Part I The Basic Stuff

1. Early Daze: Your First Week in MR …12

2. Seeing is Believing: Introduction to Image Contrast …15
3. Lost in the Pulse Sequence Jungle…14
4. The Devil’s in the Detail: Pixels, Matrices and Slices..12
5. What You Set is What You Get: Basic Image Optimization…13
6. Improving Your Image: How to Avoid Artefacts…20
7. Spaced Out: Spatial Encoding…22
8. Getting in Tune: Resonance and Relaxation…17
9. Let’s Talk Technical: MR Equipment…20
10. Ghosts in the Machine: Quality Control…19

Part II The Specialist Stuff

11. Acronyms Anonymous I: Spin Echo…22
12. Acronyms Anonymous II: Gradient Echo..18
13. The Parallel Universe: Parallel Imaging and Novel Acquisition Techniques …
23
14. Go with the Flow: MR Angiography …18
15. A Heart to Heart Discussion: Cardiac MRI…19
16. It’s Not Just Squiggles: In Vivo Spectroscopy …15
17. To BOLDly Go: fMRI, Perfusion and Diffusion…23
18. Making it Count: Quantitative MRI…19
19. But is it Safe? Bio-effects...13
20. Where Are We Going Now?...5
1.
Zonal Defence: Control and Access
Zona 1 : areas that are accessible to the general public
Zona 2: Patients may be screened in Zone II and will generally be under
supervision
Zona 3: zone under the control of designated MR personnel
Zona 4: MR examination room

2. Cuidado
Contraindicaciones: conventional cardiac pacemakers or implanted cardiac
defibrillators; abandoned cardiac leads; cochlear implants.
Precaución: Clips Quirúrgicos, neuro estimuladores, cámaras endoscópicas,
implementos cardiacos. Trabajadores de metal. Militares. Tatuajes. Pacientes
sistema termorregulador comprometido. Prótesis válvulas cardiacas.
Embarazadas primer trimestre.

3. Dealing with Implants

4. simple classification of the body tissues, which will be good enough to

describe the basic appearances:
fluids – CerebroSpinal Fluid (CSF), synovial fluid, oedema;
water-based tissues – muscle, brain, cartilage, kidney;
fat-based tissues – fat, bone marrow.

5. T2-Weighted Images
T2-weighted (T2w) images are one of the most important MR images,
because they are sensitive to fluid collections.
SE T2 images require long TR and long TE, so they have a long scan time.
For brain and spine imaging, T2w images are usually acquired with the spin-
echo pulse sequence. For liver, where a breath-hold is needed, the T2w
images can be acquired using gradient echo which is faster.

Flair
 to remove the CSF signal, known as ‘nulling the signal’, by choosing an
Inversion Recovery (IR) sequence instead of spin echo, and carefully setting
the inversion time (TI). 1800 and 2500 ms.

T1:
T1w images the longest T1s have the darkest signal. Tissues with short T1s
appear brighter. T1w images are usually quite fast to acquire, because they
have short repetition times (TR). T1w images are often known as ‘anatomy
scans’, as they show most clearly the boundaries between different tissues.

T1w Images Post-Gd

The total body dose has a half-life of around 90 min in subjects with normal
kidneys, and can be considered completely eliminated after 24 h. It has the
effect of shortening the T1 of tissues where it accumulates, so the most useful
images to acquire post-Gd are T1-weighted.

STIR Images
Images, especially for spine and for musculoskeletal imaging. STIR images
have very low signal from fat but still have high signal from fluids, fat has a T1
of 220 ms at 1.5 T, so if we set TI to 150 ms, the signal from fat can be
suppressed. At other field strengths TI will be different, e.g. fat T1 is
approximately 380 ms at 3.0 T.

Densidad Protónica
in the knee you can distinguish articular cartilage from the
cortical bone and menisco.

Imágenes Eco gradiente

the choice of excitation flip angle, α, is much more important to determine the
contrast in the images. A 30° excitation RF pulse leaves most of the protons
in their equilibrium position, aligned with the main magnetic field. They also
have very short TRs 500 ms avoid T1.

6. Gradient-Echo T1-Weighted Images

Parameters needed tO get T1w GE images. A short echo time, a flip angle of at least
50°. On GE Healthcare systems, choose SPGR; on Philips, choose T1-FFE; and on
Siemens, choose FLASH
One interesting variant with T1w GE images is the possibility to acquire in-phase and
out-of-phaseimages. This refers to the choice of echo times, which determines
whether water and fat signals are either in phase with each other, or out of phase.
This technique allows a radiologist to detect if there is diffuse fatty infiltration in the
liver or other organs, since the OP images will show a darker signal than the IP
images.

7. 3.9.2Gradient-Echo T2*-Weighted Images

the relaxation of tissues after an RF pulse, making the spin–spin relaxation time
appear shorter. We call this the apparent spin–spin relaxation time T2- To produce
GE T2* images, select GRE sequences on GE Healthcare systems; choose T2-FFE
on Philips; and FISP on Siemens scanners. You need to keep α small to avoid T1
weighting.

8. P D image.
use the same type of sequence as you did for a GE T1w image, setting the TR to be
as short as possible (for a 3D scan) or long enough to get the required number of
slices. Since we don’t want any T2 decay, the echo time must be short, and α must
be small to avoid creating T1 contrast. GE PDw images are not often used in clinical
practice, so we won’t spend

9. Contras Agent
Since gadolinium is paramagnetic, it alters the local magnetic field in areas where it
accumulates. Not enough to be measurable, but enough to affect the relaxation
times of the tissues. In particular, the tissue T1 is reduced, and this leads to higher
signal on T1w images after Gd injection.

10. Angiographic Images

Artefacts. How can a long-T1 fluid give a high signal on T1-weighted images? It is
because the blood is flowing. This blood has not been tipped by the RF pulse, so
when the next pulse is applied the blood has its full equilibrium signal
The three most important sequences are ‘time-of-flight MRA’ (also known as ‘in-flow
MRA’), ‘phase-contrast MRA’, and ‘contrast-enhanced MRA’, which uses a very
rapid imaging sequence during the injection of gadolinium. time-of-flight sequences
may not distinguish freshly thrombosed clots from flowing blood (because the
methaemoglobin is starting to affect the MR signal), stenoses may be exaggerated in
terms of length and severity, and very slow-flowing blood may disappear altogether.
 11. Diffusion-Weighted Images
Diffusion of water molecules,which changes in certain pathological conditions.
These gradients have a large amplitude and duration, which means that they force
the TE to be rather long, e.g. 80 ms.
Diffusion is also widely used in body imaging, for example breast, liver or prostate.
Focal tumours in these organs also tend to have high cell densities and show the
same high signal on DWI as brain tumours or strokes.

12. Pulse Sequense

While the end-point of an acquisition can be expressed in terms of T1 or T2
weighting.
Scan time is determined by TR, along with the number of lines (NPE) in the image
and Number of Signal Averages (NSA, called NEX on GE Healthcare systems):

Scan time ¼ TR × NPE × NSA

 13. Turbo Spin Echo: The Work Horse

In Turbo or Fast Spin Echo (TSE, FSE), multiple signal echoes are collected
following each excitation pulse (Figure 4.8) and therefore the scan time can be
reduced by the Turbo Factor (TF) or Echo Train Length (ETL).
TSE has become the standard sequence for T2-weighted imaging. It produces
excellent T2 contrast by the combination of long TR (to avoid T1 effects) and
variable TE to control the extent of T2 weighting.
to add an inversion pulse before the excitation. This sequence is called Inversion
Recovery (IR) and results in very strong T1 weighting. IR may also be used to
remove unwanted signals from the image, such as fat as in STIR or cerebrospinal
fluid as in FLAIR.
Driven equilibrium pulse is a 90° pulse applied after the MR signal has been
acquired and just before the next excitation pulse.
HASTE or Single-Shot FSE/TSE (SS-TSE). These sequences are very useful for
imaging fluid structures, e.g. the biliary system in MR Cholangio- Pancreatography
(MRCP) examinations.
One of the downsides of TSE is its sensitivity to movement artefacts. To reduce this,
radially acquired TSE can be used. Sometimes known as PROPELLER, MultiVane,
BLADE or JET, the sequence offers more limited image contrast,
 14. The Other Branch of the Tree: Gradient Echo will commonly be used for MR
angiography, contrast studies, breath-hold imaging and high-resolution 3D
imaging. EPI is the ultimate in terms of scanning speed, collecting a whole
slice in under 100 ms.

In GE a new parameter, the flip angle, α (alpha) is introduced, this mainly controls
the T1 appearance of the image. T2* differs from T2 in that it relates not just to the
tissue properties, but also to the quality of the magnetic field or inhomogeneity of the
scanner.
Spoiled Gradient Echo
used to give fast T1- weighted sequences in either 2D or 3D. They are particularly
suited to dynamic gadolinium contrast enhanced studies (FLASH, T1-FFE, or
SPGR).

Rewound Gradient Echo

Rewound GE sequences may be called FISP, FFE (FIESTA) or GRE. The sequence
may be performed in 2D or 3D, and the short TR makes it particularly suitable for 3D
scanning.

Ultrafast GE Sequences
we can use pre-pulses to improve SNR and contrast, e.g. to reintroduce T1
weighting we would add an inversion pulse at the start of the sequence. These
sequences carry the generic name RAGE, can be optimized for fast 3D breath-hold
abdominal examinations, and may be called VIBE, LAVA or THRIVE.
 15. Echo Planar Imaging
hybrid sequences
GRASE, generating both spin echoes and gradient echoes. The reason for this is
that the contrast which GRASE generates is closer to a spin-echo or TSE contrast,
i.e. T2-weighted rather than T2*-weighted.
EPI is used in applications such as perfusion, diffusion or functional MRI, where
extreme acquisition speed in a single shot is required. Spin-echo EPI is used for
DWI, while gradient echo EPI (or simply EPI) is used for perfusion and fMRI
(produce quantitative maps of diffusion, perfusion or brain oxygenation)

16. Pixels, Matrices and Slices

a. From Analogue Signal to Digital Image

The MR signal is detected by the receive coils, and is simply a voltage induced in the
coil. MR signal in the receive coil is a continuously changing voltage; no matter how
closely we zoom in, it still varies smoothly. The ADC makes a measurement of the
voltage, calculates the appropriate number and stores the digital value in the
computer, so the signal also changes from being continuous to stepped or discrete
data.
Earlier we learned that frequencies higher than fN will be aliased and appear as low
frequencies. In order to avoid this corruption of the spatial information, we use either
an electronic (analogue) or digital filter to remove all theMRsignals with frequencies
higher than fN.

b. Matrices, Pixels and an Introduction to Resolution

The pixels in MR images are organized into rows and columns in a matrix (plural
‘matrices’). Each pixel in the reconstructed image can be thought of as a location in
the computer memory or hard disk, containing a number which represents the signal
intensity. Once the raw data matrix is full it is reconstructed into the final image using
a clever piece of maths called a Fourier transform. Notice that when you set the
phase-encode matrix, you define how many times the sequence must be repeated
(how many rows there are in k-space) and therefore how long the scan will take to
acquire. you can usually recognize the PE axis by looking for ghost signals from
motion, as these always go across the PE direction
As a general rule you should not make the PEmatrix less than half the FE matrix,
because it makes the pixels too pencil-like.

c. Slices and Orientations

When we select an axial slice, we are creating images perpendicular to the Z
direction. Sagittal images are perpendicular to the X direction, and coronal images
are perpendicular to the Y direction.

d. Displaying Images
Display systems usually have 12- or 16-bit depth (4096 or 32 768 grey levels
respectively). However, the human eye can only distinguish about 200 grey levels
example, the highest pixel value has the brightest screen intensity, zero-valued
pixels are black, and everything in between is scaled accordingly.

e. What do the Pixels Represent

the sign from just a small volume of tissue within the patient’s body, known as a
voxel (a contraction of ‘volume element’). The actual signal intensity depends on
many factors, including the sequence timings and the intrinsic T1, T2 and PD of the
tissues. If

f. From 2D to 3D
So during each TR, the scanner excites and collects echoes from many slices. The
signals of different slices do not interfere with each other, thanks to the way slice
selection works usually with a small gap between adjacent slices
known as multi-slice imaging. Due to imperfections in the RF pulses we usually
have to introduce a slice gap to separate the slices. This is measured as the
distance between the slice edges.

g. True 3D scanning
technique, with phase encoding in the slice direction as well as in-plane. For every
slice encode, we must acquire all the in-plane phase encodes. So the scan time gets
multiplied by the selected number of slices (sometimes called partitions) in the 3D
volume. 3D scans have a few advantages over multi-slice 2D imaging. For example,
it’s possible to define very thin slices with no slice gaps (contiguous slices), and, if
the voxels are nearly isotropic, the resulting volume can be reformatted on a
workstation to produce images in any orientation.

Basic Image Optimization