A first clinical study on the quantitative characterization of pleural sliding using multifrequency lung ultrasound in-vivo data

Lung Ultrasound (LUS) data analysis is generally limited to the visual interpretation of imaging patterns (IPs). Recent studies have focused on a transition toward a quantitative approach by characterizing these IPs. This study aimed to characterize pleural sliding, defined as a shimmering movement on the pleural line in LUS videos. Specifically, we developed a new methodology to quantify the pleural sliding velocity (SlV). Additionally, the discriminating potential of SlV was assessed for two pathologies, cardiogenic pulmonary edema (CPE) and pneumonia. The dataset included 104 multifrequency RF videos from 28 patients; 15 affected by CPE and 13 affected by pneumonia. For each video, we estimated each pixel velocity (v p) using our proposed velocity model. This model assumed a relation between v p and frequency spectra of the slow-time intensity signal. The SlV was then calculated as the mean of v p of pixels within the pleural line region. The estimated SlV showed values within the physiological range (≈1.50 mm/s – ≈4.75 mm/s). Moreover, the two populations were found to be statistically different for three out of four imaging center frequencies (3–6 MHz). This showed, for the first time, the potential of pleural sliding velocity estimation for discriminating between CPE and pneumonia.