Pitfalls of MR Angiography due to slow flow in large Aneurysm

This case illustrates a potential pitfall of Non contrast 3D TOF MRA imaging. Contrast enhanced CT study shows a focal aneurysmal dilatation of basilar, but the non contrast 3 D TOF MR Angiography of Brain of same patient fails the aneurysm is not depicted well on lateral view.

The non visualization of aneurysm may be due to slow flow owing to its large size or thrombus formation in the aneurysm because vascular imaging in Non contrast 3D TOF MRA depends on flow related signal from a rapid blood flow in a vessel. In this study the ICAs and basilar show normal flow related signals but the large aneurysm sac appears nearly isointense with nearby static brain.
CT Angiography or MR Angiography with contrast has considerable value in such cases because it is not dependent on flow velocity to create contrast and signals.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

MRI Pitfalls: Jugular vein with high signal

MRI Axial T2w image in Neck region, left side internal jugular vein show high signal where as right side internal jugular show normal T2 flow voids, Thrombosis ? 
No !!! this is artifactual high signal due to slow flow in the left jugular vein.
This is a common finding in MRI Cervical region spine on Axial T2w images and can be passed off as normal. 

It is important to consider the anatomy of the jugular veins here, which is not symmetric.
Internal jugular veins joins sub clavian veins before draining directly into the superior vena cava via innominate veins. SVC being on right side, right innominate is short. The left innominate vein, which is nearly twice as long as the right, must cross the midline to join SVC which is on right side of the mid line.
So due to slow flow or sometimes reversed flow in the left jugular vein which occur secondary to compression of this left innominate vein by the aorta as the vein passes under the sternum, there is artifactual high signal in the left jugular vein.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

Effects of Oxygen on FLAIR

A 30 yo pt brought unconscious, required intubation and O2 support. MRI axial FLAIR brain screening show diffusely increased signal in the subarachnoid spaces in the region of cortical sulci and basal cisterns but Csf examination was normal.

An Artifactual high signal in Csf spaces due to para magnetic effect of inhaled Oxygen.
"Pseudo SAH" 

Patients who are intubated and on inhaled oxygen, once the concentration of dissolved oxygen is high enough in the blood, its level will increase in the CSF as well leading to sub optimal suppression of Csf signal on FLAIR due to para magnetic effect of Oxygen and results in diffuse Csf T2 hyper intensity.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

Neck MR Angiography Pitfall

A young pt with recurrent episodes of TIA advised MR Angiography of neck along with Brain.

Non contrast 2D TOF MRA performed, a MIP image of neck demonstrating of left carotid bifurcation and a row axial image at the level of the ICA origin.

A focal notching in ICA at its origin appears to be:

1. Plaque.

2. External compression.

3. Artifactual.

Answer: it’s an artifact, due to turbulence of flow at the bifurcation as axial sections of contrast enhanced CT Angiography of same patient at the same level are normal. 

Discussion: 

RE CIRCULATION FLOW ARTIFACT on 2 D TOF Non contrast MRA

Carotid bulb is a normal focal expansion of cervical ICA at its origin is patent and relatively prominent in young patient. Again due to vigorous flow there is formation of an eddy circulation in bulb with reversal of flow. This reversed flow leads to diminished signal on 2D TOF MRA because of the linked saturation band that suppresses all caudal flow, regardless of its nature. 

This problem can be obviated by use of contrast during MR Angiography as in contrast enhanced MRA study, flow related signal by contrast is insensitive to direction of flow and demonstrates the normal contour of the carotid bulb.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

Truncation Artifact MRI Dorsal Cord

A 40 yo pt with a history of lower limb weakness referred for mri screening of brain and whole spine for cord. MRI sagittal T2 screening of dorsal region shows a faint uniform linear high signal at the center of the cord. The signal abnormality likely to represent:

(1) Cord demyelination.

(2) Syrinx.

(3) Artifact.

Answer : Its an artifact, known as truncation or Gibbs artifact, as axial T2w images planned at right angle to cord at multiple levels are normal including sagittal T2 images repeated with increase in image matrix and lowering FOV.


Discussion:


Interpreting thoracic spinal cord on MR sometimes can be challenging because artifacts due to pulsations of CSF as well as respiratory motion can project over the spinal cord on T2WI. The artifacts from CSF pulsation can be minimized with the addition of flow compensation techniques, and the nearby fat can be suppressed using STIR or saturation bands.
There is, however, one other artifact that can contribute to artifactual high signal in the spinal cord called a Gibbs artifact. This is attributed to the difficulties of replicating the sharp changes in contrast between adjacent structures like cord and CSF using limited frequency information. While large data sets will allow a closer approximation of the edge, with limited time and data there will be some “truncation” of the information. This artifact is most evident where there are sharp contrast borders and a coarse matrix is used for the image acquisition.
It cannot be entirely eliminated, but it can be minimized by using either a smaller FOV ( Field of view) or a larger and finer matrix. It is not seen often today because fast spin echo imaging and high field scanners has allowed the use of a finer matrix without consuming too much time. As the matrix increases and the pixels are therefore smaller, this artifact is minimized because the transition at any high contrast edge is spread over more pixels.
When the Gibbs artifact is visible, it is located in the exact center of the cord as continuous faint high signal intensity line of uniform thickness on sag T2. Dilation of the central canal or a small syrinx generally should not be mistaken for a Gibbs artifact. A syrinx tends to be more sharply defined, more hyper intense and is evident on both T1WI and T2WI axial imaging. In Demyelination or Ischemia, cord involvement will be focal, multi focal or diffuse, axial T2 sections can be used for confirmation.
This artifact was named after the American physicist J. Willard Gibbs, who was called “the greatest mind in American history” by Albert Einstein. He became a professor at Yale in 1871, so his work predated NMR by nearly 100 years. There is some irony in attaching his name to this artifact since his life’s work focused on the mathematics of what is now called thermodynamics.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

Infratentorial insensitivity of FLAIR

Axial FLAIR and T2w images of same patient at the same level show a lacunar infarct in right half of Pons visible only on T2 and not on FLAIR, attributed to relative low sensitivity of FLAIR for Posterior fossa lesions. 

It is common to have conflicting findings on MR from two different sequences of same patient.
The common example is lesions like lacunar infarcts or plaques of multiple sclerosis are visible on the Proton Density / T2w sequence but not on the FLAIR.
In practice, one should put a much higher value on the proton density / T2w sequence in the posterior fossa due to low sensitivity of FLAIR as far as infratentorial lesions are concerned.
It is not entirely clear why lesions like lacunar infarcts or plaques of multiple sclerosis below the tentorium are not well visualized on FLAIR.
This limitation of FLAIR can be dangerous at times when only FLAIR sequence is offered to radiologist for interpretation, the institutions where only axial FLAIR screening of Brain is commonly advised may be to save time or with intention to save patient’s money.

Multiple sclerosis: Case 2

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

MRI Artifacts, Flow void and Signal void

A young pt admitted to casualty with history of trauma. MR study shows two signal voids in the Sylvian fissures on either side on T2 as well as T1w images.

Diagnosis ?  Bilateral MCA Aneurysms ?

Answer : No, as pt's non contrast 3 D TOF MR Angiography of brain normal, the thin axial CT sections in corresponding region nicely demonstrates that the low signal intensity in sylvian fissures on MR is signal voids due to intra cranial air and not flow void of aneurysms. Note the susceptibility Artifact due to air on MR Diffusion. 

Discussion:

FLOW VOID AND SIGNAL VOID

Low signal intensity in Sylvain fissures like those on T2WI images could be due to signal void of air or flow void aneurysms. The bright rim evident on the diffusion scan indicates susceptibility Artifact that would not be expected in association with an aneurysm. 

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

MR Angiography Artifactual flow loss

A 60 yo male for stroke evaluation.
3D TOF MRA, MCA main stem on either side show focal flow loss or narrowing.
Diagnosis ?
1. Stenoses.
2. Artifactual flow loss.

Answer: Artifactual flow loss, common in elderly non co operative patients due to motion.

Discussion:

3D TOF MRA ARTIFACTS DUE TO MOTION AND VENETIAN BLIND ARTIFACT

This case provides a reminder that MRA of the brain using 3D TOF technique is usually acquired as two to five slabs, unlike the 100 or more thin slices acquired during a 2D TOF sequence. This approach is called MOTSA (multiple overlapping thin slab acquisition). To make 3D TOF images of the intracranial vessels, rather than include all the region of interest in a single slab, multiple thin slabs are acquired with an overlap and then knit together to appear as one continuous volume. The image contrast for both 2D and 3D techniques is still the result of entry slice enhancement, i.e. unsaturated spins coming from outside into an imaged volume, and you may recall that the advantage of 3D TOF imaging in the brain is its improved depiction of curving vessels. These multiple thin slabs are necessary because the hydrogen spins become progressively saturated by the repeated 90 degree pulses as they experience as they traverse the slab. As a result, when using a single, thick slab there would be no signal recovered from the most distal portions of a vessel. These slabs are acquired sequentially i.e. one after the other.
The above MR Angiography demonstrates misregistration between slabs due to some head motion that occurred between two adjacent slab acquisitions. Another cause of this artifact is the loss of flow-related enhancement that becomes increasingly evident as the vessel approaches the exit side of each slab. This phenomenon called Venetian blind artifact, common when the slabs are stitched together.
To obviate this artifactual signal loss modify the pulse sequence to “add” vascular signal at the exit side of the slab. This is accomplished by using a variable flip angle on this gradient echo acquisition that changes during each slab acquisition. Since the vascular signal increases with an increase in the flip angle, a modulated or “ramped” flip angle can be used to correct for vascular signal loss. This approach can help to provide a seamless appearance to the vasculature across the multiple slabs.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD

MR Venography Artifactual flow loss

A 30 y o postpartum woman presents with headaches and drowsiness. The maximum intensity projection (MIP) image from coronal two-dimensional time-of-flight MR venogram (2D TOF MRV) the part of superior Sagittal sinus near torcula show poor flow related signals.
Diagnosis?
1. Superior sagittal sinus thrombosis.
2. Stenosis.
3. Normal.

Answer: It’s a common finding in 2 D TOF MR Venogram due to inflow artifact, can be passed off as normal.

Discussion:


2 D TIME-OF-FLIGHT : IN-PLANE FLOW ARTIFACTS.

2 D TOF technique of MR Venogram for demonstration of intra cranial venous system without need for any intravenous contrast. The contrast between the vascular lumen and the background is due to entry slice enhancement of blood flow. Because the image is acquired as a stack of individual slices, every slice represents an entry slice for blood flow. This stack of thin images transformed into MIP reconstructions to demonstrate MR Venogram. The advantages to using a 2D TOF technique over 3D TOF is its sensitivity to relatively slow flow and almost unlimited coverage, there are some artifacts that have to be considered as a result of its single plane of acquisition.
Entry slice enhancement is most effective when the orientation of the acquisition slice is perpendicular
to the direction of flow. For MR Venogram commonly coronal 2D TOF is used because the venous flow in the brain is largely anterior to posterior.
It need to be noticed that the superior sagittal sinus (SSS) is well demonstrated throughout with the exception of the posterior segment at its junction with the straight sinus at the torcula. While this appearance can be readily mistaken for intraluminal thrombus or stenosis, it is important to recognize that this is an artifact of in- plane flow. Since intravascular signal is highest when flow is perpendicular to the slice and lowest when flow is in the same plane as the slice, the signal in the SSS drops off in this vertical segment of the sinus that is in plane on a coronal acquisition.
It is important to know how the source data for the 2D TOF study is acquired coronal or sagittal in order to anticipate these artifacts.
Acquiring two 2D TOF MRV scans in perpendicular directions or using contrast can help resolve some of the problems created by in-plane flow.

Reference: Practical MR Physics and case file of MR artifacts and pitfalls, Alexander C. Mamourian, MD