4-Dimensional lung ultrasound imaging for lung surface motion estimation: an in vivo feasibility study

Lung ultrasound (LUS) is nowadays widely spread in the clinical world. LUS analysis generally does not rely on the dynamics of LUS videos, being thus limited to the evaluation of specific LUS frames. Only rare LUS applications rely on dynamics, but without quantifying lung surface motion (LSM). For example, the presence of pneumothorax correlates with the absence of pleural sliding. Moreover, pleural sliding is assessed through the analysis of B-Mode LUS videos (3D data), which cannot guarantee PS to reliably represent LSM. Hence, in this study, for the first time, we evaluated the possibility to estimate LSM with 4-Dimensional (4D) LUS. Specifically, we assessed whether a velocity model based on functional ultrasound (fUS) applied to 4D LUS data can be used to differentiate normal breathing [NBr] (10-20 breaths per min [BPM]) and fast breathing [FBr] (40-60 BPM). Results showed how mean velocity for FBr equals 13.17 mm/s, corresponding to 1.88 times the mean velocity for NBr (7.01 mm/s), thus showing a variation consistent with the BPM change among NBr and FBr. Moreover, estimated velocities are within physiological range, proving, for the first time, the possibility to estimate LSM with 4D LUS.