Edge-guided phase unwrapping for SWI

Scientist
Hemanth Thayyullathil

Abstract

Accurate quantification of phase is essential in many contexts in MRI, such as in susceptibility weighted imaging, flow imaging and for the calculation of Bo field-maps. A phase image measured as the inverse tangent of the ratio of imaginary and real part of the complex image comprises of artificial discontinuities called ‘phase wraps’. Classic algorithms proposed in the past are either computationally very complex for phase unwrapping or smoothen subtle features in phase images. In an attempt to formulate a computationally efficient algorithm that performs phase unwrapping without smoothening subtle features, we have formulated a new method that alters the phase values only at regions of wraps, which are automatically identified by linked edges present in the wrapped phase image by performing edge detection and linking.