Clear delineation of optic radiation and very small vessels using phase difference enhanced imaging (PADRE) Poster No.: C-2459 Congress: ECR 2010 Type: Scientific Exhibit Topic: Neuro Authors: T. Yoneda, H. Arimura, T. Takemaru, Y. Hiai; Kumamoto/JP Keywords: phase, susceptibility, MRI DOI: 10.1594/ecr2010/C-2459 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. www.myesr.org Page 1 of 14
Purpose It is important to delineate various nerve tracts in the brain image. Diffusion Weighted Imaging (DWI) and tensor imaging as a application of it are one of the excellent method to delineate them. However, spatial resolution of DWI is relatively low because of technical difficulties to detect them. Therefore it is difficult to detect thin nerve tract on the DWIimage and expected to develop new imaging theory and method. Recently, many study reports that phase information of MRI potentially possibility delineate small structure in the brain such as very small vessels and nerve tracts. Especially, susceptibility weighted imaging (SWI) can beautifully delineate the vein (vessel) in the brain by using phase information [ref. 1, 2]. This is the one of the most excellent result of phase imaging in MRI. In this study we develop new technique "Phase Difference Enhanced Imaging (PADRE)" in order to enhance various tissues in the brain such as OR and vessels in the brain MR-image. This technique selectively enhances the phase (difference) of objective tissue locating even in the white matter in which contrast of different tissues is hardly reconstructed. PADRE is to clearly delineate OR by adjusting several parameters. In this study contrast of tissues (OR and vessels) was measured and fixed such parameters. Additionally, we show an ability of PADRE to illuminate very small vessels. Methods and Materials In this study, we have taken images with 2 sequences and scanning parameters. In order to have much information of phase (differences), we need long echo time (TE) (Fig. 1 on page 3) [ref. 3]. Long TE, however, results in distortion of magnitude image contrast due to the T2* effect. Therefore, whenever we set long TE, we have to take much longer repetition time (TR) for having sufficient signal intensity. But such longer TR automatically means longer scanning time which should be evitable for clinical use. We frequently meet such kind dilemma when we want to use phase information. One of the solution to longer scanning time is using shifted-echo sequence in which TE is longer than TR. In this study we used 3D Principle of Echo Shifting with a Train Observations (PRESTO) as shifted-echo sequence. Since PRESTO is running on Philips MRI system, we have chosen Philips's 3.0 T MRI (Achieve 3.0T, Philips Medical Systems, The Netherlands). Page 2 of 14
3D-GE (FFE) acquisition was also used to see a dependence of the contrast of PADREimage. We arranged scanning parameters so as to fix scanning time (10min); TR/TE = 32/51 ms, SENSE factor = 1, FOV = 230x230 mm, Matrix = 512x512, Slice Thickness = 1 mm for PRESTO and TR/TE = 38/30 ms, SENSE factor = 1, FOV = 230x230 mm, Matrix = 512x512, Slice Thickness = 1mm for 3D-GE. Each phase image was enhanced by PADRE and measured the contrast between OR and parenchyma, and also the deep veins as small vessels. Enhancement steps were as below (Fig. 2 on page 4); 1. Low-pass filter with fixed filter size was operated on k-space data and Fourier transform was carried out to create low-pass filtered complex image (data). 2. High-pass filtered complex image was obtained from dividing original image by low-pass filtered image. (We henceforth call filtered image as high-pass filtered complex image.) 3. Filtered phase image (data) was obtained from filtered complex image. We just call this "phase" as shorthand notation. 4. Phase was selected to suit objective tissue [ref. 4]. 5. Selected phase was enhanced by depressing the signal value on the original magnitude. Enhancement achieves by using enhancing function which was designed for adjustable selection phase and enhancement of them. Region of interest (ROI) was located on objective tissue (OR) and parenchyma with fixed area. Obtained data is used to determine the contrast of OR. Here, contrast is to be defined as difference between lowest value of signal in a profile curve of OR and mean value of parenchyma in the ROI. Images for this section: Page 3 of 14
Fig. 1: Phase value proportional TE and local magnetic field (magnetic flux density) with susceptibility #. Contrast in phase image should be defined local magnetic field difference between different tissue. Page 4 of 14
Fig. 2: Phase (difference) value normally takes very small number, about several ppm. Therefore we must enhance it when we put phase information in the magnitude image. Page 5 of 14
Results Original PRESTO and 3D-GE images could slightly show OR tracks due to the T2* effect (Fig. 1 on page 6). From the theoretical point of view, it is expected that PRESTO which has longer TE than 3D-GE sequence could show higher contrast on "phase image". But, in practice, background plays important role to decide enhanced contrast because of its definition, and therefore enhanced contrast do not always increase proportional to TE. Actually, OR is to be seen well in the 3D-GE image compared with PRESTO image (Fig. 1 on page 6). It is also true in the enhanced images. We can see clear contrast of OR on PADRE image reconstructed from 3D-GE image (Fig. 2 on page 7). All acquired images were filtered with 128x128 filter size. Although PADRE-image reconstructed from PRESTO can also delineate OR, its apparent contrast is slightly lower than 3D-GE's one (Fig. 3 on page 8). In order to numerically measure the contrast of OR and parenchyma, we set profile line and ROI to measure the signal of parenchyma (see Fig. 4 on page 9). Obtained data was calculated and plotted on Fig. xx. Imaging parameters were changed so as to enhance for (1) all phase of OR, (2) large phase value of OR, (3) small phase value. PADRE-image very well delineate OR (red arrows in Fig. 2 on page 7, 3 on page 8) compared with magnitude image of PRESTO which is T2* image. Actually, all PADRE-images gave higher contrast value than PRESTO-image (Fig. 5 on page 10). As is in Fig. 5 on page 10, contrast of PADRE-image type 2 and 3 especially show high contrast. But, in those images, background noise looks be enlarged by enhancement and it becomes to be visually difficult to distinguish between noise and signal of tissue in them. By enhancing the phase of vessel, we could have vessel enhanced PADRE-image which is corresponding to SWI-image (Fig. 6 on page 11). In that image, we can see very high contrasted vessels, especially, fine deep veins is very well delineate. Same as OR enhanced images, images type 2 and 3 looked to be noisy one because PADRE simultaneously enhanced phase noise with objective phase. In fact, signal to noise ratio (SNR) of those images shows lower value (Fig. 7 on page 12). Images for this section: Page 6 of 14
Fig. 1: Both of these magnitude images show low signal intensity of OR due to T2* effect. Page 7 of 14
Fig. 2: All of images show high contrast of OR which is pointed by red arrow. There images were reconstructed by different parameter sets for enhancement. Page 8 of 14
Fig. 3: Enhanced Images generated from PRESTO images. These images also show high contrast OR. Additionally, cerebral cortex can be illuminated by PADRE technique. Page 9 of 14
Fig. 4: Contrast was defined by lowest signal of transverse profile curve of OR (red line). We obtained signal of parenchyma defined by mean value in the ROI (green circle). Page 10 of 14
Fig. 5: All of reconstructed images show high contrast of OR compared with magnitude image. Page 11 of 14
Fig. 6: Even small vessels were delineated by PADRE technique. Deep veins (enlarged in the box) can be seen very well. This vessel imnages were generated by minimam intensity projection over 12 slices. Page 12 of 14
Fig. 7: Comparison of SNR in the vessel enhanced images. Parameters to generate strong contrast simultaneously enhances signal and phase noise. Therefore, SNR decreases in the case of strong enhancement (image #3). Page 13 of 14
Conclusion We had clearly delineated OR and small vessels by using PADRE technique with various imaging parameters. Although all parameter sets lead higher contrast of OR compared with magnitude image, phase noise was also enhanced on the PADRE image so that it is not always good to observe tissues. This situation was also true in the case of vessel enhanced images. Actually, SNR of them varied depending on parameters set. But vessel enhanced PADRE-images shows very clear contrast of very fine vessel. References 1. Haacke E.M., Xu Y., Cheng Y.N., Reichenbach J.R., 2004. SusceptibilityWeighted Imaging. Magn. Reson. Med. 52, 612-618. 2. Reichenbach J.R., Barth M., Haacke E.M., Klarh ofer M., Kaiser W.A., Moser E., 2000. High-resolution MR venography at 3.0 Tesla. J. Comput. Assist. Tomogr. 24 (6), 949-957. 3. E. Mark Haacke, Robert W. Brown, Michael R. Thompson, and Ramesh Venkatesan, Magnetic Resonance Imaging: Physical Principles and Sequence Design. 1999, Wiley-Liss, USA. 4. Zhong K., Leupold J., von Elverfeldt D., Speck O., 2008. The molecular basis for gray and white matter contrast in phase imaging. NeuroImage 40, 1561-1566. Personal Information Tetsuya YONEDA, PhD, School of Health Sciences, Kumamoto University, Kuhonji 4-24-1, Kumamoto Japan. E-mail: tyoneda@kumamoto-u.ac.jp Page 14 of 14