Registry Part 5and6

Physical/Logical Gradients and Spatial Localization

Part VAmanda Golsch BSc, RT(R)(MR)

Physical Gradients• The physical coordinate system refers to the “physical body”.

• The gradient coil that varies the magnetic field from head to foot is the Z gradient.

• The gradient coil that varies the magnetic field from left to right is the X gradient.

• The gradient coil that varies the magnetic field from anterior to posterior is the Y gradient.

Slice Selection and Physical Gradients

• If the Z gradient is utilized for slice selection, this results in an axial slice.

• Think of a loaf of bread. The blue arrow shows us the Z gradient coil. If we use the Z gradient coil for slice selection we get an axial slice.

Axial slice

Physical Gradients and Slice Selection

• If the Y gradient coil is used for slice selection, the result will be a coronal slice.

• Back to the bread!

• The blue arrow shows us the anterior to posterior direction of the Y gradient. Therefore, when you cut the hoagie roll along the Y gradient the result is a coronal slice.

Note that the hoagie roll is sliced ina coronal orientation.

Physical Gradients and Slice Selection

• If the X gradient coil is utilized in slice selection, the result will be a sagittal slice.

• Let’s look at a c-spine

Logical Gradients• The logical notation is used when referring to the gradients by their function rather than their

direction/orientation.

• When speaking of logical gradients, the slice selection gradient is the Z gradient. This has This has nothing to do with physical direction or orientation!nothing to do with physical direction or orientation!

• The slice selection or Z gradient is the first gradient to be applied during the RF pulse. During a spin echo it is applied during the 90 and 180 pulses.

• The slice thickness is determined by the amplitude (slope) of the gradient and the transmit bandwidth of the RF pulse.

• Thinner slices = Higher amplitude and/or narrow bandwidth

• Thicker slices = Lower amplitude and/or wide bandwidth

• Slice location is determined by the transmit frequency of the RF pulse.

Logical Gradients• Frequency encoding (X gradient) occurs during the sampling of the echo. The sampling of the

echo occurs during the readout at the TE time.

• The sampled data is then placed into K-space.

• Let us assume that we wish to obtain an image with a 512 X 512 spatial resolution:

– The system will sample the echo 512 times in the presence of the frequency encoding gradient. This will generate 512 frequency data points or columns to be plotted into K-space in the frequency (X) direction. This creates enough data to reconstruct an image with a spatial resolution of 512 X 1.

– In order to obtain 512 X 512, we need to collect 512 echos one for each line in k-space (phase direction). Each of these 512 echos must be sampled 512 times for each point along each line (frequency and direction).

Logical Gradients• The phase encoding gradient is the Y gradient.

• The phase encoding gradient is applied during the FID and its purpose is to encode spatial information into the MRI signal. It provides the “line” in k-space on which the data will be plotted during the readout (frequency) period.

• It occurs after slice selection, but it is before the frequency encoding gradient.

• If you would like a phase resolution of 256, you would need to repeat the pulse sequence 256 times. This would occur at different amplitudes and polarities (128 positive and 128 negative steps).

• Spatial resolution is increased by acquiring more phase encoding steps. However, scan time is also increased.

K-Space• Data is collected by sampling echos.

• The data that is collected is digitized and mapped in relation to the spatial encoding gradients which are the phase (Y) and frequency (X) gradients.

• The map is known as K-Space.

• In one direction of K-Space is frequency information, and the other direction holds phase information.

• Raw data is the signal information that is stored in K-Space.

• The letter K is used when referring to frequency.

• The number of data points in k-space is determined by the number phase and frequency encodings selected by the operator.

• The more data points selected, the better the detail. This results in longer scan times.

• Once all of the data is collected, the raw data is reconstructed by the Fourier Transform in the array processor.

• RAW DATA IS NOT THE MR IMAGE; IT IS DATA SIGNAL INFORMATION.

Image Quality

Part VI

Amanda Golsch BSc, RT(R)(MR)

Image Quality and Slice Thickness

• Slice Thickness is determined by the amplitude or slope of the z gradient.

• Higher amplitudes yield thinner slices.

• Lower amplitudes yield thicker slices.

• As slices get thinner, SNR is reduced but spatial resolution improves.

Image Quality and Spatial Resolution

• Spatial resolution allows you to distinguish one structure as separate from another.

• The only way for spatial resolution to be increased is to reduce voxel volume or pixel size.

• A voxel is a 3-dimensional object. Its size is determined by FOV, the acquisition matrix, and slice thickness.

• The pixel size is synonymous with the term in-plane resolution. The term pixel is often applied to the face of the voxel and is determined by the acquisition plane and FOV. Note that slice thickness does not play a role in determining the size of the pixel.

• Magnification can make a picture look blurry because it magnifies the pixel.

SNR• Signal to noise is a ratio of MR signal to noise.

• Noise can occur from multiple sources. The patient, system electronics, and the environment are all sources of noise.

• Signal comes directly from the tissues within the voxel.

• In order to increase SNR, you need to reduce the noise and/or increase the signal.

• Operator acquisition parameters greatly effect the SNR.

Contrast• TR,TE,TI, and flip angle are all parameters that can be used to manipulate MR contrast.

• Increasing the TR will increase the amount of longitudinal magnetization allowed to recover between excitation periods. This results in an increase in the SNR.

• Therefore, decreasing the TR will decrease the SNR.

• Increasing the TE increases the amount of transverse magnetization decay between the excitation pulse and the sampling of the echo and, therefore, will result in a reduction in the SNR.

• If you decide to null the signal from fat by adjusting the TI, and the area being imaged is primarily composed of fat, the image would have a low SNR because the tissue that is providing the signal has been nulled.

• Adjusting the flip angle can increase or decrease SNR. With a low flip angle, the amount of transverse magnetization is small, so the SNR will to low.

Field Homogeneity and Shimming• The magnetic field must be as homogenious as possible. This is especially true at isocenter.

Homogeneity is maximized by shimming.

• Shimming can be accomplished actively or passively.

• Active shimming implies the use of addition coils within the magnet. Current applied within the shim coils either adds or subtracts from the static magnetic field to produce as homogeneous a field as possible.

• Passive shimming implies the use of small bits of ferrous material, known as shim plates. These plates are placed around the bore.

• Homogeneity is expressed in ppm.

RF Shielding• RF from outside sources can interfere with MR studies.

• Outside RF sources can degrade image quality, so MR scan rooms are RF shielded.

• MR rooms use a lining of copper in the walls and around the doors to shield from outside RF sources.

• RF shielding is known as a “Faraday Cage”

• Holes or tears in the shielding can result in extraneous signals entering the scan room and can appear as zipper artifacts on images. This is known as a leak in the RF shielding and can occur when scanning with the room door open or not tightly closed.

Question 1• The physical gradient coil that varies the magnetic field from left to right is the

– A. Y gradient

– B. Z gradient

– C. X gradient

– D. W gradient

Question 2• To obtain a coronal slice, which physical gradient coil should be applied

– A. Y gradient

– B. F gradient

– C. D gradient

– D. X gradient

Question 3• Which logical gradient is applied during the RF pulse?

– A. Y gradient

– B. X gradient

– C. Z gradient

– D. X and Y gradient

Question 4• Which logical gradient is the frequency encoding gradient?

– A. Z gradient

– B. X gradient

– C. Y gradient

– D. A gradient

Question 5• How many echos need to be sampled to obtain a spatial resolution of 256 X 1?

– A. 128

– B. 256

– C. 512

– D. 600

Question 6• The gradient amplitude must be for a 3mm slice than a 5mm slice.

– A. Higher

– B. Lower

– C. Equal

– D. Parallel

Question 7• K-space stores?

– A. The MR image

– B. The FID

– C. SNR

– D. Raw data

Question 8• What mathematical process converts raw data into a reconstructed image?

– A. K-Space

– B. Faraday’s Law

– C. Fourier transform

– D. The array processor

Question 9• If you increase your spatial resolution from 256 to 512 what happens to the SNR?

– A. It stays the same

– B. It decreases

– C. It increases

– D. It is the average of 256 and 512

Question 10• Decreasing the TE does what to the SNR?

– A. Increases

– B. Decreases

– C. It doesn’t effect SNR

– D. Averages