC of 2D and 3D model in NB, GBDT, and SVM was 0.773 and 0.771, 0.722 and 0.717, 0.733 and 0.743. There was no significant difference in two AUCs (p=0.960, 0.913, 0.830) between 2D and 3D model. The 2D features had a better performance than 3D features in NB models and the 3D features had a better performance than 2D features in GBDT models. The 2D features and 3D features had an equal performance in SVM models.ConclusionsThe 2D and 3D features had a comparable performance in predicting meningioma grade. Considering the issue of time and labor, 2D features could be selected for radiomics study in meningioma.Key pointsThere was a comparable performance between 2D and 3D features in meningioma grade prediction. The 2D features was a proper selection in meningioma radiomics study because of its time and labor saving.

ObjectiveTo evaluate the performance of 2D and 3D radiomics features with different machine learning approaches to classify SPLs based on magnetic resonance(MR) T2 weighted imaging (T2WI).Material and MethodsA total of 132 patients with pathologically confirmed SPLs were examined and randomly divided into training (n = 92) and test datasets (n = 40). A total of 1692 3D and 1231 2D radiomics features per patient were extracted. Both radiomics features and clinical data were evaluated. A total of 1260 classification models, comprising 3 normalization methods, 2 dimension reduction algorithms, 3 feature selection methods, and 10 classifiers with 7 different feature numbers (confined to 3–9), were compared. The ten-fold cross-validation on the training dataset was applied to choose the candidate final model. The area under the receiver operating characteristic curve (AUC), precision-recall plot, and Matthews Correlation Coefficient were used to evaluate the performance of machine learning approaches.ResultsThe 3D features were significantly superior to 2D features, showing much more machine learning combinations with AUC greater than 0.7 in both validation and test groups (129 vs. 11). The feature selection method Analysis of Variance(ANOVA), Recursive Feature Elimination(RFE) and the classifier Logistic Regression(LR), Linear Discriminant Analysis(LDA), Support Vector Machine(SVM), Gaussian Process(GP) had relatively better performance. The best performance of 3D radiomics features in the test dataset (AUC = 0.824, AUC-PR = 0.927, MCC = 0.514) was higher than that of 2D features (AUC = 0.740, AUC-PR = 0.846, MCC = 0.404). The joint 3D and 2D features (AUC=0.813, AUC-PR = 0.926, MCC = 0.563) showed similar results as 3D features. Incorporating clinical features with 3D and 2D radiomics features slightly improved the AUC to 0.836 (AUC-PR = 0.918, MCC = 0.620) and 0.780 (AUC-PR = 0.900, MCC = 0.574), respectively.ConclusionsAfter algorithm optimization, 2D feature-based radiomics models yield favorable results in differentiating malignant and benign SPLs, but 3D features are still preferred because of the availability of more machine learning algorithmic combinations with better performance. Feature selection methods ANOVA and RFE, and classifier LR, LDA, SVM and GP are more likely to demonstrate better diagnostic performance for 3D features in the current study.

Subclinical Left Ventricular Systolic Dysfunction due to Coronary Arterial Thrombosis in a Neonate with Hypoxic Ischemic Encephalopathy Undergoing Therapeutic Hypothermia

Pulmonary arterial hypertension is a progressive and fatal disease and rodents with experimental pulmonary hypertension (PH) are often used to study pathogenic mechanisms, identify therapeutic targets, and develop novel drugs for treatment. Here we describe a hands-on set of experimental approaches including ex vivo lung angiography and histology and in vivo right heart catheterization (RHC) to phenotypically characterize pulmonary hemodynamics and lung vascular structure in normal mice and mice with experimental PH. We utilized Microfil polymer as contrast in our ex vivo lung angiogram to quantitatively examine pulmonary vascular remodeling in mice with experimental PH, and lung histology to estimate pulmonary artery wall thickness. The peripheral lung vascular images were selected to determine the total length of lung vascular branches, the number of branches and the number of junctions in a given area (mm−2). We found that the three parameters determined by angiogram were not significantly different among the apical, middle, and basal regions of the mouse lung from normal mice, and were not influenced by gender (no significant difference between female and male mice). We conducted RHC in mice to measure right ventricular systolic pressure, a surrogate measure for pulmonary artery systolic pressure and right ventricle (RV) contractility (RV ± dP/dtmax) to estimate RV function. RHC, a short time (4–6 min) procedure, did not alter the lung angiography measurements. In summary, utilizing ex vivo angiogram to determine peripheral vascular structure and density in the mouse lung and utilizing in vivo RHC to measure pulmonary hemodynamics are reliable readouts to phenotype normal mice and mice with experimental PH. Lung angiogram and RHC are also reliable approaches to examine pharmacological effects of new drugs on pulmonary vascular remodeling and hemodynamics.

THE ROLE OF PROVOCATIVE TESTING DURING HEMODYNAMIC ASSESSMENT TO IMPROVE PULMONARY HYPERTENSION DIAGNOSIS IN OLDER ADULTS