Single and multiple scattering quantitative ultrasound methods in the lung

Abstract

Lung ultrasound images are qualitative and rely on subjective interpretation for diagnosis. Ideally, quantitative ultrasound (QUS) methods applied to lung ultrasound data would provide biomarkers associated with specific diseases in an absolute fashion. While QUS methods are successful in numerous organ systems in preclinical and clinical applications, their direct application to lung ultrasound data requires taking multiple scattering (MS) into account. This presentation shows QUS methods independently exploiting MS and single scattering (SS) components of radio-frequency (RF) data, acquired from lungs from rodent models of pulmonary edema and fibrosis. SS and MS components were obtained using singular value decomposition and eigenvalue thresholding. Spectral and envelope QUS parameters were computed from the SS component. MS-based QUS parameters were obtained by extracting the diffusion constant and tracking the SS intensity decay rate. Stepwise linear discriminant analyses using only three QUS parameters yielded strong correlations between wet-to-dry ratio in healthy and pulmonary edema rats. Strong correlations were also obtained between QUS parameters and modified histological Ashcroft scores in healthy and pulmonary fibrosis rats. These results demonstrate the value of QUS parameters obtained from SS and MS approaches for assessing lung diseases with high specificity and sensitivity. [Supported in part by R21HL154156 and W81XWH1810101.]