Advanced

AJNR - American Journal of Neuroradiology

This Article Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Schmitz, B. L. Articles by Grön, G. Search for Related Content PubMed PubMed Citation Articles by Schmitz, B. L. Articles by Grön, G.

Advantages and Pitfalls in 3T MR Brain Imaging: A Pictorial Review

Bernd L. Schmitz^a, Andrik J. Aschoff^a, Martin H.K. Hoffmann^a and Georg Grön^b

^a Department of Radiology, University Hospitals at Ulm, Ulm, Germany
^b Department of Psychiatry, University Hospitals at Ulm, Ulm, Germany

View larger version (10K): [in a new window]   FIG 1. Comparison of the theoretical signal intensity increase when increasing NEX (gray line) or field strength (B0, black line). To reach the same signal intensity one can get from the doubling field strength, the acquisition time is 4 times longer (circles). Numbers along the x axis refer to levels of NEX and B0, respectively.

View larger version (10K): [in a new window]   FIG 2. Theoretic increase in relative signal intensity for NEX and field strengths B0 (gray lines) as in Fig 1. Note the relative increase in SAR related to field strength (black line). Although signal intensity linearly increases with field strengths, SAR increases with the square of field strengths. Values on the y axis are arbitrary units solely to demonstrate the relationship between the increases of different parameters in one graph.

View larger version (80K): [in a new window]   FIG 3. Chemical shift artifacts at different bandwidths at 1.5 (upper row: TR, 172 ms; TE, 9 ms; field of view [FOV], 220 x 220 mm2; matrix, 256 x 192) and 3T (lower row: TR, 206 ms; TE, 10 ms; section thickness, 5 mm; FOV, 220 x 220 mm2; matrix, 256 x 192). Bandwidth increases from left to right (60, 120, 240, 480 Hz/pixel), resulting in chemical shift of 7.4, 3.7, 1.9, and 0.9 pixels, respectively, at 3T, and half of these values at 1.5T. Note the double line of the occipital subcutaneous tissue at lower bandwidths (between white arrows). Of note also is the increasing noise at higher bandwidths.

View larger version (112K): [in a new window]   FIG 4. T2 turbo spin-echo (left) and hyperecho (right) of the same section position and with all other parameters kept equal (TR, 3810 ms; TE, 78 ms; matrix, 832 x 416; FOV, 220 x 220 mm2; echo train length [ETL] 9; section thickness, 5 mm; bandwidth, 145 Hz/pixel; flip angle, 120°), demonstrating that there is no difference in image contrast while SAR was significantly reduced.

View larger version (107K): [in a new window]   FIG 5. Gradient echo T1 (left: TR, 311 ms; TE, 2.5 ms; section thickness, 5 mm; matrix, 512 x 256; FOV, 220 x 220 mm2; bandwidth, 465 Hz/pixel; flip angle, 90°) compared with T1 spin-echo (right: TR, 700 ms; TE, 10 ms; section thickness, 5 mm; matrix, 256 x 192; FOV, 220 x 220 mm2; bandwidth, 200 Hz/pixel; flip angle, 90°) in the same subject at 3T, which indicates higher contrast in gradient echo than spin-echo sequences at 3T.

View larger version (101K): [in a new window]   FIG 6. Left, T1 spin-echo image at 1.5T (TR, 600 ms; TE, 14 ms; bandwidth, 90 Hz/pixel; section thickness, 5 mm; matrix, 256 x 192; FOV, 220 x 220 mm2; flip angle, 90°). Right, T1 spin-echo at 3T (TR, 700 ms; TE, 10 ms; section thickness, 5 mm; 19 sections; bandwidth, 200 Hz/pixel; matrix, 256 x 192; FOV, 220 x 220 mm2; flip angle, 90°), which is indicative of the reduced gray-to-white matter contrast at higher fields.

View larger version (137K): [in a new window]   FIG 7. Turbo inversion recovery (TIR) sequence at 1.5T (left: TR, 8770 ms; TE, 92 ms; TI, 300; matrix, 512 x 256; section thickness, 3 mm; FOV, 220 x 220 mm2; bandwidth, 130 Hz/pixel) and 3T (right: TR, 8890 ms; TE, 95 ms; TI, 300; all other parameters equal), which demonstrates clear depiction of gray and white matter at both field strengths.

View larger version (108K): [in a new window]   FIG 8. Same axial section position of a TIR (TR, 7670 ms; TE, 68 ms; TI, 300; ETL, 15; matrix, 448 x 224; section thickness, 4.5 mm; bandwidth, 130 Hz/pixel; flip angle, 150°) and T1 sequence (magnetization-preparation rapid gradient echo TR, 1880 ms; TE, 3.7 ms; matrix, 256 x 230; FOV, 256 x 256 mm2; section thickness, 5 mm, reconstructed from 1.0 mm primarily; bandwidth, 160 Hz/pixel; flip angle, 8°) after contrast agent in a patient with a contrast-enhancing lesion. Note the absent contrast enhancement in the TIR image (white arrows in the T1 image).

View larger version (75K): [in a new window]   FIG 9. Same section position with spin-echo T1-weighted sequences at 1.5T (upper row: TR, 700 ms; TE, 10 ms; section thickness, 5 mm; matrix, 256 x 192; FOV, 220 x 220 mm2; bandwidth, 200 Hz/pixel) and 3T (lower row: same imaging parameters). Flip angles decreasing from left to right 130°, 110°, 90°, 70°, 50°. The lowest flip angle shows the best gray-to-white matter contrast. The effect is stronger at higher field.

View larger version (136K): [in a new window]   FIG 10. Patient with multiple cavernomas inducing large susceptibility artifacts in T2* imaging at 3T, which is indicative of the high sensitivity for susceptibility effects at 3T (TR, 800 ms; TE, 26 ms; flip angle, 20°; bandwidth, 80 Hz/pixel; section thickness, 5 mm; matrix, 320 x 320; FOV, 220 x 220 mm2).

View larger version (17K): [in a new window]   FIG 11. Empirical data from a blocked finger tapping fMRI study at 1.5T and 3T comparing BOLD signal intensity strength in terms of statistical effect sizes (beta values from a linear regression equation; units are arbitrary) at different TEs (54 volumes scanned, a stimulus onset asynchrony of 16 volumes, and an epoch length of 8 volumes. Beta values were computed from scaled episeries within the general linear model by using SPM 99). There is a more than linear increase in BOLD signal intensity with higher field strength.

View larger version (111K): [in a new window]   FIG 12. Comparison of susceptibility artifacts at the skull base for 1.5T and 3T at different TEs, keeping all other parameters equal (TR, 4500 ms; isotropic voxel size, 3.3 mm; 48 sections; bandwidth, 2170 Hz/pixel). Artifacts are larger at any TE for 3T and increase with rising TE for both field strengths. At low TE, however, skull base artifacts are tolerable at 3T.

View larger version (54K): [in a new window]   FIG 13. Same section position in axial EPI images of the same subject (TR, 4090 ms; TE, 54 ms; isotropic voxel size, 3.3 mm; 36 sections at 3.0T). Decrease of susceptibility effects with increasing bandwidth (in Hz/pixel) due to shorter echo spacing and duration of EPI readout. At low bandwidths (752 Hz/pixel, in this example), distortion effects from the frontal sinus are clearly obvious; however, very high bandwidths (4882 Hz/pixel) do not yield a significant further reduction of distortions that are already achieved with bandwidths at 2520 Hz/pixel.

View larger version (114K): [in a new window]   FIG 14. Comparison of EPI with different voxel size regarding susceptibility artifacts at 3T. (A) Thinner sections (matrix, 64 x 64; TR 4500 ms; TE, 38 ms; bandwidth, 2170 Hz/pixel; FOV, 190 x 190 mm2; 53 sections; section thickness, 2 and 5 mm) and (B) smaller in-plane voxel sizes (matrix, 128 x 128; all other parameters same as in A) result in fewer susceptibility artifacts, especially in frontobasal regions.

View larger version (102K): [in a new window]   FIG 15. Maximum intensity projection of a time-of-flight angiography (TR, 28 ms; TE, 4.92 ms; matrix, 704 x 576; FOV, 163 x 200 mm2; 92 sections; section thickness, 0.75 mm; bandwidth, 105 Hz/pixel; flip angle, 25°) at 3T showing the clear depiction of even very small vessels.