posters 5th Asia-Pacific NMR Symposium 2013

Optimization of phase processing for magnetic resonance venography (#220)

Phillip G. D. Ward 1 2 , Parnesh Raniga 2 3 , Amanda C. L. Ng 2 4 5 , David L. Dowe 6 , Gary F. Egan 2 7 , David G. Barnes 1 2 4 5
  1. Faculty of Information Technology, Monash University, Clayton, Vic, Australia
  2. Monash Biomedical Imaging, Monash University, Clayton, Vic, Australia
  3. The Australian e-Health Research Centre-BioMedIA, CSIRO Preventative Health National Research Flagship ICTC, Herston, QLD, Australia
  4. Monash e-Research Centre, Monash University, Clayton, Vic, Australia
  5. VLSCI Life Sciences Computation Centre, Carlton, Vic, Australia
  6. Computer Science and Software Engineering, Clayton School of I.T., Monash University, Clayton, Vic, Australia
  7. School of Psychology and Psychiatry, Monash University, Clayton, Vic, Australia

Purpose: To improve the performance of magnetic resonance imaging (MRI) based venous vessel segmentation algorithms by examining the impact of different imaging parameters, reconstruction techniques and post-processing techniques. Background: The deoxygenated haemoglobin in venous blood has a distinct magnetic susceptibility, which manifests in the phase of MRI data. The phase-contrast is sensitive to sequence parameters and reconstruction. Post-processing techniques are used to enhance the contrast prior to segmenting the vasculature. Method: Each combination of MRI sequence, reconstruction technique, and post-processing algorithm was investigated. Sequences included single echo and dual echo. Reconstruction techniques included GRAPPA1  and CSense2 . Post-processing techniques included high-pass filtering, rod-like filters3 , susceptibility-weighted imaging4  and vesselness filters5 . Each unique combination of sequence, reconstruction technique and post-processing technique (a pipeline) was tuned (reconstruction and post-processing parameters) to optimise performance and the resultant image displayed as a minimum intensity projection (mIP). The performance of each pipeline, for a small set of vessels with different orientations and diameters, was visually assessed and compared objectively using the signal-to-noise ratio in the mIP visualisation. Results: The selection of pipeline components had a strong effect upon vessel appearance. Vessel visibility was dependent on vessel size, orientation and location in addition to being sensitive to pipeline components. No strong relationship between number of tuneable parameters and vessel contrast was found. Dual echo sequences consistently outperformed single echo sequences. Conclusion: This investigation provided a guide for component selection when imaging cerebro-vasculature and suggested optimal contrast was dependent upon the size, orientation and location of the target vessels.

  1. M. A. Griswold, P. M. Jakob, R. M. Heidemann, M. Nittka, V. Jellus, J. Wang, B. Kiefer, and A. Haase. Generalized autocalibrating partially parallel acquisitions (grappa). Magn Reson Med, 47(6):1202–1210, Jun 2002.
  2. Z. Chen, L. A. Johnston, D. H. Kwon, S. H. Oh, Z.-H. Cho, and G. F. Egan. An optimised framework for reconstructing and processing mr phase images. Neuroimage, 49(2):1289–300, Jan 2010.
  3. Z. Jin, L. Xia, M. Lou, M. Zhang, and Y. P. Du. Mr venography of the brain with enhanced vessel contrast using image-domain high-pass filtering of the susceptibility phase shift. J Magn Reson Imaging, 34(5):1218–25, Nov 2011.
  4. E. M. Haacke, Y. Xu, Y.-C. N. Cheng, and J. R. Reichenbach. Susceptibility weighted imaging (swi). Magn Reson Med, 52(3):612–8, Sep 2004.
  5. A. Frangi, W. Niessen, K. Vincken, and M. Viergever. Multiscale vessel enhancement filtering. Medical Image Computing and Computer-Assisted Intervention - Miccai’98, 1496:130–137, 1998.