Using singular value distribution of backscattered ultrasound waves for tracking the pulmonary edema caused by COVID-19 in lung: A phantom-based study

A condition caused by COVID-19, pulmonary edema is an abnormal buildup of fluid in lungs. B-lines tracking of ultrasound images is a common way to track lung abnormality, however due to multiple scattering this method tracks the air volume change only on the lung surface. The aim of this phantom-based study is to assess the pulmonary edema in lungs using multiple scattering of ultrasonic waves in MHz range. A 128-element array transducer is used to form an Inter-elements Response Matrix (IRM) by firing a 5.2 MHz broadband pulse, element by element and receiving the backscattered signal by the whole array. The distribution of singular values of IRM in frequency domain is then analyzed for polymer sponges (n = 4) with different Water Volume Content (WVC) ranging from 5 to 25 ml mimicking pulmonary edema. The results of Random Matrix Theory are used to show that the mode of the Probability Density Function of singular values increases as the WVC increases in sponge indicating the change in scattering regime. The B-mode image of sponges with different WVC did not represent neither A nor B lines, however our results shows that there is a correlation between the WVC in sponge and the strongest singular value of IRM (R = 0.89) indicating the potential of using quantitative ultrasound in lung for early diagnosis of COVID-19.