Estimation of Lung Surface Roughness by Means of an Ultrasound Multifrequency Approach, in silico and in vitro results

Lung ultrasound (LUS) is an important tool to evaluate the state of the lung surface, but is limited to the visual evaluation of imaging artifacts. Specifically, it is mainly focused on the vertical artifacts, which correlate with pathologies characterized by a reduction of air-spaces’ dimensions. In contrast, there exist pathologies, e.g., chronic obstructive pulmonary disease (COPD), characterized by an enlargement of the air-spaces. This enlargement causes the lung surface to behave essentially as a perfect reflector, which does not allow the ultrasound penetration. As air-spaces of different sizes may generate a different roughness level at the lung surface, the roughness estimation could provide a way to characterize the state of lung surface. In this study, we show how a multifrequency quantitative approach has the potential to estimate the roughness of the lung surface by measuring image intensity variations along the lung surface as a function of frequency.