This project is aimed to develop methods to improve image quality in ultrasonic
array imaging using recent model based reconstruction methods for realistic array
layouts. The drawback of the traditional delay-and-sum (DAS) based array focusing
method---w hich requires the array pitch to be less than half the acoustic wavelength
to avoid grating lobes and the element sizes to be small since they are regarded
as point sources by the DAS method---is relaxed by this new approach; thus enabling
a more flexibl e design of the array layouts. The objective of the project is to further
develop and adapt these methods to medical, sonar, sparse arrays, 2D arrays, as well
as undersampled linear arrays. Issues that also will be considered are design of robust
metho ds for treating model errors, design of optimal input signals for optimizing
reconstruction performance, as well as development of methods for designing focusing
methods specifically customized to these model based reconstruction algorithms.
Furthermor e, since the methods can be computational demanding the project will also
include studies in efficient sparse representations as well as development of parallel
implementations for large scale problems.