Fig 1.
Representative lung ultrasound images used for scoring.
0 Normal pattern with horizontal reverberation of the pleural line (also known as A lines). 1 Vertical hyperechoic artifacts (also known as B lines) more than 3 per field, well spaced. Thin, regular pleural image. 2 Coalescent B lines, thick pleural image with or without small subpleural consolidations. 3 Thick and irregular pleural image with evident subpleural consolidations1.
Fig 2.
First-order statistics analysis showing ROI distributions (upper panels) and the calculated intensity histograms (lower panels).
Table 1.
Main demographic variables of the study cohort.
Fig 3.
Correlation of visual LUS score with the PaO2/FiO2 ratio (3A); its ROC curve for a cut off value of less than 200 gave an AUC = 0.83 (3C).
The correlation of visual LUS score with alveolar arterial gradient is shown in panel 3B; its ROC curve for a cut-off value of more than 150, shown in panel 3D, gave an AUC = 0.844.
Fig 4.
Gray scale analysis results for small region of interest (50K pixels).
Correlation of the gray scale analysis with the PaO2/FiO2 ratio (3A); its ROC curve for a cut off value of less than 200 had an AUC = 0.71 (3C). The correlation of the gray scale analysis with the alveolar arterial gradient is shown in panel 3B; its ROC curve for a cut-off value of more than 150, shown in panel 3D, resulted in an AUC = 0.55.
Fig 5.
Gray scale analysis results for large region of interest (100K pixels).
Correlation of the gray scale analysis score with the PaO2/FiO2 ratio (3A); its ROC curve for a cut off value of less than 200 had an AUC = 0.72 (3C). The correlation of the gray scale analysis with the alveolar arterial gradient is shown in panel 3B; its ROC curve for a cut-off value of more than 150, shown in panel 3D, gave an AUC = 0.66.
Table 2.
Performance (AUC) for selected groups of features with small and large ROI.
Fig 6.
Performance (AUC) for both indexes and ROI sizes as a function of the number of principal components kept in the feature vector.
Components are sorted by descending variance.
Fig 7.
Fraction of the total variance of the full feature vector explained by the first principal components.
Components are sorted by descending variance.