Author

Kevin Ayer

Abstract

Finite amplitude sound propagation undergoes nonlinear distortion due to continuous path interaction with the propagation medium. This distortion tends to defocus the beam causing significant lateral and contrast resolution degradation. Fundamental understanding of this interaction requires development of computational models that accurately predict the nonlinear interaction - development of media-borne harmonics - as well as produce an ultrasound image - introduction of transducer effects, interface transitions, and innovative image processing to extract harmonics. Most computational models of ultrasound propagation assume axial symmetry for computational expediency. Two notable exceptions are the K-Z-K and NLP models. A new endto- end model, NUPROP, is introduced that also incorporates non-axially symmetric geometries and simplified transducer responses to accurately predict ultrasound RF signals for image reconstruction. Nonlinearities are modeled using either the Fubini solution or Burgers' Equation coupled with angular spectrum propagation or Lommel formulation, appropriately masked by the transducer frequency response. Comparative analyses are performed on NUPROP results with high correlation with the literature. Parameter sensitivity analyses are performed to determine harmonic signal characteristics as a function of propagation distance. A-line and B-scan images are produced.

Library of Congress Subject Headings

Ultrasonic imaging--Mathematical models; Image analysis; Nonlinear waves; Image processing

Publication Date

1-1-2000

Document Type

Dissertation

Student Type

Graduate

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Rao, N.

Advisor/Committee Member

Wilcox, Theodore

Advisor/Committee Member

Schott, John

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: R857.U48 A93 2000

Campus

RIT – Main Campus

Share

COinS