Blind-Label Subwavelength Ultrasound Imaging

 Ultrasound allows non-invasive, non-radiative imaging with sub-millimeter resolution, with broad applications in sensing, communication, biomedical diagnosis and non-destructive testing. In this talk, I will present our work on combining acoustic random scattering and computational imaging for improving the resolution of ultrasound imaging. Similar to all the other wave-based imaging modalities, ultrasound imaging suffers from diffraction limit, which is a trade-off between imaging resolution and wave penetration depth. Most existing acoustic subwavelength imaging technologies addressing this limit require controlled “labels,” i.e., metamaterials or contrast agents, to be deposited close to the objects and to either remain static or be tracked precisely during imaging, restricting their practical applications. We propose a “blind-label” approach. The blind labels are randomly distributed subwavelength microstructures or microparticles. The originally evanescent components in the scattered waves from the object are first converted to propagating components and then extracted by computational algorithms. Compared to conventional ultrasound imaging systems, our approach achieves an order of magnitude improvement in the imaging resolution without sacrificing the wave penetration depth. Our “blind-label” approach relaxes the restrictions of existing acoustic subwavelength imaging technologies that rely on controlled labels, thereby substantially enhancing the practicality of acoustic subwavelength imaging in real-world applications