Pulse Sequences

Anatomical brain scanning essentially relies on proton density, differences in T1 and T2 between regions (e.g. cortex vs. subcortical nuclei), and tissue type (white matter, grey matter and CSF). Brain images can be acquired using different pulse sequences. A 3.0 T MR system implements all the pulse sequences commonly applied in clinical practice, including SE (normal, fast 2D and 3D); GE (GRASS/fast GRASS 2D and 3D, SPGR/fast SPGR 2D and 3D); EPI (single shot/multishot 2D and 3D); and IR (STIR, FLAIR, normal, fast) using common imaging parameters like flow and respiratory compensation, cardiac and peripheral gating, graphic prescription, graphic saturation, fat/water saturation, variable bandwidth and asymmetric field of view.

Since the user interface and sequence parameters are the same as those employed with 1.5 T systems, at 3.0 T one should in theory obtain the same images with the advantage of high-field strength, i.e. a double signal intensity. In practice, much depends on tissue relaxation times (T1 and T2) and the specific absorption rate (SAR) [1-6]. Since relaxation times are field-dependent and affect image contrast, 1.5 T imaging sequences cannot be transferred to 3.0 T systems. In addition, the excessive RF energy that is deposited on tis sues at3.0T needs to be taken into account when adapting and optimizing clinical imaging protocols for use with such systems.

The technical parameters of the main sequences applied with 1.5 and 3.0 T systems are reported in Table 4.1. Higher-field systems afford greater resolution quality by allowing the acquisition of a larger number of slices (usually 25-27 vs. 18-20) of smaller thickness (4 mm vs. 5-6 mm), with narrower fields of view (20-22 vs. 24), broader matrices (512x512 vs. 256x256), and acquisition times similar to those of 1.5 T systems.

Lastly, two factors that depend on magnetic field strength need to be taken into consideration: the influence of the signal loss caused by the stronger susceptibility effects and the potentially diminished sensitivity of the radiofrequency coils at higher field intensity.

T1 Imaging

Since T1 relaxation times are longer at higher magnetic fields, entailing a reduction in the relative differences among different tissue types, it is generally difficult to obtain T1-weighted images with sufficient contrast when using a 3.0 T unit [2]. In fact, not only T1 increases, but the T1 values of different tissue types actually converge, giving rise to a narrower range of distribution. For this reason, optimization of the signal/ noise ratio (SNR) in the same sample using a higher field unit requires longer TR (and a smaller flip angle), resulting in longer scanning times. Using longer TR, the gain in signal intensity in unit of time afforded by the higher magnetic field is therefore lost. T1 saturation for a given TR is greater at 3.0 T, resulting in reduced signal gain. In other words, the SNR per unit of time would be optimized with the shortest possible TR/ T1. The possibility that the SNR gain connected with the high magnetic field may substantiallybe offset by a longer T1 highlights the importance of optimizing the imaging parameters [3].

The increased T1 is the main problem in 3.0 T brain imaging using SE T1-weighted sequences, which afford very unsatisfactory T1 contrast (Fig. 4.1) [7-10]. In-

Table 4.1. Technical parameters of the main sequences applied with 1.5 and 3.0 T systems

Sequence

TR (ms)

TE (ms)

Other parameters (TI, FA, ETL)

Slice thickness (mm)

No. of slices

FOV

Matrix

NEX

Examination time (min:s)

SET1

500

Min

_

6

16

24

256X192

1

1:23

1.5 T

SET1

500

Min

-

5

18

24

320X224

1

2:00

3.0 T

FGET1

225

Min

FA 75

4

20

24

512X256

1

1:57

3.0 T

IRT1

2,075

24

TI auto

4

11X2

24

256X256

1

2:49

3.0 T

TSE T2

3,000

88

ETL 12

6

21

24

256X224

2

2:24

1.5 T

TSE T2

4,850

81

ETL 15

4

25

22

448X320

2

2:39

3.0 T

TSE T2 HD

4,975

85

ETL 15

3

25

20

512X320

4

7:20

3.0 T

FLAIR

8,800

120

TI 2200

5

14X2

26

256X192

2

4:24

1.5 T

FLAIR

11,000

130

TI 2250

4

16X2

22

288X192

2

Images Grey And White Matter
Fig. 4.1. T1 imaging: comparison of the same SET1 sequence (TR 500,0.5 NEX, 1:00) acquired at 3.0 T (a) and 1.5T(b). Note the less satisfactory white/grey matter contrast in a
Fig. 4.2. T1 imaging with SE T1 (a), FLAIR T1 (b) and FGRE T1 (c)
Fig. 4.2. (Cont.) The poor white/grey matter contrast of the SE T1 sequence (a) required performance of the other two high-contrast sequences. Note the typical high signal of the larger arterial vessels in c (arrow)

deed, the reduction in T1 differences among different types of tissues entails a loss of contrast between white and grey matter. For T1 contrast, this has led to the application of sequences other than SE which yield high T1 contrast irrespective of field strength, i.e. fast GE T1 (spoiled gradient echo, SPGR, or MP-RAGE) and fast IR or fast FLAIR T1-weighted (Figs. 4.2-4.4) [7-9].

Fast SPGR allows thinner slices to be obtained in a shorter time and provides images suitable for reformatting in all three planes using a single volumetric acquisition of the whole brain, thus disclosing a larger number of lesions than can be depicted using SE T1-weighted sequences acquired in a single plane. In addition, large arterial vessels appear hyperintense on fast SPGR sequences, although this does not impair image interpretation.

In contrast, fast IR entails longer scanning times because chained acquisitions are required to image the whole brain; this can be a problem when multiple un-enhanced and contrast-enhanced studies, or different views, need to be performed.

Unlike IR Tl-weighted sequences, which are less satisfactory and take longer to perform, a typical FLAIR Tl-weighted study of the brain coupled with parallel imaging allows higher spatial resolution protocols to be applied in a shorter acquisition time.

Greater inflow contrast and enhanced background suppression make contrast-enhanced T1 imaging a definite advantage of 3.0 T systems. For instance, when studying primary and secondary brain tumours, administration of a both single- and multiple-dose gadolinium yields greater contrast between tumour and normal brain tissue and allows the detection of more metastases, and demonstrates different patterns of enhance

Fig. 4.4. High-definition images: comparison of a 3D FSPGR IR-Prep sequence (a, b) (TR 9.1, TE 4.0, FOV 170, slice thickness 2 mm, matrix 320 x 288) with an FSE-IR sequence (c, d) (TI250 ms, FOV 160, slice thickness 3 mm, matrix 512 x 256). Note the optimum anatomical depiction of the cortico-subcortical junction on both sequences

Fig. 4.4. High-definition images: comparison of a 3D FSPGR IR-Prep sequence (a, b) (TR 9.1, TE 4.0, FOV 170, slice thickness 2 mm, matrix 320 x 288) with an FSE-IR sequence (c, d) (TI250 ms, FOV 160, slice thickness 3 mm, matrix 512 x 256). Note the optimum anatomical depiction of the cortico-subcortical junction on both sequences

Microadenoma
Fig. 4.5. Unenhanced (a) and contrast-enhanced (b) hypophyseal microadenoma acquired with FSE-XL T1 (TR 600, TE 12, ETL 5, FOV 180 mm, slice thickness 2.5 mm/0.5, matrix 224 x 192, ZIP 512,29"/phase). PEI processing using Functool2 (c)

ment that may be useful to assess the degree of malignancy and to monitor response to therapy [11-13].

The greater sensitivity of 3.0 T contrast-enhanced investigations has also been demonstrated in multiple sclerosis, with a 21 % increase in the number ofenhanc-ing lesions detected, a 30% increase in enhancing lesion volume and a 10% increase in total lesion volume relative to scanning at 1.5 T [14].

Contrast-enhanced imaging also improves the evaluation of hypophyseal macro- and microadenomas (Figs. 4.5,4.6). 3.0 T MR venography is another valuable technique that provides increased spatial resolution, and thus more detailed information, in the same time of acquisition as a 1.5 T system [15].

However, contrast-enhanced T1 semeiotics requires adjustments in TR and TE and a reduction in the dose of contrast agent [3]. The usual dosage of 0.1 mmol/kg can be halved without affecting image quality. Using a standard dose, the contrast/noise ratio (CNR) is two and a half times greater than with 1.5 T systems [15]. Double or triple doses can result in meningeal contrast uptake, a common non-pathological finding at 3.0 T

which at lower field strengths may, however, mimic carcinomatosis or meningitis, and should thus be avoided.

To enhance sensitivity or highlight minimal changes in the blood brain barrier, a double dose of contrast agent may be employed in selected investigations of brain metastases or inflammatory disease [16, 17]. In such cases, greater lesion enhancement with respect to normal brain parenchyma results in improved sensitivity and resolution.

Fast GE T1, rather than standard SE sequences, are preferred for contrast-enhanced T1 imaging, as they offer better contrast in the same acquisition time. SE T1 sequences are, however, consistently improved by contrast administration (Fig. 4.7).

Fig. 4.5. (Cont.)

Fig. 4.7 a, b. Small left acoustic neurinoma: high-definition SE T1 sequence (matrix 512, thickness 1.5 mm). Unenhanced (a) and contrast-enhanced (b) images

Fig. 4.6. Hypophyseal macro-adenoma: dynamic contrast-enhanced image (unenhan-ced image in a)

Fig. 4.7 a, b. Small left acoustic neurinoma: high-definition SE T1 sequence (matrix 512, thickness 1.5 mm). Unenhanced (a) and contrast-enhanced (b) images

T2 Imaging

In T2 contrast imaging the advantages of high magnetic field strength are best exploited by performing fast SE sequences acquired using the same technique as with 1.5 T, imagers but with shorter TR and TE due to the shorter T2 relaxation times (Fig. 4.8). High spatial resolution FSE-weighted images are especially effective at 3.0 T as they provide more anatomical detail and better contrast due to the greater SNR, which allows to employ thinner slices and broader matrices (Figs. 4.9-4.13) [7-9, 18]. Even better images are obtained with inverted contrast FSE T2 sequences (Fig. 4.14).

In practice, the greater SNR may be attenuated by

tl* V"
Fig. 4.8. T2 imaging: comparison between the same FSE T2 sequence at 3.0 T (A, B zoom) and 1.5 T (C, D zoom). Anatomical detail and contrast resolution are greater in a and b
Fig. 4.9. High-definition FSE T2 image of the median sections of the brain

Fig. 4.10. High-definition FSE T2 image of the hippocampus

Fig. 4.10. High-definition FSE T2 image of the hippocampus

Fig. 4.11. High-definition FSE T2 image of the fifth cranial nerve in coronal (a), axial (b) and parasagittal (c) section the increased deposition of RF energy, which in turn requires long TR or pulse refocalization angles well below 180°. This lower SNR can, however, be compensated for by the higher field strength. In addition, the broader bandwidth, smaller echo space, shorter minimum TR and TE, and shorter FSE sequences employed at higher field strengths result in improved image quality, artefact reduction, and enhanced diagnostic performance in much shorter examination times, to the benefit of patients.

These sequences allow excellent visualization of the basal or mesencephalic nuclei, by virtue of their iron content, despite the intrinsically lower sensitivity of FSE to magnetic susceptibility in 3.0 T compared with 1.5 T systems [3].

Different signal features are also seen in brain haemorrhage, which at 3.0 T is characterized by greater hy-pointensity in the acute and early subacute phases [19]. SNR and signal intensity being roughly similar at both field intensities, correct dating of the haemorrhage is

Fig. 4.11. (Cont.)

usually feasible. Only in the acute phase can bleeding be erroneously construed as early subacute haemorrhage [19].

T2 images can also be acquired with GE (T2 quences like CISS (constructive interference in the steady state), which is particularly suited to investigations of the inner ear. 3D CISS offers high-resolution images of the inner ear and labyrinth with greater SNR than other T2-weighted sequences used to study this area (e.g. fast recovery FSE) (Fig. 4.15) [20].

Fig. 4.12 a, b. FSE T2 imaging. The excellent detail of this se quence allows visualization of the perivascular (Virchow-Rob in) spaces at the level of the bi-hemispheric cortico-subcortical junction. Axial (a) and coronal (b) views

Fig. 4.13. Multiple sclerosis: FSE T2 sequence

Fig. 4.12 a, b. FSE T2 imaging. The excellent detail of this se quence allows visualization of the perivascular (Virchow-Rob in) spaces at the level of the bi-hemispheric cortico-subcortical junction. Axial (a) and coronal (b) views

Fig. 4.13. Multiple sclerosis: FSE T2 sequence

Fig. 4.14. High-definition FSE T2 sequence (matrix 512x320, thickness 3 mm, ZIP 1024) acquired with (a-c) and without (d-f) inverted contrast: coronal view. Note the excellent anatomical detail of both hippocampi

Fig. 4.15a. High-definition study of the inner ear using a 3D FIESTA sequence (TR 4.3, TE 1.5, flip angle 50°, BW 62 kHz, FOV 170, slice thickness 0.6 mm, matrix 256X256, ZIP 512, 2 NEX, 4:43). Single partitions (a)

Fig. 4.14. High-definition FSE T2 sequence (matrix 512x320, thickness 3 mm, ZIP 1024) acquired with (a-c) and without (d-f) inverted contrast: coronal view. Note the excellent anatomical detail of both hippocampi

Fig. 4.15 b-d. (Cont.) Detail (b); MIP image (c); post-processing with volume rendering (d)

FLAIR Imaging

FLAIR sequences, which are generally included in all imaging protocols, are also sharper and more sensitive at 3.0 T by virtue of their reduced slice thickness and broader matrices, especially in detecting small lesions. Greater sensitivity is also afforded by the longer TE required in relation to 3.0 T relaxation times, allowing good contrast and acceptable noise levels (Figs. 4.164.18). Indeed, detection of small lesions requires CNR, rather than SNR, optimization, which is obtained using longer TE (120 ms).

The basic lack of difference in CSF T1 between 1.5 T and 3.0 T entails that a similar inversion time (IT) is re quired to suppress the fluid signal (around 2,250 ms) at both field strengths [3]. This does not apply to other IR sequences, neither those that seek to enhance contrast between different tissues (e.g. between white and grey matter), nor those using fat suppression, where T1 must be increased by 25 - 40 %.

The white matter, especially periventricular, areas that are physiologically hyperintense at 1.5 T enhance even more strongly on 3.0 T images, requiring great caution in distinguishing them from diseases like amy-otrophic lateral sclerosis [7].

The semeiological features described for grey nuclei and recent haemorrhage also apply to FLAIR sequences (Figs. 4.19, 4.20) [3, 19].

Fig. 4.16. FLAIRT2 image acquiredat 3.0 T (a) and 1.5T (b).Use ofbroadermatrices andthinnerslices at 3.0 T (a) disclosed small lesions difficult to identify at 1.5 T (b)
Fig. 4.18. Left hippocampal lesion on FSE T2 (a), „inverted contrast" FSE T2 (b), and FLAIR (c) sequences. The signal alteration is best appreciated in c. Optimum anatomical cortical-subcortical detail in b

A major drawback of high-field imaging is a larger number of pulsation artefacts, especially in the posterior fossa, probably due to higher blood flow magnetization. The adoption of broader saturation bands does not address the problem because of difficulties with the head coil, but recently developed new coils or new software (parallel imaging) are expected to diminish its impact.

Fig. 4.18. (Cont.)
Relaxation Audio Sounds Autumn In The Forest

Relaxation Audio Sounds Autumn In The Forest

This is an audio all about guiding you to relaxation. This is a Relaxation Audio Sounds with sounds from Autumn In The Forest.

Get My Free MP3 Audio


Post a comment