Quantitative evaluation of phase processing approaches in susceptibility weighted imaging

Abstract

Susceptibility weighted imaging (SWI) takes advantage of the local variation in susceptibility between different tissues to enable highly detailed visualization of the cerebral venous system and sensitive detection of intracranial hemorrhages. Thus, it has been increasingly used in magnetic resonance imaging studies of traumatic brain injury as well as other intracranial pathologies.In SWI, magnitude information is combined with phase information to enhance the susceptibility induced image contrast. Because of global susceptibility variations across the image, the rate of phase accumulation varies widely across the image resulting in phase wrapping artifacts that interfere with the local assessment of phase variation. Homodyne filtering is a common approach to eliminate this global phase variation. However, filter size requires careful selection in order to preserve image contrast and avoid errors resulting from residual phase wraps. An alternative approach is to apply phase unwrapping prior to high pass filtering. A suitable phase unwrapping algorithm guarantees no residual phase wraps but additional computational steps are required. In this work, we quantitatively evaluate these two phase processing approaches on both simulated and real data using different filters and cutoff frequencies. Our analysis leads to an improved understanding of the relationship between phase wraps, susceptibility effects, and acquisition parameters. Although homodyne filtering approaches are faster and more straightforward, phase unwrapping approaches perform more accurately in a wider variety of acquisition scenarios.

Quantitative assessment of susceptibility‐weighted imaging processing methods

“…Image artifacts were analyzed and parameters for phase processing were optimized. Preliminary results of this work have been previously reported in a conference paper .…”

“…: [7] However, rounding may not work well when the difference value w u ði;jÞ-w r ði;jÞ 2p is halfway between integers. In such situations, additional phase processing may need to be performed.…”

“…All filtering methods were performed in the frequency domain. Other than filtering, spatial smoothing has also been proposed to remove background gradients after phase unwrapping . To include this method in the comparison, spatial smoothing following phase unwrapping was implemented using a boxcar averaging operator in the image domain, where the smoothing factor is defined as the width of the boxcar averaging operator .…”

Rapid, high-resolution, whole-brain, susceptibility-based MRI of multiple sclerosis

“…For the phase images, an automatic analytical phase unwrapping was employed as described in (21). Following phase unwrapping, large background gradients were removed by performing Gaussian filtering with a filter size of 32 pixels and full width at half maximum of 8 pixels (22). …”

Characterizing the spatial distribution of microhemorrhages resulting from Traumatic Brain Injury (TBI)

“…SWI images were processed for phase enhancement using in-house developed SWI processing software and included three steps. 10,11 First, a Fourier-based phase unwrapping algorithm was performed, followed by one minus Gaussian filtering to remove unwanted artifacts of the raw phase signal obtained directly from the scanner. Then a normalized phase mask was used with values between zero and one in areas of negative phase, and with a value of one elsewhere.…”