Yoojin Lee1, Yeji Han, HyunWook Park. 1. Department of Electrical Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea.
Abstract
PURPOSE: To significantly reduce the background phase effects, especially at the air-tissue interface, and to enhance the desirable local structures of veins in susceptibility-weighted imaging. METHODS: In the proposed reconstruction method called Magnitude of Complex Filtering, a complex-valued magnetic resonance image is acquired using a flow-compensated high-resolution 3D gradient-echo sequence and the magnitude of the complex-valued image is set to 1 so that the phase information, which contains details of the local susceptibility, is emphasized. Then, the nonlinear filter of the Magnitude of Complex Filtering method is applied to the complex-valued image with a constant magnitude. This filter utilizes the magnitude of the low-pass and high-pass filtered complex data to selectively reduce the background phase effects while enhancing the local structures. The filter output is then processed to generate a susceptibility-weighted image. RESULTS: Compared with the conventional susceptibility-weighted images generated by a homodyne high-pass filter, the susceptibility-weighted images from the proposed Magnitude of Complex Filtering method show significant improvement; the undesirable artifacts at the air-tissue interface regions and the brain boundaries are significantly reduced, while the contrast of the local structures of veins is enhanced. CONCLUSION: The Magnitude of Complex Filtering method successfully reduced most background phase effects without requiring additional processing or scan time.
PURPOSE: To significantly reduce the background phase effects, especially at the air-tissue interface, and to enhance the desirable local structures of veins in susceptibility-weighted imaging. METHODS: In the proposed reconstruction method called Magnitude of Complex Filtering, a complex-valued magnetic resonance image is acquired using a flow-compensated high-resolution 3D gradient-echo sequence and the magnitude of the complex-valued image is set to 1 so that the phase information, which contains details of the local susceptibility, is emphasized. Then, the nonlinear filter of the Magnitude of Complex Filtering method is applied to the complex-valued image with a constant magnitude. This filter utilizes the magnitude of the low-pass and high-pass filtered complex data to selectively reduce the background phase effects while enhancing the local structures. The filter output is then processed to generate a susceptibility-weighted image. RESULTS: Compared with the conventional susceptibility-weighted images generated by a homodyne high-pass filter, the susceptibility-weighted images from the proposed Magnitude of Complex Filtering method show significant improvement; the undesirable artifacts at the air-tissue interface regions and the brain boundaries are significantly reduced, while the contrast of the local structures of veins is enhanced. CONCLUSION: The Magnitude of Complex Filtering method successfully reduced most background phase effects without requiring additional processing or scan time.