Quantitative estimation of pulmonary artery wedge pressure from chest radiographs by a regression convolutional neural network

Abstract

Recent studies reported that a convolutional neural network (CNN; a deep learning model) can detect elevated pulmonary artery wedge pressure (PAWP) from chest radiographs, the diagnostic images most commonly used for assessing pulmonary congestion in heart failure. However, no method has been published for quantitatively estimating PAWP from such radiographs. We hypothesized that a regression CNN, an alternative type of deep learning, could be a useful tool for quantitatively estimating PAWP in cardiovascular diseases. We retrospectively enrolled 936 patients with cardiovascular diseases who had undergone right heart catheterization (RHC) and chest radiography and estimated PAWP by constructing a regression CNN based on the VGG16 model. We randomly categorized 80% of the data as training data (training group, n = 748) and 20% as test data (test group, n = 188). Moreover, we tuned the learning rate-one of the model parameters-by 5-hold cross-validation of the training group. Correlations between PAWP measured by RHC [ground truth (GT) PAWP] and PAWP derived from the regression CNN (estimated PAWP) were tested. To visualize how the regression CNN assessed the images, we created a regression activation map (RAM), a visualization technique for regression CNN. Estimated PAWP correlated significantly with GT PAWP in both the training (r = 0.76, P < 0.001) and test group (r = 0.62, P < 0.001). Bland-Altman plots found a mean (SEM) difference between GT and estimated PAWP of - 0.23 (0.16) mm Hg in the training and - 0.05 (0.41) mm Hg in the test group. The RAM showed that our regression CNN model estimated high PAWP by focusing on the cardiomegaly and pulmonary congestion. In the test group, the area under the curve (AUC) for detecting elevated PAWP (≥ 18 mm Hg) produced by the regression CNN model was similar to the AUC of an experienced cardiologist (0.86 vs 0.83, respectively; P = 0.24). This proof-of-concept study shows that regression CNN can quantitatively estimate PAWP from standard chest radiographs in cardiovascular diseases.

Keywords: Artificial intelligence; Deep learning; Diagnostic method; Heart failure.

Fig. 1

Regression convolutional neural network. The structure of the regression convolutional neural network used to estimate pulmonary artery wedge pressure

Fig. 2

Study flowchart. To develop a regression convolutional neural network and verify its generalization error, we randomly categorized all data (N = 936) as training data and test data; 80% of all data were categorized as training data (training group, n = 748) and 20% as test data (test group, n = 188). The training data were split into 5 subsets for 5-hold cross-validation

Fig. 3

Relation between ground truth and estimated pulmonary artery wedge pressure in the training and test groups. A Scatter plots showing the relation between ground truth (GT) and estimated pulmonary artery wedge pressure (PAWP) in the training group. B Bland–Altman plot of the training group data. C Scatter plots showing the relation between GT and estimated PAWP in the test group. D Bland–Altman plot of the test group data. GT ground truth, PAWP pulmonary artery wedge pressure

Fig. 4

Representative cases. Examples of visualization with a regression activation map (RAM). In each case, the original image is on the left and its heatmap is on the right. The red and yellow areas on the heatmap represent the points on which the regression CNN model focused. A Case 1: A 73-year-old man with ischemic heart disease. Ground truth (GT) pulmonary artery wedge pressure (PAWP), 6.0 mm Hg; estimated PAWP, 9.3 mm Hg. B Case 2: A 69-year-old man with ischemic heart disease. GT PAWP, 6.0 mm Hg; estimated PAWP, 7.6 mm Hg. C Case 3: A 60-year-old man with ischemic heart disease. GT PAWP, 26.0 mm Hg; estimated PAWP, 25.1 mm Hg. D Case 4: A 55-year-old man with heart failure. GT PAWP, 44.0 mm Hg; estimated PAWP, 24.0 mm Hg. GT ground truth, PAWP pulmonary artery wedge pressure