Datasets:

Karthik1080

/

Radio_MRA_V1_C37

like 0

The term angiography stands for the imaging depiction of vascular structures, and there are many ways to evaluate the vascular structures like Doppler, computed tomographic, catheter and magnetic resonance angiography (MRA) Over the last two decades MRA has been increasing in demand because of its physiological nature, on the contrary to CTA and catheter angiography, which involve catheterization, radiation and nephrotoxic iodinated contrast agents However, there are many limitations of MRA such as availability, cost, time consuming and high sensitivity to motion and flow related artifacts

The first research meeting devoted to Magnetic Resonance Angiography was hosted by Roberto Passariello in L’Aquila, Italy in 1989 Those days three-dimensional phase contrast MRA would take 19 hours from the time the patient entered the magnet until images could be seen; one hour to acquire the image and 18 hours of overnight image postprocessing Computational capabilities of modern equipment have reduced the delay to a few seconds Postprocessing now has taken a more central role in the communication of enormous amounts of data with less cumbersome two- or three-dimensional projections

Many variations on the MRA theme have been presented over the ensuing 15 years Dennis Parker developed the 3D multislab time-of-flight MRA technique which remains in routine clinical use to this day Pulse sequence design plays a major role in the continuing advancements in the field, most notably as a consequence of more sophisticated and novel k-space filling strategies The work of Kent Yucel and Martin Prince at the Massachusetts General Hospital in 1992 brought gadolinium-enhanced MR angiography to clinical utility The first-pass dynamic contrast-enhanced MRA method provides robust and reproducible imaging results that have propelled the adoption of MRA into wider clinical use

This advance reliably produced images of sufficient quality to replace invasive catheter-based X-ray contrast angiography for most diagnostic purposes

With the advent of high field clinical scanners operating at 30 Tesla, earlier noncontrast methods can be improved by virtue of two key benefits offered by 30 T MRA The spatial resolution can be improved or scan time can be shortened by up to a factor of 4 because the signal to noise ratio (SNR) achieved with 30 T is twice that of the 15 T

The longer T1’s of tissues at 30 T, ~20–40% higher than 15 T, provides better background suppression, additional inflow enhancement, and improved contrast-to-noise

FLOW PHENOMENA AND PHASE EFFECTS IN MRA Flow Phenomena Flow phenomena in blood or cerebrospinal fluid (CSF) also influence the MR image contrast; in addition to inherent tissue factors like T1, T2 and proton density Blood flow is complex and variable inside the body, so it is important to understand the various types of flow

Laminar flow is flow where the particles move along in concentric sheets and laminae, ie different but consistent velocities across the vessel It is seen in normal vessels

Plug flow is flow where all fluid particles move forward in parallel lines with the same speed and has a characteristic blunt profile It is seen in the descending thoracic aorta

Turbulent flow is flow at different velocities which varies, ie velocities across the vessel changes and is seen at vascular bifurcations

Vortex flow is flow after narrowing and is seen after stricture or stenosis In it, the high velocities are seen at the center

Stagnant flow is flow that nearly behaves like stationary tissue and is seen in occluded vessels and large aneurysms

The moving spins (spins that move during acquisition of data) show different contrast characteristics from the stationary spins The moving spins cause mismapping of the signal because of the flow phenomena and result in flow motion artifacts or phase ghosting The flow phenomena are generally categorized into time of flight, entry slice phenomenon and intra-voxel dephasing

Outflow-related Signal Loss (Washout Effect, T2 Flow Void) If a spin-echo (SE) pulse sequence is used to obtain the images, high velocity blood flow at a plane perpendicular to imaging plane produces weaker signal compared to adjacent stationary structures

This occurs because flowing spins get washed out from the slice during the imaging process A sequence of slice-selective 90° and 180° radio frequency (RF) pulses are applied in spin-echo sequences If a tissue component is affected by both the pulses, only the MR signal is produced

Flowing blood in any vessel or per se any moving material, while flowing through an excited slice at a reasonably high speed, is affected by only one of these pulses and as a result, it does not produce MR signal resulting into the so-called “flow void” (Figs 1A and B) The intensity of the vascular signal declines with decreasing slice thickness, increasing echo time (TE), and increasing flow velocity

The vessels would appear dark when the blood flow speed is so high that all spins leave the slice between the 90° and 180° pulses (v ≥s/ (TE/2), as no signal is produced This phenomenon does not affect spins flowing within the imaging plane This effect secondary to washout phenomenon is observed only for spin echo sequences and is most marked on T2-weighted sequences as long echo times are used On the other hand, in a gradient-echo (GRE) sequence, the echo is refocused without a 180° pulse simply by reversing the imaging gradients The washout effect does not occur as only one RF pulse is needed to form an echo Valuable and reliable information about blood flow can be obtained by the washout effect provided by standard SE sequences If the flow void of a particular vessel is not visualized on T2-weighted images, it generally indicates very slow flow or even vessel occlusion On the contrary, in presence of flow void on T2-SE sequence vessel occlusion can be excluded

FIG 1A Spin echo sequence depicting the “Flow void” phenomenon (A and B) Represents the tissue sample with shaded area as stationary tissue (spins) and central nonshaded area as vessel having moving spins (A) At the time of 90° excitation and 180° rephasing RF pulse for that particular selected slice location; (B) Same tissue with same slice location at the time of receiving signal or TE, now only stationary spins are giving signal in the selected slice however the moving spins have migrated out of selected slice with new nonexcited spins in that place giving no signals

FIG 1B T2W axial image of the patient with bilateral ICA aneurysms shown to explain “Flow voids”–Normal arteries, ie left ACA and basilar arteries (black arrows) are seen hypointense due to flow void, left aneurysm (white elbow arrow) post coiling shows heterogeneous hypointense signals due to partial thrombosis Right patent aneurysm (white arrows) reveals flow voids at periphery due to high flow however central part giving bright signals due to relative stasis of blood

Inflow-related Signal Enhancement (Inflow Effect) As already discussed, in a SE sequence, there is reduction in the signal of blood flowing rapidly out of the measured slice Thus, an opposite effect may occur when spins flow into a slice and it is likely to generate a higher signal than the surrounding tissue- also known as “inflow enhancement” Contrast is generated on T1-weighted imaging, by applying repeated RF pulses having a time interval (repetition time, TR) which is shorter than the T1 relaxation time of the tissue (typically TR < 700 msec) Thus, it results into unequal saturation of the tissue components, depending on their individual T1 times

This is the basic principle of generation of T1-weighted image contrast On a normal T1-weighted image, irrespective of flow effects, blood in the vessels would appear hypointense owing to its relatively long T1 time With reduction in TR, the signal emitted by the tissue diminishes In a GRE sequences, repetition times shorter than 50 msec can be achieved, allowing the majority of non-moving spins to become saturated; thereby, decreasing the background signal The RF pulses do not influence the spins outside the excited slice (or volume) As a result, fully relaxed blood enters into the slice being imaged, without experiencing not more than a few excitations on its way through the slice

Consequently, significantly higher signal intensity relative to that of the saturated spins in the stationary tissue is generated by the flowing blood producing the so-called “inflow enhancement” or “flow-related enhancement” The signal intensity of flowing blood increases with decreasing slice thickness, and increasing flow velocity

Maximal flow enhancement is achieved, if the blood flow velocity is so high that all vessel spins are replaced by unsaturated spins in the time interval TR As a result, the vessel appears bright on a gray or black background Although the inflow effect is seen with both SE and GRE sequences, SE sequences are not practical for the TOF method This is because of the fact that the competing washout effect tends to overbalance the inflow effect at higher flow velocities in SE sequences, leading to decreased flow signal Hence, production of bright blood images by virtue of flow-related enhancement using GRE sequences is the basis of time-of-flight angiography

Phase Effects Phase effects concern the transverse component of the magnetization Phase effects are produced whenever spins are moving in the presence of magnetic field gradients, as are applied for spatial encoding of the MR signal Depending on gradient strength and spin position, magnetic field gradients provoke a change in the Larmor frequency A phase shift of the transverse magnetization can be induced by a gradient pulse of certain length and amplitude This phase shift can be compensated by a second gradient pulse with identical strength and duration but having opposite sign Thus, a net phase shift of zero is encountered with stationary spins On the contrary, a non-zero phase shift is generated by the same gradients applied on a flowing spin The first spin induced phase shift cannot be completely compensated by the second gradient pulse as the spins change their position during the bipolar gradient application The remaining phase shift Φ is proportional to the velocity component v of the spins along the gradient direction On standard MR imaging, this flow-induced phase shift results into a spatial mis-encoding of the signal leading to ghost artifacts, which are typically found in the phase-encoding direction Spins in a blood vessel are moving with different velocities Often, a parabolic flow profile is found Larger phase shift is experienced by spins moving faster than those moving more slowly If there is a velocity distribution inside a voxel, phase dispersion (intravoxel dephasing) occurs leading to reduced signal in the blood vessel The strength and time interval of the gradient pulses, as well as the distribution of spin velocities are determinant for extent of spin- dephasing A very broad spectrum of velocities within a voxel can be seen with complex flow patterns, for example in vessels with turbulent flow, leading to total signal loss in the vessel

Using additional gradient pulses of appropriate amplitude and duration, flow-induced phase shifts can be compensated, thus eliminating any signal loss This technique is called “gradient motion rephasing (GMR)” or just “flow compensation”1 However, GMR is normally restricted to first-order movements, ie spins that move at a constant velocity Turbulent flow and effects of acceleration cannot be completely compensated by GMR Optimal reduction of flow-induced phase effects can be achieved by combining GMR with as short a TE as possible, in order to reduce the time available for spin dephasing Short echo times also diminish the impact of pulsatile blood flow and turbulence (Fig

Noncontrast MR Angiography (NC-MRA) Techniques Several techniques of NC-MRA are available in current practice and they can be classified into techniques exploiting flow-related enhancement, those using the phase of MR imaging signal from moving blood, and those exploiting the physical properties of blood, independent of blood flow2 These techniques are shown in the Flowchart 1 and described in detail thereafter

FLOW-RELATED ENHANCEMENT DEPENDENT NONCONTRAST MRA TECHNIQUES Time-of-Flight Angiography Techniques Inflow effect is the basis of the contrast mechanism of time-of-flight (TOF) MRA Bright signal is produced by fully relaxed blood entering the measured volume as it behaves as an endogenous contrast agent However, flow rephasing techniques (GMR) are required for the bright depiction of flowing blood, to overcome the effects of transverse magnetization induced spin dephasing GRE sequence based TOF MRA offers several advantages FLOWCHART 1 Noncontrast Mr angiography (Nc-Mra) Techniques

FIG 2 Diagrammatic representation of phase-contrast MRA pulse sequence, note the velocity encoding gradient are in slice direction

GRE sequences does not suffer from wash-out phenomenon which results in decreased signal of fast flowing blood as seen with SE techniques

A short repetition times (TR < 40 msec), necessary to efficiently saturate stationary tissue can be used in GRE techniques

Further reduction in spin dephasing can be done by keeping a short echo times (TE < 5 msec), In order to avoid opposed-phase effects at the vessel walls that could potentially impair the depiction of small vessels, generalized it is advisable to apply in-phase echo times The shorter acquisition time of GRE techniques are useful for acquiring volume (3D) datasets

Sequential 2D Multislice Method Multiple thin slices are imaged sequentially in this technique for depiction of the vessels This technique offers two advantages compared to the interleaved multislice technique a The inflow effect can be boosted by the use of very short TR times

Even in the vessel with slow flow, sufficient inflow enhancement is guaranteed with this technique and produces good contrast vessel-background contrast in the covered area, as each slice is an entry slice3 However, 2D TOF MRA can be problematic in case the direction of blood flow in the vessel of interest is not perpendicular to the plane of imaging Signal loss may occur due to the partial saturation of flow in case the vessels run partly inside the slice (in-plane) or return to the slice in a looped form (Figs 3A and B) A slice thickness of at least 2–3 mm is necessary to achieve an optimum signal-to-noise ratio, However, because of larger voxel size required, it results in increased spin dephasing and reduced spatial resolution Because of non-rectangular profile of 2D slices, they exhibit signal variations at the edges leading to step like artifacts in maximum intensity projection (MIP) reconstructions However, overlapping of the slices can be used to overcome this effect The extent of inflow enhancement varies during the cardiac cycle in vessels exhibiting highly pulsatile flow Ghost images of the vessels are produced by the periodic change of inflow enhancement Slow flow in the diastolic phase can lead to saturation of blood spins ECG triggered 2D TOF sequences may be used to overcome many of these problems by restricting data acquisition to the phase of maximum inflow

Thus, accentuation of blood signal and elimination of ghost artifacts can be achieved The mapping of vessel in each slice with equal intensity can be achieved by synchronizing data acquisition with the cardiac cycle However, this technique suffers from the drawback of longer acquisition time

3D TOF MRA-single Slab and Multislab Methods Simultaneous excitation the entire imaging volume, usually 30–60 mm thick, is done and then partitioned into thin slices by an additional phase encoding gradient along the slice-select direction4 Excellent spatial resolution coupled with high signal-to-noise ratio are the major advantages of 3D TOF MRA technique and thus it facilitates improved depiction particularly of small vessels Isotropic voxels can be acquired easily with this technique and slice thickness of less than 1 mm thickness can also be achieved However, 3D TOF suffers from the problem of progressive saturation which happens when blood flowing through the volume is subjected to repeated RF pulses It results into continuous drop in the signal intensity in the direction of flow The extent of saturation is determined by the length of time in which the blood stays inside the volume In slow flow vessels, diminution of signal begins to occur after covering only a short distance On the other hand, the signal remains visible for a greater distance in faster flowing blood Hence, matching to the size of the vessel in the region of interest, the maximum volume thickness should be kept as small as possible Saturation reduction can also be achieved by increasing the TR Subdivision of the volume of interest into several thin 3D slabs that are acquired sequentially can be used for imaging Larger vessel sections (Table 1)5 FIGS 3A AND B 2D TOF MRA (A) shows loss of flow-related signal in M1 segment of left MCA due to in-plane saturation however in same patient lt MCA showing normal caliber and flow related signal on 3D TOF MRA (B)

Multiple overlapping thin slab acquisition (MOTSA) is one such multislab technique, which retains the advantages of 3D TOF; however has the reduced saturation effects like 2D TOF Generally, to avoid saturation within the slabs, the chosen slab thickness has to be small enough However, approximately 20–30% overlap between adjacent slab are required in order to compensate for signal attenuations arising at the slab edges due to non-rectangular excitation profiles However, in this technique time efficiency is compromised leading to a longer overall acquisition times Frequently, owing to the high sensitivity to slow flow 2D TOF MRA is favored for imaging veins On the contrary, for fast arterial flow and for those cases where high spatial resolution is required 3D TOF MRA is more appropriate (Figs

2D TOF 3D TOF Less susceptible to saturation effect More susceptible to saturation effect Visualize slow as well as fast flow Good for high flow Low resolution and SNR High resolution and SNR Better for relatively larger FOVs Smaller FOVs Used mainly for cerebral venographies, abdominal and peripheral angiographies Cerebral angiography FIG 4 2D TOFMR venography reveals adequate visualization of cerebral venous system

FIGS 5A AND B 3D TOF MRA (A) showing very good cerebral vascular anatomy with high spatial resolution and SNR in comparison to 2D TOF MRA (B) in same patient

Important Points for Optimization of TOF Angiography6-8 The slices or volume should be oriented perpendicular to flow direction

Pitfalls with TOF MRA Methemoglobin in thrombosed vessels (cavernous sinus thrombosis) may mimic blood flow (ie

vessel patency) Work around—compare MIP with pre-contrast T1 images or use phase contrast MRA Tissue having short T1 time (blood, fat, tissue that can take up contrast) may simulate vessels Pulsation artifacts in CSF may mimic vessel lesions Overestimation of stenosis and artifacts in the depiction of aneurysms can occur due to signal loss because of turbulent or very slow flow

Signal loss can occur due to magnetic susceptibility artifacts (coils, clips) Signal loss can also occur in case of in-plane flow (2D) or slow flow (3D) After contrast administration, overlap of arteries and veins can occur, especially in intracranial MRA

Inflow Inversion Recovery–Magnetic Resonance Angiography (NC-MRA) In this technique of NC-MRA, stationary background tissues are suppressed by an inversion RF pulse, while the inflow of fully magnetized arterial spins is imaged9,10 A slab-selective inversion pulse followed by an inflow time (TI) of several hundred milliseconds is applied in this technique The TI is chosen to provide adequate inflow into the target vascular anatomy and a sufficient degree of background suppression A rapid readout is then used for imaging Chemical shift-selective fat saturation, short-tau inversion recovery, or water excitation RF pulses are different techniques used for fat suppression TimeSLIP, NATIVE TrueFISP, Inhance Inflow IR, and B-TRANCE are few trade names for the IFIR sequence

Vascular beds containing brisk flow are best imaged using IFIR technique as angiographic contrast depends on inflow The imaging readout is either balanced steady-state free precession (bSSFP) or flow-compensated fast spin echo (FSE) to maintain high signal from flowing spins The upstream edge of the inversion pulse is positioned close to the arterial bed under interrogation to minimize the inflow requirements Respiratory motion is compensated by gating by prospective navigator or bellows

One of the variation of IFIR technique is outflow MRA where a nonselective inversion RF pulse and then a slice-selective inversion pulse to the inflowing blood pool is applied before its entry into the target anatomy The background signal suppression and image quality can be improved in outflow MRA in some applications compared to inflow MRA11 Easy execution and good image quality are main strengths of inflow inversion recovery, especially for renal angiography while poor display of vessels with slow flow is its main drawback

Quiescent Interval Single-shot Magnetic Resonance Angiography Another NC-MRA technique that is used for imaging lower limb arteries is Quiescent-inflow single-shot (QISS) angiography,12 which applies For suppressing the static tissue-an in-plane RF saturation pulse For suppressing venous signal—a tracking venous saturation RF pulse

A waiting period (or quiescent interval) of ~230 milliseconds needs to be applied To ensure arterial inflow into the imaging slice- Fat suppression done by chemical shift-selective fat saturation and A single-shot Cartesian 2D bSSFP readout

ECG gating is done in order to ensure that the systole is overlapped by the quiescent interval while data is acquired during diastole One slice is acquired every heartbeat Thus, a peripheral MRA from abdomen to the feet can be completed within 10 minutes using 3 mm slices

A variant of QISS using a highly undersampled radial k-space trajectory can be used to acquire multiple slices within a single heartbeat, shortening scan times by a factor of 2–313 Compared with single-slice QISS, the thickness of the in-plane saturation RF pulse is increased to span all slices that are acquired in each heartbeat Fat saturation is applied before the bSSFP readout for each acquired slice

Drawbacks Suboptimal display of flow parallel to the imaging slice and poor display of retrograde flow is major drawback of QISS Partial volume averaging in the slice direction can be overcome by acquiring thinner slices, at the expense of acquisition time

Arterial Spin-labeled Magnetic Resonance Angiography Acquisition of 2 image sets which are almost identical except for magnetization of inflowing arterial spins are required in Arterial spin labeling (ASL) MRA technique The 2 image sets are later subtracted

The difference in arterial magnetization is imparted by a process known as labeling Subtraction of the 2 image sets displays the labelled arterial inflow and suppresses background signal

An inversion RF pulse is applied to achieve labeling of arterial spins during ASL MRA This variant of ASL, commonly known as pulsed ASL, applies an inversion RF pulse to arterial spins flowing into the target vasculature, followed by a TI time to accommodate inflow, and an imaging readout The difference between this data set and its counterpart acquired without the labeling RF pulse depicts the labeled arterial anatomy Signal targeting with alternating RF (STAR)14 and flow-alternating inversion recovery (FAIR),15 are two commonly used pulsed ASL approaches

Other ASL techniques apply a quasicontinuous or pseudocontinuous stream of RF energy to label arterial spins flowing through a thin (~1 cm thick) labeling plane16 The signal/noise ratio (SNR) of vessels near the labelling plane where short inversion times are realized can be improved with pseudocontinous ASL, compared with pulsed ASL17 The two techniques of ASL may be combined to improve SNR and vascular coverage18,19 Extended, multiphase readouts can also be used in conjunction with ASL to display the temporal passage of arterial spins20 Strengths High arterial contrast

PHASE DEPENDENT NONCONTRAST MRA TECHNIQUES Phase Contrast Angiography The other noncontrast MR angiographic technique is phase contrast MR angiography21-23 Velocity information can be obtained with this technique by measuring changes in the phase of protons

Currently its primary applications are in cardiovascular flow, venography, and three-dimensional (3D) vascular studies The phase contrast study can also be used for CSF flow studies Phase contrast angiography can provide informations that can be used in two ways (1) for production of routine angiographic images and (2) for quantification of flow

This technique is not routinely used in clinical practice as both TOF-MRA and contrast enhanced angiography are faster techniques Thus, it is used in practice when the results of other angiographic studies are equivocal or physiological information about flow is required To be precise, it is the most sensitive sequence for demonstration of slow flowing blood and it is the only MRA sequence that can provide data regarding flow, velocity and pressure

When an MR echo is recorded, two types of information are acquired-phase and magnitude information A corollary can be drawn with a vector where, magnitude determines the length of the vector while phase determines the direction of the vector (from 0–360°) For generating most information the phase information is not regarded and the magnitude images are displayed only; although, the phase information can be useful in some conditions For example, phase information is frequently used to localize signal in the phase-encoding direction Depending on their location along the gradient, protons along the phase-encoding gradient acquire different phases Precession of a proton at the weaker end of the gradient will be slower as it experiences a 149-T magnetic field, compared to the precession of a proton in the middle of the gradient that experiences a 15-T magnetic field The protons develop different phase shifts, owing to these differences in precession speed, which can then be used to map the location of the proton within the body This same principle is used in phase contrast sequences, where special gradients are used to eliminate the phase differences between stagnant protons while accentuating the phase differences in moving protons In phase contrast sequences, moving protons experiencing a changing magnetic field accumulate a phase which is different from stationary protons, experiencing a constant magnetic field Because faster-flowing protons travel a farther distance through the gradient than slower-moving protons, their phase differences (compared with stationary protons) will be greater than those of the slower flowing protons Thus, briefly speaking, phase differences are proportional to a proton’s velocity (that can be calculated) in phase contrast imaging Additionally, phase shift is also proportional to the magnitude of the gradient, hence, more phase shift per distance travelled is produced by larger gradients Thus, gradient size can be adjusted to manipulate the degree of phase shift for a given velocity As phase shifts greater than 180° result in aliasing, the ability to adjust the gradient is important

Fortunately, manual calculation of gradient size is not required; instead a velocity encoding (VENC) value needs to be entered, using which the computer can calculate the size of the gradient Before acquisition of any phase contrast sequence the VENC value must be defined It is defined as the highest velocity which can be measured correctly by the sequence before aliasing takes place In other words, the gradient is set in such a way by the computer so the VENC value (maximum projected velocity) will result in a ±180° phase shift The closer the VENC is set to the measured velocity, the more accurate the measurement Aliasing will occur when a too low VENC value is set If a too high VENC value is set, accurate measurement of slow velocities will not be possible and there will be a drop in SNR The ideal VENC value for arterial blood is usually more than 100 cm/s and can be as high as 300 or 400 cm/s, depending on the vessel being imaged and on the degree of stenosis On the other hand, the VENC value for imaging the veins and the venous sinus should usually be set around 20 to 30 cm/s Adjustment of the VENC value is analogous to adjustment of the maximum velocity on Doppler ultrasound For detection of slow flowing blood, phase contrast is the most sensitive MR technique as the VENC value can be manipulated

Bipolar gradients are used to acquire images in phase contrast sequence, which is a GRE sequence The two lobes of the bipolar gradient, are often referred to as the dephasing and rephasing lobes They are identical in magnitude and time; however, their directions are exactly opposite

Stationary protons experience two gradients that are exactly opposite to each other, resulting in a return to a net 0° phase and it happens irrespective of the position of the proton along the gradient

On the other hand, moving protons, will change locations in between the two lobes of the bipolar gradient, resulting into two different magnetic field strengths which influence the phase of the proton

This difference causes accumulation of a net change in phase relative to the stationary proton In practice, magnetic field inhomogeneities, that causes unwanted phase changes in stationary protons, cannot be accounted for by a single bipolar gradient Application of a second mirror image bipolar gradient is required to eliminate magnetic field inhomogeneities The second gradient is identical to the first; however is in reverse order Thus, a flow-sensitive sequence, similar to digital subtraction angiography, is generated A stationary proton that accumulates phase during the first bipolar gradient (secondary to magnetic field inhomogeneities) will then acquire exactly the same phase during the second reversed bipolar gradient Thus, after subtraction, the stagnant spin will have a net phase of 0°

However, the moving spin will accumulate phase of same magnitude but of opposite direction of phase during the second bipolar gradient compared with the first When subtracted, the opposite phase shifts will result in a net doubling of the phase shift However, the acquisition time is almost double compared to TOF-MRA as echo has to be obtained after each bipolar gradient Moreover, only one dimensional flow (oriented parallel to the axis of the gradient) is detected Hence, if we aim to obtain flow in all three dimensions, necessary for producing conventional angiographic image, the sequence must be repeated three times, making it prohibitively long

Quantification of flow can be estimated either in one dimension (through-plane flow) or two dimensions (in-plane flow) With through-plane flow, the velocity-encoding gradient is oriented along the slice selection axis (craniocaudal for an axial scan) and information is obtained about flow that is moving through the plane of the image For in-plane flow, the gradients are along the frequency- and phase- encoding axes (eg, transverse and anteroposterior dimensions for an axial slice) Information about flow occurring in the plane of the scan is obtained

A characteristic angiographic appearance similar to fluoroscopic digital subtraction angiography images is generated by phase contrast images It happens because two separate datasets are obtained and then subtracted from each other to create the flow quantification image Intermediate gray signal is assigned to all stationary spins (0° phase shift), while based on the direction of flow the mobile protons are then assigned either hypointense or hyperintense signal, with an intensity level correlating to the velocity magnitude Air containing body parts (which are relatively devoid of signal), like that found external to the patient, in the lungs, and in the paranasal sinuses, exhibit a static or snowstorm appearance, because of the background noise and random movement of protons Figures 6 and 7 demonstrate intracranial MRV and MRA acquired with phase contrast technique while Figures 8A and B demonstrate superiority of PCMRV over TOF MRV for depiction of intracranial venous sinuses

Strengths Ability to calculate velocity and flow data is the major advantage of PC-MRA Cardiac gated through plane images are acquired for obtaining dynamic information throughout the cardiac cycle Both velocity and flow information in relation to time can be determined with the placement of a region of interest and the use of commercially available computer software

Using the Bernoulli equation these flow data can then be used determine intraluminal pressures, which can be used to evaluate the degree of a vessel stenosis (eg, as in aortic coarctation)

Finally, analysis of higher-order flow data like pulsatility and jerk can also be done using more complex gradients Although, this application is seldom used in clinical practice

Drawbacks Suffers from the drawback of longer scan time (2-4 times TOF-MRA) because of need for acquisition of multiple image sets

Phase contrast imaging is highly sensitive to motion degradation like DSA, and usually a nondiagnostic image is produced in case of patient’s movement

Finally, protons in the areas of turbulent flow swirl rather than having a unidirectional movement, resulting in a signal void precluding the acquisition of velocity data Turbulent flow is encountered usually in area just distal to a site of high-grade stenosis or at the vessel bifurcation

Eddy currents, concomitant gradients and spatially varying background phase offsets cause phase imperfections- the other drawback of PC-MRA24 Optimization VENC should be adapted to maximum flow velocity

Four-dimensional Phase Contrast Imaging Blood flow in 3 spatial dimensions over the cardiac cycle can be displayed and quantified by the novel technique of 4D PC-MRA25 Velocity-encoding data are acquired in three directions in this technique compared to conventional phase contrast technique Information about spatial and temporal evolution of 3D blood flow in any particular vascular bed can be assessed by this technique However, this technique also suffers from the drawback of longer acquisition time as velocity encoding needs to be done in three directions

FIGS 8A AND B Phase contrast 3D MRV (B) showing better cerebral venous sinuses and smaller veins than 2D TOF MRV (A) in same patient

However, novel techniques like parallel imaging and other k space undersampling techniques have made it possible to acquire 4D flow MRA at an acceptable scanning time of ~5–10 minutes for different body parts like brain, chest and abdomen

Gated Subtractive Magnetic Resonance Angiography The differential appearance of arteries between systole and diastole provides the angiographic contrast in this technique of NC-MRA26 Arteries are rendered dark at the time of systolic fast flow, because of arterial spin dephasing during the echo train On the other hand, while imaging in slow flow during diastole arterial signal is preserved Arterial anatomy is displayed by subtraction The flow sensitivity of these techniques, which use FSE readouts, can be controlled by the length of the echo train, the flip angle of the refocusing RF pulses, supplementary gradient lobes for flow compensation or spoiling, and the orientation of the frequency-encoding axis27,28 Fresh blood imaging, Non Contrast MRA of ArTerIes and Veins (NATIVE), Sampling Perfection with Application optimized Contrast using different flip angle Evolutions (SPACE), Inhance Delta Flow, and Triggered Angiography Non-Contrast Enhanced (TRANCE) are few of the Commercially available gated subtraction techniques

Imaging readout is used to accomplish dephasing of arterial spins in FSE-based gated subtraction MRA However, dephasing of arterial spins during systole is achieved by applying a flow-suppressing motion-sensitized driven equilibrium (MSDE) magnetization preparation before the imaging readout, when imaging is carried out with a bSSFP readout which preserves signal from flowing arterial blood

These MSDE preparations spoil signal from arterial blood based on its velocity29 or acceleration,30 leaving signal from stationary or nonaccelerating tissue largely unaltered The degree of flow spoiling depends on the timing, strength, and duration of the gradient lobes applied within the MSDE preparation, and when the preparation is applied within the cardiac cycle

Operator involvement to synchronize the acquisition to phases of maximal and minimal flow and tune the degree of flow spoiling (ie to optimize arterial dephasing and minimize venous contamination)

FLOW INDEPENDENT NONCONTRAST MRA TECHNIQUES Angiographic contrast is derived by using the T2 and T1 properties of the blood pool in the non flow- dependent methods of noncontrast MRA Spin-echo and gradient-echo readouts were used in initial reports of flow-independent MRA techniques31,32 However, balanced steady state free precession (bSSFP) is the predominant readout for flow-independent MRA in modern practice

Drawbacks High signal from fat and fluids, which may preclude routine use of standard maximum intensity projection (MIP) processing and dark band artifacts from B0 inhomogeneity However, chemical shift- selective fat saturation can be used to achieve fat suppression Otherwise, the phase difference between fat and water can be exploited to achieve fat suppression Inversion-recovery and T2 magnetization preparations can be used for suppressing signal from fluids and muscle tissue respectively

Subtraction-based bSSFP techniques may also be used for flow-independent angiography Better extraneous nonvascular background tissue signal suppression of can be achieved with subtractive technique compared to non-subtractive technique Signal targeting alternative RF and flow- independent relaxation enhancement (STARFIRE)33 is one of the subtractive flow-independent MRA technique, that preferentially demonstrates blood based on its long T2 and T1 Sensitivity to subtraction artifact due to motion, and increased acquisition times caused by the need to collect 2 data sets are drawbacks of STARFIRE technique

Contrast-enhanced MR Angiography Contrast-enhanced MR angiography (CE-MRA) has evolved as the method of choice for vascular images The T1shortening effect of gadolinium based contrasts are exploited by this technique contrary to the flow based TOF and PC-MRA technique relying on the inherent motion of blood flow for generating vascular signal34 A gradient system with high slew rates for ultrafast acquisition is required unlike 3D TOF MRA, along with additive tools for proper scan timing with regard to the bolus arrival35 The basic sequence used for CE-MRA is ultra short TR/TE 3D spoiled gradient echo sequence

The TR is made as short as possible to speed up the image acquisition and capture the first pass of contrast The presence of Gd makes the blood immune to saturation, so the short TR tends to saturate only background tissue In general, the highest flip angle achievable by the MRI machine is used The TE is also made as short as possible—ideally approximately 10 msec, so as to limit any flow-induced dephasing effects36 For achieving optimal contrast of the arteries different techniques like bolus timing, automatic bolus detection, bolus tracking and care bolus, can be are used37 The peak gadolinium concentration, and hence the achievable arterial SNR is affected by the rate of injection Generally, higher arterial SNR can be achieved with a faster injection rate, but it is associated with shorter bolus duration and earlier venous enhancement Most CE MRA can be acquired with injection rate of 2 mL/sec with higher rates demonstrating very little advantage38 Most CE MRA examinations can be optimally performed with a contrast dose of 015–02 mmol/kg (typically 20–30 mL) If good timing is achieved, optimum images can be generated using even a lower dose (01 mmol/kg) In general, additional advantage of prolonging the arterial phase and providing the operator with an additional buffer to compensate for errors in timing can be obtained if a larger contrast agent dose is used38,39 Administration of saline flush is also an important consideration for optimum CE-MRA technique

Practically, a large saline flush (at least 30 mL) needs to be used for all CE MRA examinations It ensures administration of the entire contrast volume beyond the tubing and bolus transit through the peripheral veins into the right heart Thereby, sufficient gadolinium concentrations in the more distal arteries is ensured The increase in the slope of the enhancement curve, as well as the duration of the arterial phase of the bolus (up to 50%), and delaying significant venous enhancement, the factors which are preferrable for arterial CE-MRA, can be achieved with a large saline flush38,40 The goal is to record the central region of k space during the maximum enhancement of the artery

The center of k space contains the lowest (spatial) frequency wave data, so it represents the major structures of the image and thus most of the gross image form and contrast; therefore, the center of k space should be acquired during the time of highest contrast agent concentration41 Also, for preventing ringing artifacts, arising in the Fourier transform, a high rate of change of the contrast agent concentration during the acquisition of central k space should be avoided41 So, the enhancement is maximized by timing the contrast agent injection such that the period of maximum arterial concentration corresponds to the k space acquisition37 Special considerations with respect to timing must be undertaken with certain vascular problems such as aneurysms since the flow can be much slower through an aneurysm, more time must be allowed between the injection of contrast agent and the image acquisition41 A high resolution with near isotropic voxels and minimal pulsatility and misregistration artifacts should be striven for The postprocessing with the maximum intensity projection (MIP) enables different views of the 3D data set (Figs 9 to 11)

CE-MRA Advantages37 The 3D MRA can be acquired in any plane, which means that greater vessel coverage can be obtained at high resolution with fewer slices (aorta, peripheral vessels) The possibility to perform a time resolved examination (similarly to conventional catheter angiography) Most of the artifacts of time of flight angiography or phase contrast angiography can be reduced or eliminated by CE-MRA

Minimal invasiveness with no associated risk of neurologic complications, reduced patient discomfort and inconvenience, greater cost savings No use of ionizing radiation; paramagnetic agents have a beneficial safety The CE-MRA does not prove to be better in evaluation of carotid artery atherosclerotic stenosis42 however it appears to give better depiction of post coiling larger residual/recurrent aneurysms43 Other very important use of CE-MRA is that it differentiates between active and inactive disease in case of Takayasu arthritis by showing the intense enhancement of vessel wall44 CE-MRA Disadvantages The time window for scan acquisition is limited in the cerebral circulation, owing to the short arteriovenous transition time leading to early enhancement of venous structures Thus, a trade-off is therefore necessary between scan time, volume coverage and spatial resolution Another drawback is the possibility of a false neck remnant in post coiling patient with a intracranial aneurysm, which may be explained by the peripheral contrast enhancement of the organized thrombus or by the vasa vasorum within the aneurysm wall35,45 Time Resolved/4D CE-MRA Vascular malformation imaging work up always needs a dynamic depiction of the disease However, the conventional MRA, dynamic TOF, PC and CE-MRA lacks the ability in showing the different angiographic phases as early arterial, arterial and venous At the same time conventional angiography suffers from motion artifacts specially in unstable patients

The time resolved CE-MRA includes combination of k space segmentation into central, mid, and peripheral zones with sampling the central zones of the k space at a more frequent rate

Scan parameters used for rapid acquisition short TE of 09 ms and a short TR of 22 ms, fractional echo acquisition, 26 mm section thickness MR contrast is injected at a rate of 2 mL/s to a total of 15 mL, followed by 30 mL of saline Sagittal or coronal section are obtained with a FOV of about 28 × 224 cm, matrix of 192 × 192, NEX of 075, and flip angle of 20° at every 18 to 20 seconds About 30 to 35 section are acquired (Figs 12A to D)46 Advantages of Time Resolved MRA Useful for unstable and moving patients47 Better depiction of AVM feeders, nidus and venous drainages48 Gives an idea about flow dynamics of AVM/AVF and aneurysms47 Good for follow-up of postembolization or gamma surgery residual AVMS48 FIGS 9A AND B Contrast enhanced MRA showing good neck and cerebral arterial anatomy

FIGS 10A TO D Case of superior sagittal venous sinus thrombosis showing nonvisualization of sinuses on PC 3D MRV (A and B), postcontrast T1W (C and D) confirms the filling defects in sinus

Limitations High dose of contrast is required Time taking sequences Nonvessel selective on the contrary to catheter angiography

Parallel Imaging35 Recently developed parallel acquisition techniques (PAT) are now commonly used for vascular imaging Multichannel coil arrays are used in PAT for shortening the measurement time by reduction of the number of phase encoding steps Instead the sensitivity profiles of the coil elements are used to extract the spatial information

Sensitivity encoding (SENSE), simultaneous acquisition of spatial (SMAS) harmonics, and generalized auto-calibrating partially parallel acquisitions (GRAPPA) are few of the many PAT image acquisition algorithms that have been described

When used in intracranial 3D TOF MRA, PAT with an eight channel phased-array head coil has been used to reduce 43 percent of measurement time without comprising the image quality However, one must be aware that use of PAT is associated with reduction of SNR, by approximately the square root of the acceleration factor Owing to the inherently high SNR of CE-MRA, PAT is particularly suitable for CE-MRA

In arterial CE-MRA, PAT can be exploited for accelerating the acquisition (and thus avoiding venous contamination), to increase spatial resolution and/or volume coverage, or to combine the two35 Compressed sensing (CS) is another newer technique which is based on k-space undersampling49 can also be used for acquisition of TOF-MRA at a reduced time of scan Few studies have demonstrated the use of compressed sensing based MRA for intracranial vessels in cases of intracranial aneurysms and Moyamoya disease and shown the image quality can be comparable to that of conventional TOF MRA50-52 Compressed Sense (CS-SENSE) is a further development in which compressed sensing is combined with the parallel-imaging or SENSE infrastructure53 CS-SENSE can be used to achieve further reduction of time and can be applicable for TOF-MRA intracranial and extracranial vessels Figures 13A to F show use of CS-SENSE technique for TOF MRA intracranial vessels

FIG 11 CE-MRA bilateral lower limbs showing adequate visualization of lower limb arteries, early draining veins are seen on left side due to underlying vascular malformation

APPLICATIONS OF MRA High spatial resolution, vessel to background ratio and sufficient coverage for accurate assessment of the vessels of interest at a reasonable time of acquisition are few of the basic requirements of successful MRA in routine clinical practice A brief review of applications of NC-MRA in different vascular bed is outlined below

Neurovascular Imaging Limited volumetric coverage with high spatial resolution is desired for intracranial MRA Flow dependent techniques of MRA face the challenge of multidirectional flow in intracranial MRA as it can affect the vessel to-background contrast TOF MRA is the predominant technique used for intracranial MRA for depicting different vascular pathologies of brain