Neuroimaging Flashcards

Question

Consider the classic (non-EPI) pulse sequence in which one phase-encoding line is acquired per excitation. If I had a TR of 1s and wanted to collect data on a matrix of 128 by 128 how long would it take to acquire an image?

Answer 1

T1 is the relevant relaxation time. The image would take TRx(number phase- encoding steps) to acquire which is 128s.

Answer 2

Generally T2 gives less activation because only a subset of the contrast mechan- isms contribute to BOLD, for example none of the static dephasing mechanisms will contribute to T2.

Answer 3

The SI unit of magnetic field strength is the Tesla (T). A typical bar magnet generates a field of about 0.01 T. For reference the earth’s magnetic field is about 50 μT.

Answer 4

Draw a picture and check in the lecture notes how it looks, inside x,then,y,z,Shim,B0

Answer 5

(answer, yes it will be because the gradient maps frequency to position, and hence changing the gradient strength will change the position) If the RF pulse has a fixed bandwidth then how will the slice thickness be affected if the magnetic field gradient strength is halved? (ans. it will be doubled).

Answer 6

Both FDG and water have been used to investigate brain activity. Increased meta- bolism is found in stimulated regions, and of course increased blood flow. Unlike BOLD the senstivity to the blood flow increase is not confined to the post capillary vessels

Answer 7

1. A short TR and a short TE will give a T1 contrast. 2. A long TR and a long TE will give a T2/T2* depending on whether the acquisition is spin/gradient-echo. 3. A long TR and a short TE will give a proton density contrast. 4. A short TR and a long TE will not give a good contrast because tissue with a long T1 tends to have a long T2. A short TR leads to a stronger signal for tissue with a short T1 whereas a long TE leads to strong signal for tissue with a long T2. Hence the two effects tend to cancel.

Answer 8

Background Animation explanation The video shows the spin system being excited by a 90° (π/2)pulse. The spins then dephase under the influence of the inhomogeneities in the static field. The application of the 180° (π) pulse then flips the spins over, reversing the phase dispersion. The action of the inhomogeneities is then to bring the spins together resulting in a spin echo at a time TE/2 after the 180° (π) pulse. The total duration between the excitation 90° (π/2)pulse and the spin echo is known as the echo time and is conventionally denoted by TE. T ∗ 2 -contrast 1. T ∗2 - contrast is the easiest to generate: you simply have to excite the spin system and wait for the contrast to develop. 2. A gradient-echo sequence is typically used. 3. The contrast is allowed to develop for TE 4. If you want to quantify then data should be acquired at multiple TEs and an exponential fit performed. Relaxation times 3T 7T T1 T2 T ∗ 2 T1 T2 T ∗ 2 WM 887 77 49 1226 56 27 GM 1295 67 48-61 1781 50 28-33 Times are given in milliseconds. CSF will have considerably longer relaxation times than white matter and grey matter, but values are not widely available in the literature. As T ∗ 2 is much shorter than T2 this shows that the T ′ 2 effect dominates. Further reading: The best coverage is in the Pooley paper, but Currie and IMAIOS also have useful information

Answer 9

* We understand now that if we put a head in a strong magnetic field the water protons (amongst others) will be weakly magnetised. * To get a signal from the head we need to tease the magnetisation away from being parallel to the magnetic field, because then it will precess about the magnetic field. * We can do this by applying an additional magnetic field, the B1 field. This field is perpendicular to the main magnetic field. It rotates at the Larmor frequency (for protons in a 3T magnet this is 125 million times per second). * This field is generated by the resonator.

Answer 10

* After slice selection has been completed we can detect a signal from spins within the slice. * If a magnetic field gradient is applied within the slice then the frequency of the signal will be proportional to the position coordinate along the gradient direction. * The total signal will hence contain a mix of frequencies

Answer 11

* If you want to measure T1 then the experiment is repeated at a number of different TI values. A fit of the signal intensity as a function of TI will give T1. * If you saturate the magnetisation rather than invert it then this is a saturation recovery experiment. The contrast is less, but there is never negative magnetisa tion.

Answer 12

The signal increase varies with TE. The optimum T E is given by T E = T2*. Measurement performed at 1.5T so max signal change at T E u 60 ms.

Answer 13

* This is the characteristic time for the longitudinal magnetisation Mz to return to its original value M0. * As the equilibrium state is by definition that of lowest energy. The spin system has to give up energy to its surroundings to return to equilibrium. * Hence the name spin-lattice relaxation time. * After a 90° pulse 63% of the longitudinal magnetisation has recovered after T1.

Answer 14

* We have seen how the information from a 1D projection can be represented in both space and frequency. * This can be generalised to two or more dimensions. * The information in the second dimension is obtained by twisting the magnetisa- tion with a magnetic field gradient pulse before we acquire the signal (the total angle rotated is proportional to the position coordinate). * This is the phase-encoding gradient. It is applied orthogonal to the frequency encoding gradient. * In order to obtain an image on an m by n matrix n data points must be acquired in the frequency encoding direction, and m acquisitions with different values of the phase encoding gradient applied.

Answer 15

1. For protons there are two allowed states: precessing parallel or antiparallel to the main field (diagram). 2. There are slightly more protons in the parallel state because this has a lower energy level. 3. As the protons are randomly oriented on the cone the components of magnetisation in the transverse plane cancel. 4. There is hence only a net magnetisation parallel to the main magnetic field

Answer 16

1. The temporal resolution is defined by the ability to distinguish successive activ- ations in the same region. 2. From about 2 s separation between stimuli it is possible to distinguish BOLD responses. 3. Temporal shifts between activations in different regions can be measured with an accuracy of hundreds of milliseconds. However the speed of the haemodynamic response varies by up to a second between regions.

Answer 17

* Slice. The slice selection is the same as previously. * Read. Rather than collect one echo in the read direction a string of echoes is generated by repeatedly reversing the read gradient. Where the total gradient integral along read is zero corresponds to the centre of an echo. Data collected with a negative read gradient have to be reversed in time because this corresponds to swopping left and right. * Phase. Each echo is independently phase-encoded, so a small phase-encoding blip is applied before each echo. The large negative gradient before the first echo is cancelled out when half of the positive blips have been applied. This means that the zero integral of the phase-encoding gradient occurs half way through the echo train. * This type of EPI sequence is known as gradient echo EPI

Answer 18

1. Seen at all field strengths. 2. Takes 25-40 s to fully return to baseline, negative ’peak’ at 15-20 s. 3. Three possible explanations, summarised on next two slides: Picture p. 41 1. slow return of venous blood volume to baseline. 2. sustained elevated oxygen metabolism, or 3. flow undershoot.

Answer 19

* The positron is the antiparticle to the electron. * It is created when a certain class of unstable nuclear isotope decays according to the reaction: A/Z X −→ A/Z−1X + e+ + ν. * The other particle released is the neutrino ν which is extremely difficult to detect * The positron travels a short distance until it encounters an electron and annihilates according to: e+ + e− −→ γ + γ. * The two gamma photons produced are of equal energy, which is always the same, and shoot off in opposite directions. These characteristics form the basis of PET detection.

Answer 20

1. In order to excite a slice then a gradient has to be applied that is perpendicular to the plane of the slice. 2. The exact position of the slice is modified by changing the frequency of the RF pulse. Remember that for excitation to work the frequency of the RF pulse must be at the Larmor frequency. 3. Data from different slices is collected sequentially. 4. The slice thickness is proportional to the bandwidth of the pulse and inversely proportional to the strength of the slice selection gradient. The bandwidth of an RF pulse is inversely proportional to its duration. The carrier frequency is the frequency with which the B1 field oscillates. The bandwidth of the pulse is the range of frequencies over which the pulse is effective. Outside of this range the pulse should have no effect on the magnetisation.

Answer 21

The echo time TE is the time between the centre of the RF pulse for excitation and the centre of the echo. – In a gradient echo sequence TE determines the degree of T ∗ 2 contrast. – In a spin echo sequence it determines the degree of T2 contrast

Answer 22

The goal is to have a magnet with a very homogeneous magnetic field in the region where you are imaging. An infinite solenoid will do this (a finite solenoid is shown). However, this is not a practical design, and compromises have to be made to restrict the dimensions and allow access

Answer 23

The precession frequency is given by ω0 = γB0 (1) Here ω0 is the angular frequency (often called the Larmor frequency), B0 the main magnetic field and γ the gyromagnetic ratio which is nuclear specific. This tells us that the frequency of rotation is always proportional to the mag- netic field

Answer 24

* Consider as an example the frequency encoding of two small bottles, one of which is at the origin. * If the bottles are narrow there are only two frequencies present

Answer 25

TE=20 ms (left) versus TE=70 ms (right). Data obtained at 1.5T. Note the signal dro- pout in the inferior slices at TE=70ms and generally lower signal to noise ratio.

Answer 26

* Radiation * Health risk * No repeated experiments * Intravenous injection * Intra-arterial cannulation (only for quantitative blood flow measurements) * Few data points (12/subject) * Fixed slices * Poor temporal resolution * Expensive

Answer 27

Viewed separately the signals will be as shown in the diagram. The total signal that you measure will just be the sum of these. (Please note only the differences to the Larmor frequency are shown) VIDEO: The video shows two signals, but in contrast to the first example, the bottles do not contain the same amount of water, (the one giving the red signal has more) and neither is at the origin. 2. When we are looking at a realistic object, we acquire signals from a whole range of frequencies. The signal is hence difficult to interpret by eye

Answer 28

The classical method of generating T1 contrast is by means of the inversion recovery experiment. The magnetisation is initially inverted, and after an inversion time TI interrogated by an excitation pulse

Answer 29

1. Current whole body systems reach gradient strengths of about 80 mT m−1. 2. The maximum field generated by these is about 40 mT or two orders of magni- tude smaller than the main field. 3. The gradient fields can be switched in about 100 μs.

Answer 30

The ideal answer would explain that 1) is a T2*-weighted image because there is little contrast between grey matter, white matter, and CSF. 2) Is a T1-weighted image be- cause the CSF is darker than all other tissues which must be a result of nulling out the signal of CSF by exciting the spin system at the time that the CSF signal passes through null. A further indication that this is T1-weighted is that grey matter is darker than white matter, which would never be possible with pure T2 or T2*-weighting. 3) 54 is then T2-weighted because CSF has by far the brightest signal. A further indication is that white matter is darker than grey matter owing to its shorter transverse relaxation time. You would get one point for getting the correct contrast and a further two points for the quality of the explanation. There would be a bonus point for not making any incorrect statements in the whole question.

Answer 31

In modern systems separate coils are used for transmit and receive. Multiple chan- nel receiver coils have two advantages: 1. They improve sensitivity. 2. They can encode some degree of spatial information. This can be used to speed up image acquisition.

Answer 32

* The echo time TE is the time between the centre of the RF pulse for excitation and the echo where the integral of the phase-encoding gradient is zero. This example is gradient echo EPI. * The echo time TE is the parameter associated most with the functional contrast. A certain TE is needed to allow the functional contrast to develop. * However, a long TE is also associated with signal loss. * Hence in fMRI a compromise is often made, and a TE somewhat shorter than the optimum is used to avoid signal losses.

Answer 33

* Blood flow is the dominant effect * There is hence a net REDUCTION in the amount of de-oxyhaemoglobin during activation * Consequently T2* increases * During activation the MRI signal INCREASES

Answer 34

* Spins may lose coherence by exchanging energy with each other. One goes faster, the other slower. This is the spin-spin interaction and is characterised by T2. * They will also lose coherence if there are spatial variations in the Larmor fre- quency in the object. These are caused by inhomogeneities of the static magnetic field. So in a perfectly homogeneous field only T2-relaxation will occur. * The first mechanism is random (stochastic) and hence irreversible. The second is constant and hence reversible. * Both contribute to T ∗ 2 , but only the first to T2

Answer 35

1. If we rotate with the radiofrequency field then the picture is simplified. 2. The application of a radio-frequency field can rotate the magnetisation about any angle. 3. A 90° pulse generates the maximum amount of transverse magnetisation and hence the maximum signal. 4. The RF pulse typically lasts a couple of milliseconds, and is then turned off. Signal can not be acquired whilst the pulse is being transmitted. 5. Excitation is most effective at the Larmor frequency. * The rotating magnetisation induces a voltage according to Faraday’s law. * Do not imagine one large magnet in the middle of the object, but rather a distri- bution of small objects. * Essential behaviour is like a bicycle dynamo.

Answer 36

It is weak because the difference in number of spins precessing parallel to z compared to those precessing antiparallel to z is very small, about 1 in 100,000 at 3T. As the spins can be found anywhere on the cone, and will hence be equally distributed over the cone, the components of the field in the transverse plane will cancel, leaving only the component along z.

Answer 37

(ans. in this unlikely case there would be a flat plateau to the activation if viewed in 1D). In reality the PSF is unknown and the measurement has noise, which makes all estimates of ultimate spatial resolution uncertain

Answer 38

A change in the magnetic environment of a wire loop (i.e. the field lines from a magnet moving through the coild) will give rise to a voltage across its terminals and current will flow. Think of a bycicle dynamo!

Answer 39

* Speed. Single image in about 50ms, whole brain in 2s. * Sensitivity. Gradient echo EPI has T ∗ 2 contrast which is what we need for fMRI. * Spatial Resolution. Adequate at most field strengths * Accuracy. Image distortion due to B0 inhomogeneity and signal voids due to susceptibility differences at tissue-air interfaces

Answer 40

pictures p.48/49

Answer 41

The static magnetic field is orange, and remains stationary. The rotating B1-field is yellow. Because this rotates at the Larmor frequency the angle with the magnetisation (white) is fixed at 90°

Answer 42

* Intravascular effect of deoxyhaemoglobin is to reduce intravascular T2 and T2* * Extravascular dephasing occurs reducing T2* * Both T2 and T 2* are shortened by the presence of deoxyhaemoglobin, however the effect on T2* is the greatest, and this is generally utilised for fMRI

Answer 43

We can shape the magnetic field by shaping the wire. we use this effect to generate all the magnetic fields used in MRI

Answer 44

1. Thought to be due to oxygen metabolism increase BEFORE blood flow increase. Results in small negative signal change. 2. Difficult to detect below 4T. 3. Peaks 1 – 2 s after stimulation, visible for up to 3s. 4. Potential high spatial resolution (0.5 mm). 5. Never established as a routine fMRI technique. Only considered for complete- ness.

Answer 45

Blood flows from the arterial to venous side and becomes de-oxygenated as it passes through the capillary bed. If neural activity increases in a small region then the blood flow may increase over a larger region (“watering the whole garden for the sake of a single thirsty flower!”)

Answer 46

An early demonstration of this effect was on a rat breathing either air or oxygen The left image shows the T2*contrast arising from breathing oxygen. In this situation sufficient oxygen dissolves in the blood plasma to satisfy the metabolic requirements, and little de-oxyhaemoglobin is formed. The right frame (rat breathing air) demonstrates the effect of de-oxyhaemoglobin on a T2*-weighted image. Signal loss is seen around the venous blood vessels. This contrast is known as blood oxygenation level dependent (BOLD) contrast.

Answer 47

Doubling the current will double the field while reversing the current will reverse the polarity of the field.

Answer 48

1. Measurements of hemodynamic PSF give values of 2-3 mm. This can be con- sidered the intrinsic spatial resolution of fMRI. 2. The intrinsic spatial resolution improves with increasing field strength. 3. Technically we can easily achieve a spatial resolution of better than 2 mm with EPI, and with advanced techniques better than 1 mm. 4. In standard fMRI the spatial resolution is limited by the haemodynamic PSF and not the technical limitations of spatial resolution. Advanced techniques for investigating cortical layers and columns do exist, and the intrinsic resolution can then be under 1 mm.

Answer 49

Echo Planar Imaging * The classical imaging methods are too slow for functional imaging because each phase-encoding step requires a separate excitation. * Echo Planar Imaging or EPI collects data for a complete image after one excita- tion.

Answer 50

There are two fundamental factors that determine the spatial resolution: 1. The average distance (defined as root-mean-square distance) travelled by the positrons before annihilation. This can vary significantly between nuclei. 2. The residual energy of the electron-positron pair. This means that the two photons do not go in exactly opposite directions. The distribution of directions is near Gaussian with a half value of about 0.3 degrees. There is also the practical limit of the size of the scintillator crystals. In addition smaller crystals require more electronics and longer measurement times

Answer 51

If you put these nuclei in a magnetic field then the magnet will try to align itself with the field, but the rotation will force it to precess about the field. This is a motion like that of a child’s spinning top.

Answer 52

1. The signal is acquired during a gradient of constant amplitude. This is known as the readout or frequency encoding gradient. 2. In order to obtain a time symmetric signal a negative gradient is applied prior to the frequency encoding gradient. This is used to shift the maximum of the signal to the centre of the acquisition window. This makes it easier to analyse the data. 3. A time-symmetrical signal is known as an echo. As this type of echo is generated by gradient reversal it is known as a gradient echo. 4. The centre of the echo will occur when the time integral of the gradient is zero.

Answer 53

A cyclotron is a device for accelerating protons to a high velocity, and then allowing them to strike a target nucleus. The proton is then added to the nucleus creating a radioactive isotope like 18F which can then be incorporated into a biochemically relevant substance (like FDG). Background There is a constant magnetic field perpendicular to the plane of the cyclotron. This makes the particles rotate in a circle. An electric field is applied across the two ’D’s to accelerate the particles.

Answer 54

* Some atomic nuclei have a property known as spin. * As the nuclei are also electrically charged the spin makes them behave like small magnets. This is because the charge rotates in a circle, which is equivalent to a current loop.

Answer 55

* Multiple rings of detectors make it possible to do multi-slice imaging using either axial slices, or cross planes. * If the septa are removed from the detectors so that the angle of entry is not limited then 3D imaging can be performed with a considerable increase in sensitivity. This is now standard * A separate development is time of flight PET. Advances in electronics and detec- tion crystals make it possible to measure the time difference between the arrival of the two photons with a precision of a couple of hundred picoseconds

Answer 56

* Field distortions cause frequency variation - The signal we measure comes from water protons resonating in phase. * Frequency ∝ field strength

Answer 57

1. The BOLD signal can saturate at high levels of stimulation but is generally con- sidered to be additive for most situations encountered in practice. 2. Similarly it is possible to predict the response of longer stimuli based on that of shorter stimuli, and vice versa. 3. This is not true for very short stimuli (/4 s) which exhibit a disproportionately large BOLD response, and those for stimuli longer than about 24 s for which the BOLD response will be smaller. 4. Linearity of BOLD response is important, as most analysis techniques rely on General Linear Model

Answer 58

Increasing the gradient strength will increase the spatial resolution and hence the object will appear larger.

Answer 59

Principle of PET. A nucleus emits a positron (e+) which at some point encounters an electron (e−). These annihilate to produce two gamma rays (photons) each with an energy of 511keV. On the right an annihilation is shown in the head and the two photons are detected in the detection ring. Background The size of the photon is of the order of 100s of picometres.

Answer 60

BOLD measures neuronal activity indirectly. This has consequences in both the spatial and temporal domains. * The blood flow increase is present in the entire vasculature from arteries to veins, but the BOLD signal change is driven by the amount of deoxyhaemoglobin, and is only in the capillaries and downstream venous vasculature. * There is a time lapse between the neuronal activity and the blood flow response.

Answer 61

The diagram shows a theoretical field distortion about a cylindrical vessel. The vertical cylinders are the blood vessels and the distortion to the field outside the cylinders are shown. The distortion is zero for 0% deoxyhaemoglobin and the plane is flat. The local change in field is proportional to the local distortion away from the flat plane.

Answer 62

1. The presence of a paramagnetic substance in the blood will clearly alter its relaxation properties. Within blood vessels the presence of paramagnetic deoxyhaemoglobin will reduce the homogeneity of the field and lead to a reduction in T2*. 2. The deoxyhaemoglobin will also modify the magnetic field outside the blood vessel.

Answer 63

* In the majority of imaging experiments we obtain a two dimensional image. * To do this we need to selectively excite spins within a well-defined slice. * The basic principle is to apply a radiofrequency pulse at the same time as a magnetic field gradient. * The flip angle of the pulse is adjusted by changing the amount of current flowing in the transmitter coil. * The most important flip angles are 90◦ which gives the maximum transverse magnetisation, and 180◦.

Answer 64

1. T1. Spin-lattice relaxation time. Gives the time to return to thermal equilibrium. This means that the magnetisation recovers back to the longitudinal axis. 2. T2. Spin-spin relaxation time. Lifetime of transverse magnetisation in a perfectly homogeneous main field. 3. T ∗ Effective lifetime of transverse magnetisation.

Answer 65

A is the atomic mass, Z is the atomic number (=charge of the nucleus).

Answer 66

Check out pictures p. 46

Answer 67

This is because the recorded signal is generally strongest when it is not dephased by a gradient. This signal determines the image contrast.

Answer 68

TE is modified without changing the ETL. A delay (T) is inserted between excitation and readout

Answer 69

- electrical currents generate magnetic field - IMPORTANT: relationship of piece of wire and magnetic field generated by it - evaluate current generation to create magnetic field - if the strength of the current varies in time, then so does the strength of the field

Answer 70

The image was acquired with a spin-echo type sequence. CSF has the longest T2 and appears brightest, followed by grey matter which is here brighter than white matter

Answer 71

* Good access to subject * Relatively insensitive to movement * Little noise * Little electromagnetic disturbances * Intracorporeal metal no problem * No signal loss in the inferior temporal lobe * Quantitative cerebral blood flow measurement possible * Easy combination with specific tracer measurements

Answer 72

* We can break T ∗ 2 into its constituent parts 1/T2* = 1/T2 + 1/T2' where T 2' is the contribution to T2* of the inhomogeneities in the static magnetic field. * We hence can write T2 => T2* * In a situation where there was no dephasing contribution to T2 then all T2 relaxation would be driven by the same mechanisms as T1. * Hence T1 => T2 * The next video shows simultaneous transverse and longitudinal relaxation

Answer 73

* Oxygen is transported from the lungs to the brain via the arterial blood supply. Oxygen can dissolve in the blood plasma, but this is generally inadequate to meet demand. Hence it is also transported via oxyhaemoglobin. * The walls of the arteries and veins are impermeable to oxygen, but it can diffuse through the endothelium of the capillary bed. The greater the difference in concentration of oxygen between tissue an blood, the more rapidly oxygen diffuses out of the blood and into the tissue. * Initially the blood diffuses out of the plasma and then the oxyhaemoglobin dissociates to oxygen and de-oxyhaemoglobin. The de-oxyhaemoglobin is paramagnetic and hence shortens the local T2* -relaxation time.

Answer 74

The left shows how frequency maps onto position, and shows the signal for three exemplary positions (note how the frequency increases with r) The right part of the diagram shows the total signal at the top, and the magnitude of the FT at the bottom. Note. The bottom part of the diagram does not represent a 2D image but rather intensity (y) as a function of position (x).

Answer 75

* Imaging is made possible by the application of pulsed magnetic field gradients. * The axis system is defined so that the main magnetic field is along the z-axis, the x-axis is left-right for someone lying on their backs, and the y-axis is front-back. * The gradients, produce a small linear variation in the z-component of the main field as a function of x,y,z. * As the frequency of the MRI signal is proportional to the magnetic field the signal frequency is then linearly proportional to position

Answer 76

The scintillator used in nearly all new systems is Lutetium Oxyorthosilicate (LSO or LYSO) as it has favourable properties for PET like high effective Z and density, combined with good energy resolution, very good light output.

Answer 77

Look at picture page 47 * Coincidence events are grouped into projection images known as sinograms. * Pre-processing required to correct for random coincidences, estimation and subtraction of scattered photons, detector dead-time correction. * Traditionally used filtered back projection (essentially the scheme illustrated in the following slides). * Now iterative expectation maximisation algorithms are commonly used: give better noise profile and no streak artefacts, however need more computing resources.

Answer 78

1. The early visual areas are retinotopically organised. 2. I can hence impose a spatially regular pattern of activation on the cortex. 3. If activated regions are well separated then blurring due to the hemodynamic response will not reduce the contrast, but as the separation between activated regions is reduced then the blurring will lower the contrast

Answer 79

* Approx. 3.9 mSv for 12 scans * For comparison: * Natural radiation load per year: approx. 2 mSv * Head CT: approx 6 mSv * EU standard for professional exposition: max 50 mSv / year

Answer 80

1. The simplest means of generating T2-contrast is a spin-echo experiment. 2. True T2 relaxation is stochastic, and hence irreversible, but the effects of static field inhomogeneities are static in time and their effects can be reversed. 3. The signal intensity at the centre of the echo should be modulated solely by T2 relaxation. 4. Fitting the intensity as a function of TE to an exponential decay will yield T2.

Answer 81

It is common in medicine to use T1 and T2 contrasts for diagnosis. T2* contrast is used for fMRI. 1. T1 contrast can be added into a pulse sequence for example by inverting the magnetisation some time before data acquisition and doing something similar to an inversion recovery experiment. Using a short TR experiment is not common in cognitive neuroimaging. 2. T2 contrast is attained usually by means of a spin echo. 3. T2* -weighted images can be acquired by ensuring that there is a delay between the excitation pulse and the signal acquisition in a gradient echo sequence.

Answer 82

There are three physiological parameters that increase upon activation and will af- fect the amount of de-oxyhaemoglobin present: * Oxygen Consumption (Sometimes known as CMRO2, the rate of metabolic consumption of oxygen). An increase in this parameter leads to an increase in the blood de-oxyhaemoglobin. * Blood Flow (known as CBF, cerebral blood flow). Reduces the concentration of de-oxyhaemoglobin. * Blood Volume (known as CBV, cerebral blood Volume). An increase in venous blood volume leads to an increase in the amount of de-oxyhaemoglobin within a voxel without changing the concentration in blood. The blood volume changes in the arterial, venous, and capillary compartments

Answer 83

- it is impossible in practice to build a magnet with a perfectly homogeneous field. Furthermore the presence of an object will distort the field. - In order to compensate for these imperfections additional weak static fields are applied using room temperature coils. These are known as shim coils. - The current in these coils is adjusted before a measurement to give the best ho- mogeneity possible.

Answer 84

Your sketch should show arrows in the z-direction that have a length that is pro- portional to the position coordinate in the gradient direction, i.e. for a z-gradient the length is proportional to the z-coordinate and of course reverses direction for negative coordinates. The B1field is always rotating in the x-y plane.

Answer 85

The sketch should show a sinusoidal oscillation. If you reverse the direction of rotation you will still get a sinusoidal oscillation. The larger loop of wire the greater the voltage

Answer 86

1. The coil used to generate the RF field is often called a resonator. It can generally both transmit the RF field and detect the MR signal. 2. A common design is the birdcage coil. This generates a weak magnetic field of about 30 μT that rotates about the long axis of the coil at radiofrequency (MHz range). The long axis of the coil is parallel to the z-axis of the magnet. 3. Within the coil the transmission field should be as homogeneous as possible.

Answer 87

FDG= 2-deoxy-2-fluoroglucose. Is metabolised the same way as naturally occurring glucose. Half-life 110 min. Oxygen has a half life of two minutes and is hence only viable as a tracer when there is a cyclotron next to the scanner.

Neuroimaging Flashcards

(114 cards)