학술논문

Lesion Imaging and Target Detection in Complex Heterogeneous Media: Application to Lung Nodules
Document Type
Conference
Source
2019 IEEE International Ultrasonics Symposium (IUS) Ultrasonics Symposium (IUS), 2019 IEEE International. :959-962 Oct, 2019
Subject
Bioengineering
Components, Circuits, Devices and Systems
Engineered Materials, Dielectrics and Plasmas
Signal Processing and Analysis
Lesions
Scattering
Media
Lung
Imaging
Ultrasonic imaging
Surgery
multiple scattering
lesion detection
ultrasound imaging
random media
lung nodule
lung cancer
Language
ISSN
1948-5727
Abstract
Ultrasound imaging in strongly heterogeneous media is challenging due to the aberrations caused by multiple scattering. The linearity between distance and time is lost making echolocation-based imaging elusive. In this study, we present a method for the imaging of anechoic lesions and targets in highly scattering media taking advantage of ultrasound multiple scattering. First, in silico, heterogeneous media are simulated with 10% area fraction of plastic scatterers in water. Circular lesion (areas with no scatterers) with diameter 8mm are placed at a depth of 10mm. Experimentally, a petroleum jelly lesion is placed in a sponge partially saturated with water, and in a pig lung ex vivo. Inter-elements Response Matrix are acquired using a tranducer array. The backscattered intensity is calculated and split into its coherent and incoherent parts. From the incoherent intensity, a diffusion map is generated and combined with a depression detection algorithm to predict the presence, location and size of the lesions.