Lung ultrasound aeration semi-quantification in invasively ventilated ICU patients - an alternative to B-line counting

Abstract

Introduction: Reduced lung aeration is currently semi-quantified at the bedside by visual counting of vertical artifacts called B-lines on lung ultrasound (LUS) images. The accuracy of this method can be debated. We hypothesized that visual estimation of B-line coverage is a more accurate alternative to visual B-line counting in invasively ventilated ICU patients. Method: Invasively ventilated patients who received a clinical chest CT scan were included. A 12-region LUS exam was performed minutes before the CT scan. Lung regions scanned through LUS were compared to corresponding regions on CT. Lung aeration on CT was expressed in Hounsfield Units (HU). LUS images with a B-pattern, were visually scored by (a) estimating B-line coverage, expressed as the percentage of the screen covered by B-lines, and (b) counting the number of separately spaced B-lines. Associations between HU from the CT scan and the two B-line scores were tested with linear mixed effects model analysis. Results: The study included 23 patients, who had 87 LUS images with a B-pattern (median 3 [2-4] per patient). The B-line coverage was associated with the HU value from the CT scan; B-line count was not associated with the HU value. Conclusion: The visual B-line count on LUS images is an inaccurate score for lung aeration. Visual estimation of B-line coverage is a promising alternative for semi-quantification of aeration in LUS images.