Perfusion MRI

Perfusion MRI

Perfusion

Perfusion is a fundamental biological function that refers to the delivery of oxygen and nutrients to tissue by means of blood flow.

Perfusion normally refers to the delivery of blood at the level of capillaries and measures in ml/100gm/min.

Perfusion is closely related to delivery of oxygen and other nutrients to the tissue.

Two main perfusion MRI approaches have been developed:

A. Those using exogenous contrast agent

B. Those without using contrast agent

The first group of technique include Dynamic Susceptibility Contrast (DSC-MRI) and Dynamic Contrast Enhanced (DCE-MRI).

The second group relates to Arterial Spin Labeling (ASL).

DSC-MRI Principle

DSC-MRI is one of  the exogenous contrast based method and relies on IV injection of paramagnetic contrast agent, Gd-DTPA.

This technique utilizes very rapid imaging to capture the first pass of contrast agent, k/a Bolus Tracking MRI.

After the bolus of contrast agent is injected, hemodynamic signals of DSC-MRI depends on T2 or T2* relaxation time and transiently decrease because of increasing susceptibility effect.

DSC-MRI is based on susceptibility changes after injecting contrast agent.

The contrast agent is a paramagnetic material, which distorts the magnetic field and reduces T2 around the vessel because of an increased susceptibility effect.

Two compartments must be considered: intravascular (IV) and extravascular(EV).

When the tracer remains IV, the compartmentalization of contrast agent creates strong, microscopic susceptibility gradients which extend beyond the vessel size.

The contrast agent modifies the blood T2/T2* relaxation rates (R2/R2*).

The changes of T2/T2* (∆R2*) is the subtraction of R2* b/w post and pre-contrast injection.

DSC-MRI Parameter

Here, a bolus of Gd-based contrast agent administered as a short venous injection of 10-15 sec, by the time it reaches the ROI (brain), creating a signal dip of about 10-20 sec or longer.

To faithfully record the tracer conc during this passage, images must be acquired at a rate much faster than the time it takes the bolus to pass through the tissue (i.e. MTT), which is usually of order of several sec.

Adequate coverage (whole brain) with T2*WI at a time resolution TR<2sec needs rapid imaging seq like EPI.

TE is chosen long enough to produce sufficient CNR due to susceptibility effects, but not long enough to dephase all signal during the max contrast agent conc.

A relatively large FA is used, although not long enough to introduce unwanted T1 contamination.

DSC-MRI Signal Dynamics

Serial images are acquired before, during and after injecting the contrast agent.

While passing through the microvasculature, a bolus of contrast agent produces decrease in SI.

The time course images can be divided into 3 stages: the baseline, the first passage of bolus and the recirculation period.

DCE-MRI

This is the other exogenous contrast based method.

After the bolus of contrast agent is injected, hemodynamic signal of DCE-MRI depend on T1 relaxation time, and increase because of T1 shortening effect associated with paramagnetic contrast agent.

DCE-MRI uses rapid and repeated T1W images to measure the change in signal induced by paramagnetic tracer in the tissue with change in time.

In this method, contrast agent is also IV injected to generate bolus.

Shortening of T1 relaxation rate caused by contrast agent is the mechanism of tissue enhancement (so called T1 or relaxivity based method).

This is usually scanned with T1W in 2D or 3D.

GE methods (FLASH) T1W high resolution isotropic volume acquisition are used.

For quantitative DCE, a pre-contrast T1 value for each voxel is often performed.

These values are used to calculate the post injection T1 values for each voxel.

DCE-MRI Parameter

DCE-MRI is scanned with very short TR and TE to generate T1W images.

FA is also small due to short TR.

With current MRI technique, the volume scan time is usually b/w 5-10sec.

For breast and large FOV, it may extend upto 20sec.

So, it may require compromise in terms of spatial resolution, temporal resolution and coverage.

DCE-MRI Signal Dynamics

The time of enhancement is related to the changes, which depend on the physiological parameters of microvasculature in the lesion and on the volume fractions of the various tissue compartments.

For a bolus inj, there is always an initial rise in its conc in the plasma and possibly some leakage into the interstitium for the duration of the inj.

Afterwards, the plasma conc continuously decreases because of diffusion into the body and clearance through the kidneys to the urine.

The dynamic acquisition pattern is similar to that of DSC-MRI.

Images are acquired before, during and after injection of contrast agent.

Most benign and malignant lesions show signal enhancement in first few min after bolus administration.

The normal tissue may also show enhancement.

ASL

ASL gives absolute values of perfusion of tissue by blood.

This technique utilizes arterial water as an endogenous diffusible tracer which is usually achieved by magnetically labeling incoming blood.

ASL is completely non-invasive, using no injection of contrast agent or ionizing radiation and is repeatable for studying normal or abnormal physiology and its variation with time.

ASL required the subtraction of two images, one in which the incoming blood has been magnetically labeled (k/a label or tag image) and the other in which no labeling has occurred (k/a control or reference image).

The signal diff is the ASL signal and which removes the static tissue signal.

ASL signal is approx <1% which makes SNR.

Types of labeling methods

Continuous ASL (PASL)

Pulsed ASL (PASL)

Velocity Selective ASL (VSASL)

Vascular Territory Imaging (VTI)

CASL

In CASL, the magnetization of the arterial blood flowing through a major artery is continously labeled (usually by inversion) using RF pulses.

It is based on steady state approach.

A continous RF field is applied for a few sec along with a field gradient.

PASL

PASL involves a relatively short RF pulse, which results in labeling (usually inversion) of the blood in a large region adj to the imaging volume.

1. Signal alternating with alternating radio frequency (STAR)

2. Flow alternating inversion recovery (FAIR)

3. Proximal inversion with control for off-resonance effects (PICORE)

4. Transfer insensitive labeling technique (TILT)

VSASL

VSASL does not label the blood based on its spatial location (as in CASL and PASL) but rather does so based on its velocity, e.g. using binomial pulses to selectively  label blood flowing below a certain velocity.

This labeling is not spatially selective, but can be made sensitive to different blood velocities.

ASL Parameter

ASL contrast is related to the preparation module (blood labeling), but does not rely on T2/T2*/T1 contrast of the acquisition module.

Short TE to maximize SNR

Long TR to allow the labeling blood to reach an imaging plane.

Large FA is used.

Long scan duration due to NSA around 40.

ASL Signal Dynamics

In routine PASL and CASL, the signal at a single TI or post labeling delay of 1.5-1.8 sec is often used.

Two labeling times (Quantitative Imaging of Perfusion using a Single Subtraction- QUIPSS I/II) are used to minimize the definition error of bolus width.

The ASL signal curve also shows 3 phases.

Summary

DSC-MRI is good for quick measurement of transit time, whole brain coverage and fast scan time.

DCE-MRI is good for measurement of BV, perfusion permeability and capillary permeability and for reducing image artifacts.

ASL is good for BF measurement, repeatable studies and provide unique opportunity to provide CBF information without inj of contrast agent.

However, bolus methods with inj of contrast agent provide better sensitivity with higher spatial resolution.

References

Perfusion Magnetic Resonance Imaging: A Comprehensive Update on Principles and Techniques, Korean Journal of Radiology, July 2014.

Perfusion MRI: Technical Aspects, 2009 Lecture.

Perfusion Imaging by Monil Shah.

Dynamic Contrast Enhanced MRI by Sungheon Gene Kim, PhD, NYUSOM/Radiology/CBI.

MR Perfusion Imaging Techniques and Applications, Chen Lin, PhD, Indiana University School of Medicine & Clarian Health Partners.

ASL Basics I, University of Navarra, Pamplona, Spain.

Alternatives to BOLD for fMRI, Harvard-MIT Division of Health Sciences and Technology.

Perfusion MRI: The Five Most Frequently Asked Technical Questions, AJR Am J Roentgenol, 2013.

Contrast-Enhanced MR Perfusion Imaging, University of Wisconsin, Madison, RSNA Chicago, 2008.

Advanced MRI Techniques (and Applications), University of California, Los Angeles, January 2012.

Noncontrast Perfusion MR Imaging with Arterial Spin Labeling (ASL) of cerebrovascular disease, European Society of Radiology, 2011.

Magnetic Resonance Imaging of Perfusion, D. Le  Bihan, 1990.

Skull Radiography- Lateral and PA views

SKULL-LAT VIEW

Patient position

-         Patient lies supine to the x ray table.

-         Median sagittal plane is parallel to the table.

-         The interorbital line is perpendicular to the cassette.

Central ray:

Horizontal ray Centre midway between the glabella and the external occipital

protuberance to a point approximately 5 cm superior to the external auditory meatus.

Essential image characteristics

-         The image should contain all of the cranial bones and the first cervical vertebra.

-         Both the inner and outer skull tables should be included.

-          A true lateral will result in perfect superimposition of the lateral portions of the floors of the anterior cranial fossa and those of the posterior cranial fossa.

-         The clinoid processes of the sella turcica should also be superimposed.

1. Coronal Suture

2. Orbital Plates of the Frontal Bone

3. Posterior Clinoid Process

4. Auricle or Pinna (Ear)

5. Lambdoidal Suture

6. External Occipital Protuberance

7. Posterior Arch of C1

8. Sphenoid Sinus

Note: Skull lateral can also be done while the patient is in erect position using vertical ray in the uncooperative patient.

Skull PA (Occipito frontal)

Patient position:

-         The patient sits facing the erect Bucky.

-         Median sagittal plane is perpendicular to the cassette.

-         Neck flex so that the orbito-meatal base line is perpendicular to the Bucky.

Direction of central ray

Centre in the midline of the occiput so that the exit ray emerges in the midline at the level of the glabella.

Essential image characteristics

-          All the cranial bones should be included within the image, including the skin margins.

-         It is important to ensure that the skull is not rotated which can be assessed by measuring the distance from a point in the midline of the skull to the lateral margin. If this is the same on both sides of the skull, then it is not rotated.

1. Lambdoidal Suture

2. Ethmoid Sinus

3. Petrous Ridge

4. Maxillary Sinus

5. Body of the Mandible

6. Condyle of the Mandible

7. Mastoid Air Cells

With the degree of beam angulation, the position of the petrous ridges within the orbit can be evaluated:

– Occipito-frontal: the petrous ridges should be completely superimposed within the orbit, with their upper borders coincident with the upper third of the orbit.

– OF10° caudally: the petrous ridges appear in the middle third of the orbit.

– OF15° caudally: the petrous ridges appear in the lower third of the orbit.

– OF20° caudally: the petrous ridges appear just below the inferior orbital margin.

Note: Never use skull PA when there is a possibility that the facial bones may be fractured or when the patient is unconscious. When the patient cannot be X-rayed in a prone position, then Skull AP- Supine (frontal-occipital) is used.

Reference:

1. Clark's Positioning Radiography, 12th Edition

Acoustic Noise in MRI

Acoustic Noise in MRI

-Parajuli K, MSc Medical Imaging, IOM, Nepal

Why noise is a problem in MRI¹

·       It causes discomfort, anxiety and distraction to the patient.

·       It causes problem in verbal communication.

·       It can cause discomfort in sedated patients.

·       Some medications can increase the sensitivity of our ear to sound.

·       Temporary shift in the threshold of hearing has been reported.

·       Neonates have increased sensitivity to acoustic noise.

·       Unwanted activation of auditory cortex during functional MRI causes spurious patterns and is a serious problem.²

How is noise produced in MRI

There are many sources of noise in mri.

They can be categorized into four groups

a. Due to gradient current

b. Due to eddy current

c. Due to radio frequency waves and slice selection pulse

d. Due to ambient noise

Noise due to gradient current³

Gradient, spatial variation in magnetic field is produced by the electric currents flowing through coil. Large pulsed electrical current ranging from 200 A to 600A with rise time in the order of sub millisecond or millisecond is applied to copper windings held together by epoxy resin and fiber glass. Gradient coil is itself situated in the magnet bore assembly and thus experience the main magnetic field, a force called Lorentz force develops in between the gradient current and static magnetic field, which causes bending and buckling the coil producing compressional force. Force of 2000 N per meter of gradient coil is produced resulting in vibrations with acceleration levels in the order of 100 m/s². Lorentz force is given by the relation:

where dl is the element inside the magnetic field.

Fig 1: grey arrow showing the direction of gradient current and white arrow shows direction of Lorentz force.

Figure (2) shows two mode of vibration: cone shaped vibration in Z gradient (A) and banana shaped vibration in X-Y gradient (B).

Studies have shown similarity between spectrum of gradient current and that of noise. Also, amplitude of vibration is higher when frequency of gradient current matches the resonant frequency of gradient coil which depends on properties of length and elastic properties of material in coil assembly. It is thought to be due to better acoustic transfer function of the gradient system at resonant frequency. The time course of the gradient current depends on various pulse sequence parameters (TR, TE, FOV). Thus pulse sequence can be modified such that frequency of gradient current (also the frequency of vibration) does not match the resonant frequency of the coil. In EPI the frequency of readout gradient can be adjusted such that it is not same as resonant frequency of the gradient coil.

In addition, the wave for gradient can be changed from trapezoidal to sinusoidal.

Noise due to eddy current⁴

Gradient magnetic field induces eddy current, according to faradays law of electromagnetic induction, in the metallic structures located in the vicinity of the gradient coil for eg. inner cryostat in superconducting systems and Rf body coil, Rf shield etc. These structures conducting the eddy current are situated in a magnetic field because of which Lorenz force develops and induces vibration. Recent study by Eldestein et al ⁴have shown that main source of noise in MRI is due to eddy current induced vibration of the conducting structures located near the gradient coil.

RF Hearing¹

It is due to pressure waves generated because of expansion of tissues as a result of rise in temperature (1x10^-6 degrees C) caused by the RF induced heating. It is very nominal compared to gradient field induced acoustic noise and there is no proof regarding harmful effects due to RF hearing.

Ambient noise³

Inside the scanner room there are various other sources of noise for example:

1.    Noise due to air blower of air handling system that can produce noise upto 20 dbA.

2.    Pounding sound of low frequency due to liquid helium pump that can reach up to 80 dB.

How much is the noise produced in MRI^{5, 6}

Noise level in MRI depends upon the hardware construction and surrounding environment i.e the presence and absence of various structures that reflect sound. Thus sound level is measured in specific pressure level or sound pressure level which can be expressed in linear or A-weighted scale. The sensitivity of human ear depends upon the frequency of sound. A-weighted sound takes into account the different sensitivity of human ear to various frequencies of sound and is expressed as dBA or dB(A).dB(A) is better correlated with the risk of noise induced hearing loss.

Noise levels of 70 to 130 dB have been reported. Price et al found noise level to be 82.5+-0.1 dB (A) in 0.23 to 0.5T systems and 118.4 +-1.3 dB (A) in 3T systems. Hattori et al⁶ found that noise in 3T MR systems exceeded that of regulatory limit of 99 dB given by IEC. The peak sound pressure level was 125.7 dB for MR angiography, 130.7dB for SS-EPI (single shot spin echo EPI) on the linear scale.

What factor affects level of noise produced?

It depends upon the scan parameters like section thickness, FOV, TR and TE. It is because change in these parameters bring about change in gradient output (rise time and amplitude).Smaller the section thickness and FOV and shorter the TR, higher is the level of noise produced. It increases logarithmically with increase in magnetic field.

Studies have shown presence and absence of patient can also create difference in level of acoustic noise produced. The value was found to be larger in the former case.

What are the regulatory limits?

According to US FDA guidelines peak un-weighted sound pressure level should not be greater than 140 dB. Also, A-weighted root mean square sound pressure level should be greater than 99dBA with hearing protection in place.

UK has provided guidelines for operators as well, according to which hearing protection should be provided to the operators who can be exposed at the average of 85 dB over 8 hour a day.

Exposure limits given by Occupational Safety and Health Administration (OSHA)¹

85dB(A)         16hrs a day

90dB               8 hrs a day

95db               4 hrs a day

100dB             2 hrs a day

105dB             1 hrs a day

110dB             0.5 hrs a day

115dB             0.25 hrs a day

What are the noise reduction strategies?

1. Modification in design of hardware

a)  Use of active-passive shielded, vibration isolated and vacuum enclosed gradient coils¹

Gradient assembly can be shielded so that eddy current is not induced in the surrounding metallic structures in the MRI assembly. Active shielding refers to the use of secondary coil other than primary gradient coil(X, Y, Z) that oppose and cancel out the fringe field. Like primary coils there are three shielding coils. Passive shielding refers to the use of conducting layers that are attached on the outer radius of the gradient coil.

Gradient coil are arranged such that structural or mechanical propagation of vibration into other component is reduced this is called vibration isolated gradient coil. Also they can be enclosed in vacuum to prevent air conduction of the vibration.

b)    Use of quiet gradient coil

Current in the gradient coil are passed such that net Lorentz force is reduced.

2. Passive noise control

Ear plugs and headphones can be used to lessen the intensity by 10 to 30 dB. Disadvantage of this method is that it interferes the verbal communication with patients.

3. Active noise control⁸

It utilizes the sound, also called anti phase sound that interferes destructively with the sound perceived by the subject undergoing MRI. There are two main topologies for active noise control. Feed forward and feed-back.

Feed-back topology:

·       A microphone is placed near to the area where attenuation is required.

·       It receives the sound generated by scanner which is perceived by the subject and sends reference signal to the loud speaker.

·       Loud speaker, on the basis of reference signal sends control sound that interferes with the noise and produces cancellation effect.

·       Disadvantage of this method is that because of the time delay between production of control sound and reception of the same by the subject, high frequency noise cannot be adequately suppressed, making this sort of topology not suitable for EPI sequences where noise of frequency 3 -4 kHz are generated.

Feed forward topology

·       Microphone is located near to the sound source which sends the reference signal in advance to the loudspeaker.

·       Another microphone called “error” microphone is located near the subject and provides electrical copy of residual noise.

·       Advantage of this topology is that higher frequency noise can be attenuated because the reference signal is available on advance and can compensate for the delay due to time required for the propagation of sound from loudspeaker to error microphone. Thus this sort of topology can be used to reduce high frequency noise from EPI sequences.

4. Use of quiet pulse sequences⁹

Pulse sequence with less gradient pulsing can be used. For example Gradient pulsing can be reduced by using single shot sequences based on stimulated echo (STEAM).Further reduction in gradient pulsing can be done by modifying the projection reconstruction method in which the two gradient pulses are replaced by a single mechanically rotating DC gradient coil.

Use of soft gradient pulse is another option. Hennel et.al¹⁰ used band limited pulses that used sinusoidal ramp with least number of slope (ramp) and maximum duration of slope (ramp duration). Noise level was brought to as low as 30dBA in GRE and SE and 60dBA in RARE sequence. Authors named such pulses as soft pulse which can be obtained by convolving the hard pulses i.e the pulses with rectangular waveform with the cosine window of length ᴦ (Figure 3)

Thus there is a growing concern regarding the detrimental effects of acoustic noise generated in the MR systems and the ongoing advancements in MRI hardware, especially in  the gradient coil technology, are dedicated towards development of a quiet scanner.

References

1.      MC.Jury M,Shellock F.G Auditory noise associated with MR procedure:A Review J Magn Reson Imaging.2000 12:37-45

2.      Measuring MRI noise

3.      Adriaan Moelker and Peter M.T. Pattynama Acoustic Noise Concerns in Functional Magnetic Resonance ImagingHuman Brain Mapping .2003;20:123–141

4.      Edelstein WA, Hedeen RA, Mallozzi RP, El-Hamamsy SA, Ackermann RA, Havens TJ (2002): Making MRI quieter. Magn Reson Imaging 20:155–163.

5.      Roozen N.B ,Koevoets A.H and den Hamer A.J Active Vibration Control of Gradient Coils to Reduce Acoustic Noise of MRI Systems.

6.      Hattori Y,Fukatsu H,Ishikagi T :Measurement and evaluation  of the acoustic noise of  a 3 Tesla MR scanner Nagoya J.Med.Sci,2009;69.23-28

7.      Tomasi DG¹, Ernst T Echo planar imaging at 4 Tesla with minimum acoustic noise. J Magn Reson Imaging. 2003;18(1):128-30.

8.      Chamber J,Bullock D,Kahana Y.Developments in active noise control sound systems for magnetic resonance imaging.Applied Acoustics 2007;68.281-295

9.      Alibek S,Vogel M,Sun W etal.Acoustic noise reduction in MRI using silent scan:an initial experience.Diagn Interv radiol 2014;20:360-363

10.  F. Hennel,* F. Girard, and T. Loenneker ‘‘Silent’’ MRI With Soft Gradient Pulses Magnetic Resonance in Medicine ,1999;42:6–10