Working principle of MRA and TOF MRA

Presented by :

Nisha Karna

M.Sc. MIT 1^st year

Introduction

​

• 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).

​

• MRA has been increasing in demand because of its physiological nature on the contrary to CTA and catheter angiography which involves catheterization, radiation and nephrotoxic iodinated contrast agents.

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.
• Laminar flow is flow where the particles move along in concentric sheets and laminae, i.e. 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.

Flow phenomena

​

• Turbulent flow is flow at different velocities which varies, i.e. 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.

Flow phenomena

​

• The moving spins (spins that move during acquisition of data) show different contrast characteristics from the stationary spins.

​

• The flow phenomena are generally categorized into time of flight, entry slice phenomenon and intra-voxel dephasing.

Time of flight phenomenon

• Nuclei must receive both RF excitation and rephasing pulses to produce signal.
• Stationary nuclei always receive both pulses, but flowing nuclei present in the slice for the RF excitation pulse may have exi ted the slice before the RF rephasing pulse is applied.
• Alternatively, they may receive the RF rephasing pulse but were not present in the slice for the RF excitation pulse.
• In both cases, no signal is received from these nuclei.
• This is called TOF phenomenon
• Its effects depend on the type of pulse sequence.

TOF vs TE

​

TOF

• Flow related enhancement increases as:
• The velocity of flow decreases
• The TE decreases
• The slice thickness increases
• High velocity signal void increases as:
• The velocity of flow increases
• The TE increases
• The Slice Thickness Decrease

​

​

​

TOF

• The main limitations of time-of-flight MRA are signal loss linked to spin dephasing when the flow is complex or turbulent (stenosis), when the flow is too slow or oriented parallel to the slice plane and poor signal suppression of the stationary tissues when substances with very short T1 relaxation time are present (fat, blood degradation products) .

​

Entry slice phenomenon

• ESP is related to the excitation history of nuclei.
• Stationary nuclei within a slice become saturated after repeated RF pulses, especially when the TR is short.
• Nuclei flowing perpendicular to the slice enter the slice fresh, as they were not present during repeated excitations.
• This is called entry slice phenomenon or inflow effect as it is most prominent in the first slice of a ‘ stack ’ of slices.

ESP

ESP depends on :

• TR: A short TR reduces the magnitude of entry slice phenomenon.
• Slice thickness: Entry slice phenomenon decreases in thick slices compared with thin slices.
• Velocity of flow: Entry slice phenomenon is decreased as the velocity of flow decreases.
• Direction of flow:
• Co-current flow
• Counter current flow

Direction of flow

Entry-slice phenomenon: slice 1 (most inferior).

ESP

• Entry slice phenomenon increases:
• at the first slice in the stack
• when using a long TR
• in thin slices
• with fast flow
• in counter - current flow

Intra voxel dephasing

• If a flowing nucleus is adjacent to a stationary nucleus in a voxel, there is a phase difference between the two nuclei.
• because the flowing nucleus has either lost or gained phase relative to the stationary nucleus due to its motion along the gradient.
• Therefore nuclei within the same voxel are out of phase with each other, which results in a reduction of total signal amplitude from the voxel.
• Is called intra - voxel dephasing
• The magnitude of intra – voxel dephasing depends on the degree of turbulence.

Intra voxel dephasing

Flow phenomena compensation

• The methods for reducing flow phenomena are:
• Even echo rephasing
• Gradient moment nulling
• Spatial pre- saturation.

Even echo rephasing

• If two or more echoes are produced in a spin echo pulse sequence, intra - voxel dephasing may be
• Is reduced by acquiring the second and succeeding even echoes at a multiple of the first TE;
• can be used to reduce artefact in a T2 weighted image.

Gradient moment nulling

• compensates for the altered phase values of magnetic moments of nuclei flowing along a gradient.
• uses additional gradients to correct the altered phases back to their original values
• Gradient moment rephasing is performed by the slice-select gradient and/ or the frequency-encoding gradient.
• Gradient polarity changes from positive to double negative and then back to positive again.

GMN

GMN

• Gradient moment rephasing predominantly reduces intra-voxel dephasing.
• Gradient moment rephasing assumes a constant velocity and directi on across the gradients at all times. therefore often termed first order motion compensation
• Pulsatile flow is not strictly constant, so gradient moment rephasing is often more effective on venous rather than arterial fl ow.
• As gradient moment rephasing uses extra gradients, it increases the minimum TE.

Without GMN

With GMN

Spatial pre-saturation

• Spatial pre-saturation pulses nullify signal from flowing nuclei so that the effects of entry-slice and TOF phenomena are minimized.
• Spatial pre-saturation delivers a 90° RF pulse to a volume of tissue outside the FOV.
• A flowing nucleus within the volume receives this pulse.
• When it then enters the slice stack, it receives the 90° RF excitation pulse and is saturated.
• If it is fully saturated, it has no transverse component of magnetization and produces no signal

Pre-saturatiom

Pre-saturatiom

• Spatial pre-saturation pulses should be placed between the origin of flow and the imaging volume so that signal is nulled from flowing nuclei that enter the FOV.
• In sagittal and axial imaging, pre-saturation pulses are usually placed above and below the FOV so that arterial flow from above and venous flow from below are saturated.
• Pre-saturation pulses increase the amount of RF delivered to the patient that may increase heating effects and decrease the number of slices available

​

Even-echo rephasing uses balanced echoes in which even echoes demonstrate less dephasing than odd echoes. It

reduces intravoxel dephasing and is mainly used in T2-weighted sequences

Gradient moment rephasing uses additional gradients to correct altered phase values. It reduces artifact from

intravoxel dephasing and is most effective on slow, laminar flow within the slice

Presaturation uses additional RF pulses to nullify signal from flowing nuclei. It reduces artifact due to time-of-flight

and entry-slice phenomena. It is effective on fast and slow flow, and increases the RF deposition to the patient

MRA

• Utilizes artifactual signal changes caused by flowing blood to depict vessel lumen
• May include spin preparation to suppress signal from stationary tissues or discriminate venous from arterial flow
• Does not require exogenous contrast administration, but contrast agents may be used to enhance MRA for fast imaging

​

MRA technique

1. Time of flight MR Angiography (TOF-MRA) or Signal Amplitude Methods.
2. Phase contrast MR Angiography (PC-MRA) or Signal Phase Methods.
3. Contrast enhanced MR Angiography ( CE-MRA)
4. Parallel acquisition techniques (PAT)
5. Time Resolved/4D CE-MRA

​

​

TOF MRA

​

• The contrast mechanism of time-of-flight (TOF) MRA is based on the inflow effect.
• Fully relaxed blood entering the measured volume behaves as an endogenous contrast agent, by producing a bright signal.
• TOF - MRA uses an incoherent (spoiled or T1) gradient echo pulse sequence in combination with gradient moment rephasing to enhance flow.

​

TOF MRA

​

• TOF MRA using GRE sequences has several advantages:
• Firstly, GRE sequences are not affected by the wash-out phenomenon that diminishes the signal of fast flowing blood when using SE techniques.
• Secondly, GRE techniques permit the use of short repetition times (TR < 40 msec), which are needed to efficiently saturate stationary tissue.
• Thirdly, echo times can be kept short (TE < 5 msec), thus further reducing spin dephasing.

TOF MRA

• In TOF, maximal enhancement of flow occurs  when the vessel is perpendicular to the plane of imaging.
•  TOF techniques are thus somewhat insensitive to in-plane flow
• Also, because of saturation effects, maximum slab thickness in 3D TOF MRA may be limited.
• Various modifications of the TOF technique have been developed to reduce in-plane saturation effects and improve visualization of smaller vessels
•  These include the use of magnetization transfer saturation pulses, overlapping of multiple slabs (MOTSA), fat suppression, and variable (ramped) flip angles

TOF MRA

​

TOF techniques can be divided into three groups:

​

• Sequential 2D multi slice method,
• 3D single-slab method,
• 3D multi slab method.

Sequential 2D methods:�

​

• The vessel is imaged by sequentially scanning multiple thin slices.
• Has two advantages in comparison to the interleaved multi-slice technique:
• Firstly, very short TR times can be used which boost the inflow effect,
• And secondly, partially saturated blood is hindered from flowing from one slice to another.

Sequential 2D methods:�

​

• Problems arise with 2D TOF MRA if the vessels to be imaged do not flow in a perpendicular direction to the imaging plane.

​

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)

3D TOF MRA

​

• In 3D TOF MRA, the entire imaging volume, usually 30 to 60 mm thick, is excited simultaneously and then partitioned into thin slices by an additional phase encoding gradient along the slice-select direction.

​

• 3D TOF MRA has the advantage of high spatial resolution together with high signal- to-noise ratio, thereby facilitating the improved depiction of particularly small vessel structures.

3D TOF MRA

​

• One major problem of the 3D technique, however, is the progressive saturation that occurs when blood flowing through the volume is subjected to repeated RF pulses.
• 3D inflow studies are therefore not suitable for imaging venous flow.
• limits the maximum thickness of the slab.

3D multislab technique

• To improve coverage and at the same time maintain the signal from blood flowing within the larger volume
• a hybrid technique known as multiple overlapping thin section angiography
• ( MOTSA ) can be utilized.

​

TOF MRA

MOTSA

• Multiple Overlapping Thin Slab Acquisition, is a hybrid between 2D and 3D TOF techniques
• Involves the sequential acquisition of a several overlapping 3D volumes (or "slabs“
•  Each slab is typically less than 5 cm in thickness, so the number of contained slices is small (e.g. 16-32)
•  Because of this restricted slab thickness, loss of signal due to saturation effects is relatively limited, even at the exit slices
•  MOTSA thus offers a method to cover a relatively large anatomic area using 3D TOF with preserved intravascular signal intensity

MOTSA

• Some variation in signal still occurs at the end slices, so MOTSA extracts only the central portions for each of the overlapping acquisitions to make up the final data set for
• For optimal contrast, each MOTSA slab should contain at least 10 slices processing into the MRA projections.

MOTSA

Because each MOTSA slab is acquired at separate times, exact registration of position and signal intensity of adjacent slabs may not be possible. This gives rise to the so-called venetian blind artifacts

MOTSA

• Although most vendors refer to this technique as MOTSA, Philips calls sequence MultiChunk 3D Inflow MRA
• One or two of the overlapping slices may be retained and averaged with the corresponding slices from the next slab acquisition for correcting venetian blind artifact
•  Other methods such as SLINKY ("SLiding Interleaved K_y) have been proposed that modify reconstruction process have been proposed
• Philips also offers a special algorithm to merge chunk borders and reduce/eliminate venetian blind artifacts called CHARM ("CHunk Acquisition and Reconstruction Method")

Magnetization transfer contrast

•  Magnetization transfer (MT) is the physical process by which macromolecules and their closely associated water molecules (the "bound" pool) cross-relax with protons in the free water pool
• The interaction between the two pools can be modulated by applying radiofrequency (RF) energy exclusively to the b ound pool using specially designed off-resonance MT pulse(s)
• Some of this deposited energy is then transferred to the free water pool primarily via dipole-dipole interactions.
•  Depending on the degree of coupling between the pools, the free water pool becomes partially saturated
• The suppression of background tissue by MT pulses makes the technique especially useful as an adjunct to MR angiography (MRA)

Magnetization transfer contrast

Time of flight MRA without (left) and with (right) MT suppression. Note better visualization of small vessels after MT pulse.

LIMITATIONS OF MTC

• Limitations and disadvantages of MT-assisted MRA include:

1) a slight prolongation of imaging time (extra time is required to perform the MT pulses); and

2) tissue heating due to energy deposition from the MT pulses

​

Magnetization transfer contrast

• All the major vendors use the generic terms magnetization transfer, MT, or MTC to refer to this technique.
• Toshiba, however, uses a much longer but more descriptive trade name, SORS-STC, which stands for "Slice-Selective Off-Resonance Sinc Pulse Saturation Transfer Contrast".

FAT SUPPRESSION

• Since time-of-flight MR images are T1-weighted and reformatted using a maximum intensity projection (MIP) method, high signal from fat may obscure vessels or mimic pathology
• particularly problematic at the skull base where abundant fatty marrow may create artifacts around the carotid and vertebral arteries
•  chemical shift (CHESS), Dixon, or water excitation (WE)

​

TOF MRA (conventional). Note high signal from fat in marrow, orbits and subcutaneous tissues.

TOF MRA using water excitation. The high signal is gone and ophthalmic arteries (arrows) are now well seen.

Fat suppression

•  New TOF saturation technique called BeamSat TOF is available on Hitachi
•  Beam Sat TOF allows users to place a cylindrical beam sat pulse over a specific artery while performing a 3D TOF study over a region supplied by several arteries
• Signal from that one blood supply will be suppressed, allowing location of sources of blood flow to be identified

Variable (ramped) flip angle pulses

• In 3D time-of-flight (TOF) MR angiography flow-related enhancement is highest for vessels first entering the slab. 
•  The farther into the slab they travel, the more these vessels become saturated by repetitive RF-pulses
• results in progressive loss of MRA signal from vessels extending deep into the imaging volume.
•  Ramped RF-pulses are a special class of asymmetric pulses with flip angles that vary as a function of position
• In MRA, flip angle is made to increase linearly along the direction of flow

Variable(Ramped) flip angle

• SIEMENS and PHILLIPS- TONE( “Tilt-Optimized Nonsaturated Excitation“)
• GE-  ramped excitation.
• Hitachi- SSP ("Sloped Slab Profile")
• Canon-  ISCE ("Inclined Slab for Contrast Enhancement")

3D TOF MRA without TONE

3D TOF MRA with TONE shows improved visualization of peripheral vessels (arrow)

TOF MRA ARTIFACTS

STAIR STEP ARTIFACT (2D TOF only)�

• simplest form of this artifact is a subtle pixelated appearance to obliquely oriented vessels
•  occurs because the slices in 2D are relatively thick (1-3 mm
•  Stair-step artifacts can be minimized by overlapping slices by 25-30%

Fig: Mild "stair-step" artifact due to non-isotropic

voxels

Mild motion artifact causing horizontal banding on this 2D-TOF MRA of the aorta

Severe artifacts with jagged edges due to gross motion for carotid TOF MRA study

In-Plane Saturation Artifact

• TOF MRA relies on the inflow of fresh (unsaturated) blood to produce high intravascular signal.
•  When vessels travel within plane, their blood may become saturated like stationary tissues, resulting in decreased signal

​

3D-TOF MRA showing artifactual in-plane signal loss within both middle cerebral arteries

2D-TOF MRA shows artifactual in-plane signal loss in horizontal portions of both anterior tibial arteries

Shine-through Artifacts

• B
• oth 2D and 3D TOF MR angiographic images are displayed using a maximum intensity projection (MIP) algorithm.
•  any other material with high signal intensity will "shine through" and "contaminate" the MIP image

3D TOF MRA showing shine through of high signal from hematoma (H) and fat (F) at skull base

CSF inflow phenomenon (on source image left) creates artifact on MIP image (right)

Flow reversal artifact

• This is morea of a pitfall rather than artifact
• TOF employs travelling sat pulses to eliminate signal from veins flowing in the opposite direction.
• If an artey has retrograde flow due to some abnormal condition , it will be suppressed as well.

​

2D TOF MRA shows only right vertebral artery. No flow related signal in the left vertebral artery is seen.

Contrast-enhanced MRA shows retrograde filling of left vertebral artery (subclavian steal phenomenon)

Venetian blind artifact(3D mra only)

• Seen only with MOTSA technique
• represents overlap between adjacent slabs that are acquired as separate acquisitions and then fused together

Susceptibility artifacts

• Susceptibility refers to the magnetic field distortion that occurs especially at air-tissue interfaces and near metallic objects
•  In MRA susceptibility artifacts are commonly encountered around surgical clips, endovascular coils, and stents
•  When mild, these artifacts may suggest vascular narrowing or stenosis when none exists.
•  When severe they may cause complete loss of flow-related enhancement falsely suggesting occlusion.

Source image shows susceptibility field distortion due to aneurysm clip

MRA shows spurious loss of flow in entire right middle cerebral artery due to susceptibility artifact from clip

Pitfalls

​

• Methemoglobin in thrombosed vessels (cavernous: sinus thrombosis) may mimic blood flow (i.e. vessel patency)
• Work around—compare MIP with pre-contrast T1 images or use phase contrast MRA
• Short T1 tissues (fat, bleeding, tissue that take up contrast) may simulate vessels
• Pulsation artifacts in CSF may simulate vessel lesions
• Signal loss occurring with turbulent or very slow flow causes overestimation of stenosis and artifacts in the depiction of aneurysms
• Signal loss due to susceptibility artifacts (coils, clips)
• Signal loss in case of in-plane flow (2D) or slow flow (3D)
• Overlap of arteries and veins after contrast administration, particularly in intracranial MRA.

Parallel imaging

​

• Parallel acquisition techniques (PAT) have recently been developed and are now being increasingly used for vascular imaging
• PAT uses multichannel coil arrays to shorten the measurement time by reducing the number of phase encoding steps.
• The spatial information is instead extracted from the sensitivity profiles of the coil elements.
• Sensitivity encoding (SENSE), simultaneous acquisition of spatial (SMAS) harmonics, and generalized auto-calibrating partially parallel acquisitions (GRAPPA).
• However, PAT reduces SNR by approx. square root of the acceleration factor.

Other new non contrast MRA methods

Flow sensitive dephasing

• Is a non-contrast subtraction technique similar to the 3D-gated FSE MRA
• The FSD sequence simultaneously acquires a diastolic image (where both arteries and veins are bright) and a systolic image (where veins remain bright but arteries are dark).
• By subtracting these two images a pure arterial image remains.

Flow sensitive dephasing

FSD module is based on a cluster of RF-pulses in a driven equilibrium configuration. Bipolar gradients (B) induce accelerated arterial dephasing in systole. A spoiler gradient (S) is destroys residual transverse magnetization to prevent unwanted later echoe

Hand MRA using FSD prepared balanced SSFP

Quiescent-Interval Single-Shot (QISS) MRA

• QISS is a cardiac-gated, non-contrast inflow technique bearing some similarities to 2D time-of-flight (TOF) and inflow-enhanced SSFP MRA
• especially designed for peripheral MRA.
• The sequence begins with a pair of closely-spaced 90º-RF pulses, one to saturate signal in the slice to be imaged and the other more distally located to suppress venous inflow.
•  Next comes a quiescent interval (QI) of about 230 ms, during which fresh (unsaturated/fully magnetized) blood enters the imaging slice.
• Then fat suppression pulse is applied to destrpy any fat signal that has recovered during QI.
• Finally, the desired arterial signal is acquired from the slice using a 2D single-shot, balanced SSFP sequence. 

QISS MRA

QISS MRA of the lower exterimities. Note venetian blind artifact at junction

of adjacent slabs

​

​

​

Gated Time-of-Flight (TOF) Inflow MRA

• Simple cardiac gating can improve the appearance of a 2D TOF MR angiogram
•  k-space ordering is adjusted so that the central lines of k-space occur during peak velocity (systolic phase)
• The primary advantage is the reduction of pulsatile flow artifacts causing physical movement of the artery and spatial blurring
• Philips offers a dual-gated method which acquires segmented data both in systole and diastole which may offer advantage in characterizing the degree of stenoic lesions.

Future perspectives

​

• Presently MRA techniques are well established tool for vascular imaging however still there is need of decrease acquisition times, increase temporal and spatial resolution with more dynamic information.
• there are evidences of work under research reveals that TOF MRA at 7T have increase SNR and spatial restoration, especially for the assessment of brain AVM and high grade gliomas.
• Newer techniques like SWI-MRA are also in developmental phase, it makes possible to image both artery as well as vein simultaneously and can be evaluated separately.27 Veins are dark due to T²* effect with SWI processing, same time arteries are seen bright due to TOF inflow effect

Future perspectives

​

• it can also provide the information about hemorrhage and calcification whether parenchymal or mural in case of atherosclerotic disease.

​

• there is lot more in future for MRA application due to advent of ultra-high field scanners and development of advances imaging sequences.

​

PC MRA application of SW imaging, images (A and B) are showing intracranial arterial anatomy due to TOF inflow effect and images (C and D) reveals the venous structures on SWI in same patient

Refrences

• Jerrold T Bushberg , J. Anthony Seibert, The essential physics of Medical Imaging 3^rd Edition
• Catherine Westbrook , MRI in Practice 4^th and 5 th Eduition
• Arun kumar Gupta- Diagnostic Radiology – Recent Advances and Applied Physics in Imaging 2013
• Mri questions answers.com

​

THANK YOU