Simulating lung ultrasound imaging with body wall and alveolar anatomy

The majority of diagnostic ultrasound imaging is performed in regions with modest variations in the impedance (∼5%) which results in relatively small reflections from soft tissue interfaces. Lung ultrasound imaging is unique because it attempts to image an air-filled organ which is almost totally reflective. Thus the interpretation of lung ultrasound imaging depends on the analysis of the scattering, multiple scattering, and reverberation physics occurring at or near the complex soft-tissue/air interfaces. Even though this physics determines the image content, its relationship to the image, and its dependence on the diseased states of the lung are poorly understood. To establish a link between body wall and lung anatomy and the resulting ultrasound image we use a custom ultrasound simulation tool (Fullwave) in conjunction with acoustical maps of human body derived from the visible human project and maps of alveolar structure derived from lung histology. The experimentally validated simulations of ultrasound propagation includes attenuation, dispersion, nonlinearity, and the multiple interactions of sound at soft tissue/air interfaces. Ultrasound images are then simulated based on the first principles of propagation and reflection. We demonstrate that key features of clinical lung ultrasound imaging such as A- and B-lines can be observed. Anatomical and diagnostically relevant parameters, such as fluid percentage, alevolar density, are pleural thickness are varied to determine their effect on the final image.