Selecting Imaging Parameters for RARE Pulse Sequences

Help! I put a cool sample into my ilumr MRI system and hit run on one of the dashboards, but the image isn't looking as amazing as I had hoped. Where do I go from here? How do I improve my MR image?

Understanding the imaging pulse sequence you are using and how the parameters you have selected will affect image quality and contrast, is key to creating an awesome MR image. Most of ilumr's imaging dashboards use the RARE pulse sequence so that is the topic for today's blog post! This short guide will introduce the RARE pulse sequence, discuss its key parameters, and provide examples of how to select parameters based on a sample's relaxation properties.

What is RARE?

RARE stands for “Rapid Acquisition with Relaxation Enhancement” and is a modification of the conventional SE (Spin Echo) pulse sequence. Instead of a 90-degree radio frequency pulse followed by a single 180-degree pulse, RARE uses multiple 180-degree pulses and a series of phase encoding gradients. On some commercial scanners, RARE imaging is also known as FSE (Fast Spin Echo) imaging or TSE (Turbo Spin Echo) imaging. ilumr uses RARE pulse sequences as the base for most of its imaging protocols as this type of pulse sequence significantly reduces scanning times and improves both SNR (signal to noise ratio) and spatial resolution. 

Key RARE Imaging Parameters

Fig.1. Simple version of the RARE pulse sequence. 

The first step in parameter selection is to make sure you are choosing parameter combinations that do not violate timing constraints. Therefore, it is helpful to understand what the RARE pulse sequence looks like and how the pulse sequence parameters relate to each other.

Echo Train Length (ETL): ETL is the number of echoes produced. It is also known as “turbo factor” or “shot factor” in some commercial scanners. Generally, the acquisition time is inversely proportional to the ETL, therefore, increasing the ETL will reduce the time the experiment takes. However, increasing the ETL too far is associated with a decrease in SNR. This parameter is only editable in the advanced 2D and 3D RARE dashboards. In the standard dashboards, this parameter is calculated automatically. In the advanced dashboards, the ETL must not exceed the value selected for Phase Encoding Resolution. If the ETL is set higher, the dashboards will automatically set it to match the Phase Encoding Resolution. 

Effective Echo Time (TE(eff.)): TE(eff.) is the echo time acquired at the lowest order phase encode step and is the echo time that has the most effect on the overall contrast of the image. This parameter is only editable in the advanced 2D and 3D RARE dashboards. In the standard dashboards this parameter is calculated automatically. This value must be long enough to accommodate the selected ETL and number of phase encoding steps. This value must always be shorter than TR. If the TE(eff.) or TR values selected violate pulse sequence timing constraints or exceed the hardware’s timing limits, the RARE dashboards will automatically select the closest allowable values and set the parameters accordingly. If the dashboard is limiting TE(eff.) and you want it to be shorter, you can reduce the ETL to allow a shorter TE(eff.). 

Repetition Time (TR): TR is the time between excitation pulses. It is a key parameter for selecting image contrast. This parameter can be edited in both the standard and advanced 2D and 3D RARE dashboards. TR must be long enough to accommodate the selected TE(eff.) and ETL. If the TE(eff.) or TR values selected violate pulse sequence timing constraints or exceed the hardware’s timing limits, the RARE dashboards will automatically select the closest allowable values and set the parameters accordingly. To achieve very short TR values, ETL and/or TE(eff.) may need to be reduced.

Echo Spacing (ESP): In the dashboards, this value is calculated automatically based on the other parameters you have selected and can not be modified directly. In the standard RARE dashboards, the ESP is kept as short as possible to improve SNR. In the advanced RARE dashboards it is determined from TE(eff.) and ETL. 

Parameter Selection to Enhance Signal Acquisition

You now understand the timing constraints associated with the RARE pulse sequence, but how do you make sure you are selecting parameters that will produce a bright and clear image?

Signal amplitude is determined primarily by the net magnetization in the transverse plane during signal acquisition. If you want to maximise signal amplitude, you need to do two things:

1. Select a TR value that allows the longitudinal component of the magnetization to fully recover before it is tipped into the transverse plane during the 90-degree excitation pulse.

2. Select a TE(eff.) value so that the transverse component of the signal is acquired before it decays too far.

   
   

Fig.2. T1 and T2 Relaxation Plots

Since the time constants that describe the relaxation of these components vary from sample to sample, it is useful to measure the T1 and T2 values of your sample using ilumr's T1-T2 IRCPMG or SRCPMG relaxometry notebooks. When imaging, you can enhance the amount of signal you are acquiring, and therefore the image quality, by selecting a TR value that is significantly longer than your acquired T1 value and a TE(eff.) value that is significantly shorter than the T2 value. Follow along with this example to observe this process for yourself!

Enhanced Signal Example

For this example, we will use the "Shim Sample" from your ilumr kit which is a vial of 5mM Copper Sulphate solution.

Acquiring T1 and T2 Values:

• Navigate to home/notebooks/example notebooks/Relaxometry and open the T1-T2 IRCPMG Relaxometry notebook. If your ilumr does not have the latest T1-T2 IRCPMG and SRCPMG notebooks you can download them from our GitHub.

• In the top, left panel select Run > Run All Cells.

• Once the experiment interface loads, press the green RUN button.

• Hover your curser over the center of the bright spot in the middle of the T1-T2 map. The T1 and T2 values will appear in a pop up box. For my sample, I measured T1 = 192ms and T2 = 161ms. We can now use these values to select our TR and TE(eff.) parameters.

Fig.3. T1-T2 IRCPMG Notebook (Shim Sample)

Poor Signal: 

First, an example of poor parameter selection! The image below shows an experiment that uses TR = 300 ms and TE(eff.) = 121 ms for our sample where T1 = 192ms and T2 = 161 ms. The TR of 300 ms does not allow the longitudinal component of the magnetization to recover fully, reducing the amount of magnetization that can be tipped into the transverse plane during excitation. The TE(eff.) of 121 ms allows the transverse magnetization to decay too far before acquisition, resulting in further loss of signal. Not ideal!

Fig.4. Poor quality image of shim sample.

Good Signal:

The image below shows better parameter selection for imaging the 5mM Copper Sulphate solution. The TR parameter is over four times larger than the measured T1 time constant, allowing the longitudinal component of the magnetization to recover significantly before it is tipped into the transverse plane at the next 90-degree pulse. The TE(eff.) parameter has been made as low as the dashboard will allow, based on the other parameters that have been selected. This minimises the signal loss caused by T2 relaxation. Much better results!

Fig.5. Good quality image of shim sample.

Parameter Selection for Basic Image Contrast

You have now succeeded in creating a nice bright and clear image, but how do you create contrast between different materials or different types of tissue in a sample?

In the advanced 2D and 3D RARE dashboards, three types of image contrast can be created by varying the TR and TE(eff.) parameters: Proton Weighted, T1-Weighting, and T2-Weighting. For each type of image weighting, the parameters must be selected relative to the sample's relaxation properties (as described by the time constants T1 and T2).

The table below provides a guide for how TR and TE(eff.) should be selected relative to T1 and T2.

Weighting	TR relative to T1	TE (eff.) relative to T2
Proton Density Weighted	Long	Short
T1 Weighted	Short	Short
T2 Weighted	Long	Long

To learn more about image weighting and why varying pulse sequence parameters affects image contrast, check out our MRI Fundamentals Contrast Imaging Lab.

The following example illustrates how this table can be used to create contrast images using a green bean sample.

Contrast Example

For this example we will use a green bean. Green beans make great imaging samples as the different parts of the bean have different relaxation properties. If green beans are not available you could use a thin red chilly or a sample cut from a kiwifruit or cherry tomato.

Acquiring T1 and T2 Values:

• Once again, it is important to assess the T1 and T2 properties of your sample before selecting imaging parameters.

• This time when you run the IRCPMG notebook, you will see more than one bright patch appear on the T1-T2 map. This is because there are multiple components in the bean sample.

• Hover over the brightest patch and record the T1 and T2 values. I measured T1 = 1.8 s and T2 = 1.1 s. Now we can select parameters to create three different types of contrast.

Fig.6. T1-T2 IRCPMG Notebook (Bean)

Fig.7. T1-T2 IRCPMG Map (Bean)

Proton-Weighted Image:

Scans = 16, ETL = 16, TE(Eff.) = 15ms, TR = 4s, Dummy Echo = Unchecked

A long TR value (relative to T1 = ~2s) allows all components of the bean to recover their longitudinal magnetization before the excitation pulse. This means that initially all components in the bean have the same amount of signal that they can contribute to the image. A short TE(eff.) value (relative to T2 = ~ 1s) results in the signal being acquired before too much transverse relaxation can take place. This results in an image where the brightness correlates to the amount of local hydrogen spins that can contribute to the magnetization.

Fig.8. Proton-Weighted Image of a Green Bean

T1-Weighted Image:

Scans = 32, ETL = 16, TE(Eff.) = 15ms, TR = 0.5s, Dummy Echo = Checked

A short TR value does NOT allow all components of the bean to recover their longitudinal magnetization before the excitation pulse. This means that initially each component of the bean will have a different amount of signal that they can contribute to the image. A short TE(eff.) value means that all the signals are acquired before transverse relaxation has a strong effect on the image results, therefore the contrast that is based on different states of longitudinal recovery will be preserved. The signal from components with longer T1 values are attenuated while components with short T1 values (which were able to recover more before excitation) are highlighted.

Fig.9. T1-Weighted Image of a Green Bean

T2-Weighted Image:

Scans = 16, ETL = 64, TE(Eff.) = 202ms, TR = 4s, Dummy Echo = Unchecked

A long TR value (relative to T1 = ~2s) allows all components of the bean to recover their longitudinal magnetization before the excitation pulse. This means that initially all components in the bean have the same amount of signal that they can contribute to the image. However, a long TE(eff.) value (relative to T2 = ~ 1s) allows the signal from each component to decay before acquisition. Therefore, components with short T2 values (which experience signal decay much quicker) are attenuated while components with long T2 values are highlighted.

Fig.10. T2-Weighted Image of a Green Bean

Now you understand the basic RARE pulse sequence parameters, have learnt how to measure the relaxation properties of a sample, and know how to use this data to pick imaging parameters that enhance signal acquisition and create three types of image contrast. It's time to collect as many samples as you can find and start making some awesome MR images!

Join the Discussion!

Our forum is a great place to discuss topics with the Resonint team and other ilumr users. If you have further questions about pulse sequence parameter selection or want to share the awesome MR images you created, send us a message in the "General" topic!

If you have questions about our product range, are interested in receiving a quote, or would like to schedule a product demonstration, send us an email at info@resonint.com.