Vessel enhancement filtering in three-dimensional MR angiography

Area of Science:

• Medical Imaging
• Image Processing
• Neuroimaging

Background:

• Small vessels in magnetic resonance angiography (MRA) often exhibit low signal intensity.
• Low signal intensity can render small vessels invisible on maximum-intensity-projection (MIP) images, especially when below background tissue signal levels.

Purpose of the Study:

• To develop and evaluate a novel nonlinear second-difference spatial filtering technique for enhancing small vessel visibility in MRA.
• To compare the performance of the proposed nonlinear filters against a linear Laplacian filter and existing magnetization transfer contrast (MTC) techniques.

Main Methods:

• Development of two similar nonlinear second-difference spatial filtering techniques.
• Application of filters to three-dimensional time-of-flight (3D TOF) MR angiographic data of intracranial vessels.
• Comparison of nonlinear filter results with a linear Laplacian filter and MTC techniques.

Main Results:

• Nonlinear filtering significantly enhances small vessel details and suppresses noise and background tissue.
• Both nonlinear filtering and MTC techniques provide similar improvements in small vessel visibility and apparent continuity.
• Quantitative analysis shows a substantially greater improvement in contrast-to-noise ratio with nonlinear filters compared to the Laplacian filter.

Conclusions:

• Nonlinear second-difference spatial filtering is an effective method for improving small vessel visualization in MRA.
• The developed nonlinear filters offer superior performance in enhancing contrast-to-noise ratio for small vessels compared to linear filtering methods.
• Nonlinear filtering presents a promising alternative or adjunct to MTC for improving MRA diagnostic quality.