MR Basics: Module 6 Pulse Sequences

Transcription

1 Module 6 Transcript For educational and institutional use. This transcript is licensed for noncommercial, educational inhouse or online educational course use only in educational and corporate institutions. Any broadcast, duplication, circulation, public viewing, conference viewing or Internet posting of this product is strictly prohibited. Purchase of the product constitutes an agreement to these terms. In return for the licensed use, the Licensee hereby releases, and waives any and all claims and/or liabilities that may arise against ASRT as a result of the product and its licensing.

2 MR Basics: Module 6 Pulse Sequences 1. MR Basics Pulse Sequences Welcome to Module 6 of MR Basics Pulse Sequences. This module was written by J. Randall Carr, R.T.(R)(T)(CT)(MR), CRA, and Lisa Deans, R.T.(R)(MR). 2. License Agreement and Disclaimer 3. Objectives After completing this module, you will be able to: List parameters related to tissue characteristics that affect image quality, such as spin density and T1 and T2 relaxation. Identify other parameters that affect image quality, such as repetition time, echo time, inversion time and flip angle. Apply pulse sequence principles to magnetic resonance (MR) imaging. Explain the concepts of image formation in MR imaging. Describe image contrast appearance according to image weighting. 4. Image Balancing Act The MR technologist, under the supervision of the radiologist, creates optimum image quality by balancing image resolution, signal and scan time. Focusing on any single component jeopardizes the other components. For example, if the technologist s goal is to obtain maximum image resolution, he or she can adjust the overall signal of the image but may create an exam with a very long scan time. If the goal is to reduce scan time, the technologist must be careful to avoid reducing image resolution. Finally, if the goal is the maximize image signal quality, doing so can compromise overall image resolution. 5. Hydrogen Atoms Before addressing specific pulse sequences, it helps to review some basic MR imaging principles. Without an external magnetic field present, the hydrogen atoms in the body spin randomly at different speeds and in different directions. The nuclei produce no net magnetization. 6. Magnetic Field Effect Look at this animation. When the body is placed in an external magnetic field, referred to as B 0 or static field, the magnetic moments of the hydrogen protons align either in the same direction as the magnetic field or in the opposite direction to the magnetic field. The hydrogen protons with lower-energy states align in parallel to the main magnetic field. Hydrogen protons with a higher-energy state oppose the static field in an antiparallel direction. Increasing the magnetic field strength causes more hydrogen protons to align with the field in the parallel, low-energy position. 7. Net Magnetization Vector When pairs of low-energy and high-energy nuclei cancel each other out, the magnetic moments of the unpaired nuclei create a sum called the net magnetization vector. Only the unpaired nuclei of the net magnetization vector form the MR signal ASRT. All rights reserved. 1 MR Basics: Module 6

3 8. Precession in the Magnetic Field The hydrogen nuclei spin on their axes. Placing the nuclei in a magnetic field produces a second spin or wobble. Nuclei precess around the direction of the field. The direction the nuclei spin is the precessional path, and the speed at which they spin is the precessional frequency. 9. The Larmor Equation In MR, the Larmor equation is used to calculate the precession rate of the hydrogen nuclei. The precessional frequency (ω) is equal to the gyromagnetic ratio of hydrogen (γ) times the field strength (B 0 ). In this equation, the gyromagnetic ratio is a constant and the precession rate is directly proportional to the field strength. Precessional frequency is measured in megahertz (MHz). One MHz is 1 million cycles per second. 10. The Gyromagnetic Ratio The gyromagnetic ratio is the ratio of the magnetic moment (field strength) to the angular moment (frequency). The gyromagnetic ratio of hydrogen is MHz/tesla (T). 11. Calculating Precessional Frequency To calculate the precessional, or Larmor, frequency, multiply the gyromagnetic constant for hydrogen (42.58 MHz) by the field strength of the main magnet. These two calculations demonstrate how the precessional frequency increases as the strength of the magnet increases. 12. Knowledge Check Calculate what the precessional frequency would be for a 1.5T magnet. 13. Knowledge Check Calculate what the precessional frequency would be for a 3.0T magnet. 14. Longitudinal and Transverse Magnetization In this slide, the blue arrow represents the net magnetization vector aligned with the main magnetic field in the longitudinal plane. Transverse magnetization occurs when a radiofrequency (RF) pulse is applied causing the net magnetization vector to flip into the transverse plane. The RF energy used to flip the net magnetization vector must be transmitted at the same frequency as the hydrogen atom. In MR imaging the RF range is generally 1 to 100 MHz. 15. Faraday Law of Induction The Faraday law of induction states that an electrical current is induced by changing the magnetic field. This series of illustrations demonstrate that when the net magnetization vector is flipped, it rotates in the magnetic field producing a current in the receiver coils. The coil generates an analog signal, which is subsequently converted into a digital signal. The digital image is then able to be viewed on the computer screen. 16. T1 and T2 Relaxation Times Tissues vary in how they behave in a magnetic field. Depending on surrounding conditions, one tissue magnetizes to a greater degree than other tissues. Each tissue has a characteristic T1 recovery time that varies according to the MR scanner s magnetic field strength and an inherent T2 decay time. The tissues precessional differences set off a local field gradient that causes 2012 ASRT. All rights reserved. 2 MR Basics: Module 6

4 dephasing. The spins of the nuclei move out-of-phase with one another (phase incoherence) and transverse magnetization decreases over time. This table lists the approximate times that a particular tissue will take to relax in a 1.5T magnet. 17. Image Contrast The MR technologist must produce the desired image contrast as directed by the radiologist. MR uses three main techniques to create a specific image contrast or weight: T1, T2 and proton, or spin, density. During individual MR pulse sequences, specific chains of RF and gradient pulses excite the hydrogen protons while spatially encoding the image information. The basic MR pulse sequences of early MR imaging were spin-echo, inversion-recovery and gradient-echo. Later, variations of these basic sequences were developed, including fast spinecho, fast gradient-echo and echo planar imaging (EPI). Technologists use each pulse sequence to display specific anatomy contrast and pathology. T1-weighted scans provide excellent resolution and create a general map of the anatomy. T2- weighted scans demonstrate pathology. Proton-density-weighted scans minimize the effects of T1 and T2 weighting, resulting in an image that depends primarily on the density of protons in the imaging volume. Proton-density scans can better distinguish the difference between white and gray matter, which can be very useful if the patient has a disease such as multiple sclerosis. 18. Pulse Sequences Let s start discussing pulse sequences by describing the conventional spin-echo sequence, an early, basic MR pulse sequence. The signal from the spin-echo sequence consists of multiple frequencies, each representing a different position along the gradient body. The conventional spin-echo pulse sequence starts with a 90-degree RF pulse to flip the net magnetization vector from the longitudinal plane to the transverse plane. After the 90-degree pulse is turned off, the net magnetization vector starts to recover and realign with the main magnetic field in the longitudinal plane. The signal generated as the vector starts to recover is known as free induction decay. The free induction decay signal is not sampled. A 180-degree RF pulse then is applied to send the net magnetization vector back into the transverse plane. The signal formed when the 180-degree RF pulse is turned off is referred to as a spin-echo. The spin-echo is produced when the free induction decay, or partial echo, is spun by the 180-degree RF pulse. The full echo also can be viewed by making a mirror image of the free induction decay and flipping it over to create the full echo. The full echo is sampled and the analog data is converted into a digital image. 19. Repetition Time (TR) In a pulse sequence, the time period between the beginning of the sequence and the beginning of the succeeding, identical pulse sequence is known as the repetition time, or TR. TR is measured in milliseconds. Look at this illustration. TR is the amount of time between successive pulse sequences applied to the same slice. TR controls how much T1 relaxation is visible between tissues. A shorter TR, say 300 to 600 milliseconds, increases the T1 weighting of tissues. TR is the primary controlling factor for T1 contrast ASRT. All rights reserved. 3 MR Basics: Module 6

5 20. Echo Time (TE) Look at this illustration. Echo time (TE) is the time in milliseconds between the 90-degree pulse and the peak of the echo signal. TE is the primary controlling factor for T2 relaxation. Longer TE times, say 80 milliseconds and longer, increase a tissue s T2 weighting. 21. Knowledge Check 22. Knowledge Check 23. Knowledge Check 24. Knowledge Check 25. Scan Timing Parameter Chart This chart outlines scan parameters for contrast weighting, but remember that scan parameters may be equipment and facility specific. The parameters for a T2 sequence are a long TR of more than 2,000 milliseconds and a long TE of more than 60 milliseconds. For T1 weighting, parameters are a short TR (700 milliseconds or less) and a short TE. A proton-density sequence has a long TR (more than 2,000 milliseconds) and a short TE. 26. Pulse Sequence Diagram Each pulse sequence can be represented by a pulse sequence diagram. The first line represents the timing of the RF pulse. The second line (Gz) represents a gradient pulse used for slice selection. The third line (Gy) represents a gradient pulse used for phase encoding. The fourth line (Gx) represents a gradient pulse used for frequency encoding. 27. Knowledge Check 28. Knowledge Check 29. Slice Selection The slice selection gradient alters the magnetic field strength at a specific slice location. The field strength changes along the slope of the gradient. Altering the magnetic field strength at each slice location excites only the hydrogen atoms at that specific location. Let s look at some illustrations. This first illustration represents a sagittal image of a brain, with the diagonal line corresponding to a gradient slope. Let s use 1.5T in this example. This second illustration shows the field strength of the MR scanner is 1.5 T, so the 1.5-T field strength is in the middle of the gradient slope ASRT. All rights reserved. 4 MR Basics: Module 6

6 If we are acquiring 10 slices, altering the magnetic field strength at each slice location excites only the hydrogen atoms at that specific slice location and causes the net magnetization vector to precess at different frequencies, making it possible to differentiate individual slices. These frequencies are shown in the right-hand column of this third illustration. Compare the frequencies. Without slice selection, all of the hydrogen atoms would be excited, which would make it impossible to differentiate slice 1 at the base of the brain from slice 10 at its vertex. Because each slice has a different magnetic field strength, it also has a unique precessional frequency, enabling the scanner to obtain anatomical information from specific locations. 30. Gradients Gradients change the magnetic field along the slope by adding or subtracting magnetic field strength. The change in field strength either increases or decreases the precessional frequency, or speed, of the spinning protons. 31. Phase and Frequency Encoding Without phase and frequency encoding, none of the pixels in this 3 x 3 matrix could be distinguished. Each individual pixel looks exactly the same. Applying a phase-encoding gradient pulse helps to spatially locate each pixel. 32. Phase-encoding Gradient The phase-encoding gradient is applied after slice selection and excitation, but before the frequency-encoding gradient. It is applied at a right angle to the other two gradients. Spatial resolution is directly related to the number of phase-encoding steps. 33. Frequency-encoding Gradient The frequency-encoding gradient causes a range of frequencies to exist along the applied direction. After the MR signal is measured (a combination of all signals from a slice), the Fourier transform is used to convert the raw data in k-space to its spatial location on the matrix. 34. Matrix Size The scan matrix and scan field of view determine the pixel size, which in turn affects image resolution. The scan matrix is a grid consisting of rows and columns. The MR technologist can choose a variety of matrix sizes. The phase matrix size determines how many times the pulse sequence is repeated. For discussion purposes, we ll refer to the columns as the frequency direction and the rows as the phase direction. In MR, frequency matrix sizes are 512, 384 or 256. Phase matrix sizes can be 512, 384, 256, 224, 192, 160 or 128. The MR technologist may use any combination to form a variety of matrix sizes, such as 512 x 512 and 384 x K-space The raw data collected from the MR signal are mapped on to k-space. The signal intensity at each point in k-space contains the information about both the contrast and location of the data. The contrast data are placed in the middle section of k-space, and the resolution data are placed on the outer edges. A 256 phase matrix has 128 lines above and below the center of k-space ASRT. All rights reserved. 5 MR Basics: Module 6

7 36. K-space Look at this illustration. The vertical axis is the phase axis and the horizontal axis is the frequency axis. To acquire a 20-slice dual-echo brain scan with a 256 x 192 matrix, the MR technologist must have 40 lines, or 2 lines for each slice of k-space. K-space must be filled with 256 columns (frequency encoded) and 192 lines (phase encoded) of information. Each solid horizontal line represents a line of k-space. The horizontal dotted line would be the middle of k- space; therefore in this example, if this were a 256-matrix, there would be 128 lines above and 128 lines below this dotted line. 37. K-space There are many ways to fill k-space. Conventional spin-echo, fast spin-echo and gradient-echo pulse sequences fill k-space lines from the left to the right and stop. The sequences then return to the left each time to fill a line of k-space. This animation shows how k-space is filled for a conventional spin echo scan. All lines of k-space are filled for this pulse sequence. 38. K-space In a fast spin-echo sequence, scan time is reduced because not every line of k-space is filled. 39. K-space EPI scans fill from right to left. The scan time is very short because all the lines of k-space are filled during a single repetition. 40. Spiral Filling In spiral filling, k-space begins to fill at the center. The readout gradient or phase gradient rotates rapidly to fill lines from left to right and then right to left. 41. Centric Filling In centric filling, k-space is filled line by line starting at the center. The shallowest phaseencoding gradients are applied first and the steepest ones are applied at the end of the pulse sequence. This technique improves contrast and is used in fast gradient-echo sequences. After all k-space is filled from the center to the top, it begins to fill from the center to the bottom. 42. Zero Filling Zero filling is used to reduce scan time, but signal is also reduced. In this method, not all the lines of k-space are filled with data, but at least half the lines must be filled. 43. Knowledge Check 44. Knowledge Check 45. Number of Signals Averaged The number of times the image is sampled is called number of signals averaged, or NSA. This concept also is known as NEX, or number of excitations. Either term is correct, but for consistency, we ll use NSA in our discussion ASRT. All rights reserved. 6 MR Basics: Module 6

8 If the MR technologist chooses a 256 x 192 matrix with a 1 NSA, 192 lines of k-space are sampled one time. With a 2 NSA, the technologist samples 192 lines of k-space two times, for a total of 384 lines. The computer averages the first sample and second sample together to increase signal and decrease motion artifact. Doubling the NSA from 1 to 2 or from 2 to 4 provides an approximately 40 percent increase in overall signal-to-noise ratio. Doubling the NSA also doubles the imaging scan time for the sequence. 46. Gradient Slew Rate The slew rate is the rate at which the gradients reach their maximum amplitudes. Slew rate is calculated by dividing the gradient amplitude by the gradient rise time. It is measured in tesla (T) per meter per second. Rise time is how quickly the gradient reaches maximum amplitude. Amplitude is a measure of the change in the magnetic field over distance and is measured in millitesla per meter (mt/m). The higher the slew rate, the shorter the echo space. 47. Field of View The field of view covers the anatomy being imaged. Increasing the field of view decreases echo space and resolution. Look at these illustrations. The image on the left was acquired with a 24- cm field of view and the image on the right was acquired with 12-cm field of view. In the case of the image on the left, the field of view has been doubled in 2 dimensions (right to left and anterior to posterior). Doubling the field of view yields a 22-mm or four-fold increase in voxel size; this increases signal-to-noise ratio by a factor of four. The image on the right, acquired with a smaller field of view, has a grainy appearance compared with the 24-cm image on the left. This means that the image has a lower signal-to-noise ratio and a correspondingly lower contrast-to-noise ratio. Much like thickness or any parameter that influences voxel size, an increase in field of view results in less spatial resolution. This reduction in spatial resolution is caused by partial volume averaging. 48. Echo Spacing Echo spacing is the time from the middle of one echo to the middle of the next echo. Several factors determine echo spacing, including: Gradient slew rate. Receive bandwidth. Number of shots. Frequency-encoding steps. Frequency field of view. 49. Receive Bandwidth In conventional MR imaging, the receive bandwidth can range from 3 to 16 khz. Fast spin-echo sequences use receive bandwidths of 16 to 32 khz. EPI sequences use receive bandwidths of 32 khz, 64 khz or 184 khz, depending on the system s hardware and software configuration. Using a higher receive bandwidth reduces echo space because the frequency gradient is on for a shorter amount of time. Using a higher receive bandwidth decreases the signal-to-noise ratio and the images are grainier. 50. Single- or Dual-Echo Conventional Spin-echo 2012 ASRT. All rights reserved. 7 MR Basics: Module 6

9 Let s go back to the conventional spin-echo pulse sequence. Remember that the sequence starts with a 90-degree RF pulse to flip the net magnetization vector to the transverse plane. Then a 180-degree RF pulse is applied as the vector recovers and realigns with the main magnetic field. The signal formed when the 180-degree pulse is turned off is the spin-echo. The conventional spin-echo pulse sequence typically produces one echo per TR period. If the radiologist requests 20 slices, the exam produces one image per slice location, for a total of 20 images. Spin-echo pulse sequences were used routinely to acquire single-echo T1-weighted images, but also could be used to obtain dual-echo images. A dual-echo image contains both a proton-density image and a T2-weighted image in a single slice. The proton-density-weighted image is considered free because it does not take any extra scan time to obtain it. The T2- weighted image usually is acquired when the TE is 90 milliseconds. Stopping at the 30- millisecond interval to acquire an extra image does not add additional time to the scan. 51. Calculating Spin-Echo Scan Time The formula on this slide is used to calculate the imaging scan time for a conventional spin-echo sequence. The acquisition time is equal to the TR times the phase-encoding matrix size times the number of signals averaged divided by 60,000 milliseconds. In this example, we re given a TR of 2,000, a matrix size of 256 and an NSA of 2. Solving the equation gives us a scan time of about 17 minutes. Fast spin-echo has replaced the conventional dual-echo spin-echo sequence. In a 2-D conventional spin-echo scan, the TR, phase-encoding matrix size and NSA are the only parameters that directly affect scan time. Other scanning parameters, such as the number of slices, bandwidth and field of view, may affect how the MR technologist chooses the TR, matrix size and NSA, which means they indirectly affect scan time but are not included in the scan time calculation. 52. Knowledge Check 53. Knowledge Check 54. Knowledge Check 55. Knowledge Check 56. Fast Spin-Echo Fast spin-echo imaging was developed to produce high-resolution images but in a much shorter scan time. A 20-slice T2-weighted brain series using a spin-echo with a matrix of 256 x 192 may take 18 to 20 minutes. A fast spin-echo sequence can acquire the same series with a higher matrix size (say 512 x 256) and therefore higher resolution, in less than four minutes. Look at this illustration. As with conventional spin-echo pulse sequences, fast spin-echo imaging uses a 90-degree RF pulse, but fast spin-echo applies multiple 180-degree RF pulses. The long row of 180-degree pulses is called an echo train length. Another big difference between a conventional spin-echo sequence and a fast spin-echo sequence is the phase-encoding pulse ASRT. All rights reserved. 8 MR Basics: Module 6

10 Only one phase-encoding pulse is used per TR in a conventional spin-echo sequence, but fast spin-echo imaging uses multiple phase-encoding pulses per TR. The echo train length determines how many phase-encoding steps are performed for each TR period. In the above example, eight 180-degree pulses are used per TR period, which translates to an echo train length of eight. The scanner also uses eight phase-encoding pulses per TR period. The phase-encoding pulse fills one line of k-space per TR period. If the MR technologist selects a phase matrix of 192, then the spin-echo pulse sequence is repeated 192 times to fill each line of k-space. In a fast spin-echo sequence, multiple lines of k-space are filled per TR period. In this example, eight lines of k-space are filled per TR, which means the 192 lines of k- space are filled eight times faster than with conventional spin-echo imaging. So, the longer the echo train length, the faster the scan time. 57. Fast Spin-Echo and K-space In fast spin-echo imaging, multiple lines of k-space are filled for each TR period. In this example, the echo train length is four echoes. For each TR period, four lines of k-space data are collected or filled. Remember that contrast data is placed in the middle section of k-space and resolution data is placed in the outer edges of k-space. In fast spin-echo imaging, the phase-encoding gradients are reordered by varying the amplitude of the gradient. Low-amplitude gradients fill the center lines of k-space, and higher-amplitude gradients fill the outer lines of k-space. For a fast spin-echo scan, the MR technologist selects an effective TE based on the type of image contrast desired. In this example, the technologist chose the second echo as the desired contrast weighting. The MR scanner places the information contained in this echo in the center of k-space. The system applies a higher amplitude gradient to the first echo, and that image data is placed toward the outer lines of k-space. 58. Calculating Fast Spin-Echo Scan Time Calculating the acquisition time for a fast spin-echo sequence uses the same parameters as a conventional spin-echo sequence: TR, phase matrix size and number of signals averaged. Multiplying the three parameters gives us a product that can be divided by the echo train length, and dividing that number by 60,000 milliseconds produces the scan time, which can be converted to seconds. Solving the equation gives us an acquisition time of 32,000 milliseconds. Dividing that result by 60,000 milliseconds and then multiplying by 60 seconds gives us a time of approximately 32 seconds. For the fast spin-echo sequence calculation, the TR, phase matrix size, NSA and echo train length are the only parameters that directly affect scan time. 59. Knowledge Check 60. Knowledge Check 61. Knowledge Check 2012 ASRT. All rights reserved. 9 MR Basics: Module 6

11 62. Knowledge Check 63. Fast Spin-Echo Pulse Sequences When using a fast spin-echo pulse sequence, the MR technologist should select the minimum TE. TE is controlled by many variables, such as gradient hardware, the slew rate and rise time, matrix, field of view, slice thickness and bandwidth. Once the MR technologist selects these individual parameter settings, the MR scanner s hardware system dictates the minimum TE. For example, if the minimum TE is 20 milliseconds, echo 1 is 20 milliseconds, echo 2 is 40 milliseconds, echo 3 is 60 milliseconds and so forth. Each echo is a multiple of the minimum TE. A 6 echo train length with a minimum TE of 20 milliseconds produces a sequence with the following echoes: 20, 40, 60, 80, 100 and 120 milliseconds. TE is the primary controlling factor for T2 contrast weighting. Higher TEs produce more T2 contrast in the tissues. In fast spin-echo imaging, longer echo train lengths (that is, eight or more) increase T2 weighting. The later echoes contain more T2 contrast information. 64. Fast Spin-Echo Tradeoffs Increasing the echo train length increases the speed of the scan, which reduces the overall scan time. As the echo train length increases, the MR system might reduce the number of available slices for imaging. To regain those slices, the MR technologist may choose to increase the TR parameter. Increasing TR also increases scan time. Fast scanning can produce images with increased edge blurring. The technologist may choose to increase the receiver bandwidth to compensate for the blurring. Increasing the receiver bandwidth can decrease the image signal. Increasing the echo train length also can decrease image contrast weighting. If the technologist wants to use proton-density weighting, using an echo train length of more than six echoes might result in mixed contrast weighting. 65. Fast Spin-echo Tradeoffs Along with the advantages of faster scan times there comes disadvantages. In a fast spin echo the repetition time is increased allowing for more longitudinal relaxation which will produce more T2 weighting in the image. This will cause fat to appear brighter than usual on a T2 weighted image. Due to the increase in echo train length later echoes will be included in the information in K space. These later echoes will contain less signal to noise. This sequence is more sensitive to image blurring, flow and motion artifact, but less sensitive to detecting hemorrhages. Fast spin echo sequences have become part of a routine protocol in most clinical settings because of the faster scan times. However always critique your images and be willing to substitute with another pulse sequence if necessary. 66. Fast Spin-Echo Sequence In fast spin-echo imaging, edge blurring increases as the echo train length increases. To reduce edge blurring, the MR technologist can adjust several parameters, such as: Decrease echo train length. Increase bandwidth. Increase frequency field of view; decrease phase field of view. Reduce slice thickness ASRT. All rights reserved. 10 MR Basics: Module 6

12 Increase phase-encoding matrix (for example, from 128 to 256). Reduce frequency-encoding matrix (for example, from 256 to 224). Applying these parameters reduces the space between the echoes. These same parameter adjustments reduce blurring in the EPI pulse sequence, which we ll discuss later in this module. 67. Gradient-Echo Pulse Sequences The gradient-echo pulse sequence was an early sequence that was much faster than conventional spin-echo imaging. Whereas spin-echo sequencing uses a 180-degree RF pulse to refocus vectors, a gradient-echo sequence uses gradients to refocus vectors. As with conventional spin-echo sequences, the technique only fills 1 line of k-space per TR period. Gradient-echo sequences are best used for T2* contrast weighting, but do not correct for inhomogenieties to the magnetic field. 68. Gradient-Echo and Spin-Echo Pulse Diagrams This slide compares gradient-echo and conventional spin-echo pulse sequence diagrams: Both use an RF excitation pulse and gradients for slice selection and phase and frequency encoding. Both also generate echoes. The RF pulse is different in the two diagrams, however. Whereas the conventional spin-echo diagram contains both a 90-degree and 180-degree RF pulse, the gradient-echo diagram doesn t have a 180-degree RF pulse. 69. Gradient-Echo Pulse Sequences A gradient-echo sequence starts with an RF excitation pulse that ranges from 5 degrees to 60 degrees. The smaller pulse places less of the net magnetization vector in the transverse plane. When the frequency-encoding gradient (Gz) is turned on, a negative gradient field is applied, causing the net magnetization vector to dephase. This is represented by the red portion of the frequency encoding gradient on the gradient-echo pulse sequence diagram. The frequencyencoding gradient quickly flips and applies a positive gradient field (the blue portion) to rephase the signal, producing an echo. The frequency-encoding gradients (Gx) refocus the echo rather than the 180-degree RF pulse used in spin-echo sequences. By using a smaller flip angle, the sequence is faster and allows for a much shorter TE. Using gradient pulses eliminates the need for 180-degree refocusing pulses, which means less RF energy is deposited in the patient. A smaller amount of RF energy means less tissue warming, or a lower specific absorption rate. 70. Inversion Recovery The inversion recovery pulse sequence diagram looks similar to a conventional spin-echo pulse sequence. The inversion recovery sequence starts with a 180-degree inverting pulse; a 90- degree excitation pulse then is applied after a period of time elapses. The inversion time is the period between the middle of the inverting 180-degree RF pulse and the middle of the subsequent exciting 90-degree pulse. From that point, the inversion recovery sequence resembles a conventional spin-echo sequence. 71. Inversion Recovery The inversion recovery pulse sequence is used when the MR technologist wants to suppress the signal from a specific tissue to better display or isolate another tissue. The sequence is used to suppress the signal from fat so that any tissue containing fat is nulled and appears dark. Using 2012 ASRT. All rights reserved. 11 MR Basics: Module 6

13 an inversion recovery sequence can help radiologists determine whether a mass is filled with fat or fluid. Look at this illustration. Both fat and water have a net magnetization vector in the longitudinal plane that flips with the 180-degree RF pulse into the negative longitudinal plane. However, the net magnetization vector must be in the transverse plane so that the receiver coils can interpret the echo or signal. Neither fat nor water generate a signal while in the negative longitudinal plane. When the initial 180-degree RF pulse is turned off, both net magnetization vectors return to align with the positive longitudinal plane. The selected inversion time allows both of these tissues to recover and then the 90-degree RF pulse is applied. In this example, the MR technologist wants to suppress the fat signal. The 180-degree pulse flips fat and water into the negative longitudinal plane. The inversion time for fat is 250 to 300 milliseconds, so after 300 milliseconds, fat has returned to zero and water has returned to the positive longitudinal plane. When the 90-degree RF pulse is applied, water is flipped into the transverse plane, but fat remains in the negative longitudinal plane. Since the net magnetization vector must be in the transverse plane for the receiver coils to collect the signal, the sequence only acquires signal from water, which is in the transverse plane. 72. Calculating Gradient-Echo Scan Time The calculation for the gradient-echo sequence scan time uses the same parameters as a conventional spin-echo pulse sequence: TR, phase matrix and number of signals averaged. The product of the three parameters is multiplied by the number of slices, and that total is divided by 60,000 milliseconds to get a scan time in minutes. In this example, we re given a TR of 30 milliseconds, a phase matrix of 192, an NSA of 2 and 10 slices. Solving the equation gives us a scan time of approximately 1.92 minutes. In a gradient-echo sequence, TR, number of phase encodes, NSA and number of slices are the only parameters that directly affect scan time. Flip angle has no direct effect. 73. Knowledge Check 74. Knowledge Check 75. Knowledge Check 76. Knowledge Check 77. Gradient-Echo Sequences Gradient-echo pulse sequences are useful for a number of different examinations, such as vascular flow and cerebrospinal fluid (CSF) imaging, and dynamic studies. 78. Gradient-Echo Sequences These sequences are faster than conventional spin-echo sequences, with TRs as low as 18 milliseconds, shorter TEs and the ability to use excitation pulses from 1 degree to 180 degrees. Faster speeds also permit the use of breath-hold techniques that help minimize motion artifacts ASRT. All rights reserved. 12 MR Basics: Module 6

14 79. Gradient-Echo Sequences A 180-degree pulse contains four times more energy than a 90-degree pulse. Because the gradient-echo sequence doesn t use a 180-degree refocusing pulse, less RF energy is delivered to the patient. The low specific absorption rate makes gradient-echo the pulse sequence of choice for pregnant patients. 80. Gradient-Echo Sequences MR technologists use gradient-echo pulse sequences for abdominal imaging. Because the pulse sequences are very fast, images of abdominal anatomy can be obtained while the patient holds his or her breath. Holding the breath keeps the abdominal anatomy from rising and falling, which creates fewer motion artifacts and less ghosting. 81. Gradient-Echo Sequences MR angiography uses gradient echo pulse sequences to acquire multiple images quickly. The images are referred to as raw data. The raw data is then reformatted to construct a Maximum Intensity Projection (MIP) in a coronal, axial and sagittal plane. The image on the left is one of 114 images acquired in 6 minutes and 40 seconds. These 114 images are all raw data that was used to construct the sagittal MIP, which is the image on the right. 82. Gradient-Echo Pulse Sequence A gradient-echo uses a smaller flip angle, usually between 20 degrees and 35 degrees, but less than 90 degrees. A smaller flip angle means less time for longitudinal relaxation. When the RF pulse is removed, the free induction decay signal is produced and T2* dephasing occurs. T2* is made up of both magnetic inhomogeneities and the dephasing T2 signal. The magnetic moment is rephased by the gradient instead of a 180-degree refocusing pulse used in spin-echo sequences. 83. Gradient-Echo T2-Weighting Tissues vary in how they behave in a magnetic field. Each tissue has an inherent T2 decay time. The spins move out-of-phase with one another (phase incoherence) and transverse magnetization decreases over time. The tissues precessional differences set off a local field gradient that causes dephasing. T2 effects are caused by outside, or extrinsic, factors. These factors make normal T2 dephasing occur more quickly. In the case of air/water interfaces such as the sphenoid sinus, there is a resulting signal drop off. Gradient-echo sequences cannot correct for T2 effects. T2* is the type of image contrast created using a gradient-echo sequence. T2* is the product of T2 time and the T2 effects that influence normal decay time. A tissue s T2* is shorter than the normal decay time. T2* effects include magnetic field inhomogeneties, magnetic susceptibility and chemical shift artifact. When the radiologist asks for T2*, he or she expects the MR technologist to run a gradient-echo pulse sequence. 84. Echo Planar Imaging Echo planar imaging was developed using gradient-echo principles and making them even faster; EPI is one of the fastest pulse sequences available. As with fast spin-echo imaging, EPI fills multiple lines of k-space per TR period, but instead of using multiple 180-degree RF pulses, EPI 2012 ASRT. All rights reserved. 13 MR Basics: Module 6

15 uses alternating frequency encoding pulses to generate multiple gradient echoes. Both fast spinecho imaging and EPI use multiple phase-encoding steps per TR period to fill multiple lines of k- space per TR. Filling multiple lines of k-space per TR significantly reduces imaging time. The phase-encoding matrix chosen for image resolution determines how many lines of k-space must be filled. If a phase matrix of 192 is selected, then 192 phase-encoding steps must be completed. The 192 phase-encoding steps may occur one at a time per TR period, as in conventional spin-echo and gradient-echo imaging or multiple times per TR period, as in fast spin-echo imaging and EPI. In EPI, single-shot means that all lines of k-space are filled in a single TR period. Multishot means that lines of k-space are filled in two or more TR periods. 85. Calculating EPI Scan Time To calculate the scan time for an EPI pulse sequence, the following parameters are used: TR, number of shots and number of signals averaged. These three parameters are multiplied together and the total is divided by 60,000 milliseconds to calculate a scan time in minutes. In this example, we re given a TR of 2,500 milliseconds, the number of shots is eight and the NSA is four. Solving the equation equals a scan time of 1.33 minutes. 86. Knowledge Check 87. Knowledge Check 88. Knowledge Check 89. Knowledge Check 90. EPI vs. Fast Spin-Echo Imaging Both fast spin-echo sequences and EPI fill multiple lines of k-space per TR. Fast spin-echo uses an echo train of multiple 180-degree RF pulses, and EPI uses an echo train of gradient echoes. In addition, fast spin-echo is not sensitive to susceptibility artifacts such as metal and blood, whereas EPI is sensitive to susceptibility artifacts. 91. Increased Resolution Because EPI is a much faster MR sequence, the matrix size can be increased to provide better resolution in relatively short scan times. This image is an axial view of the brain that was acquired using both a 512 x 512 matrix and a 192 x 192 matrix. Notice the increased detail in the orbital area on the image on the left. 92. White Matter Sensitivity The EPI image on the right demonstrates more sensitivity to changes in the white matter in the brain s anterior compared with the fast spin-echo image on the left. 93. Faster Imaging Applications EPI is so much faster than other MR sequences that patient breath-hold techniques can be used. The image on the left is a conventional spin-echo T2-weighted image with an 8-minute scan time. The fast spin-echo T2-weighted image in the image in the center had a 20-second breath ASRT. All rights reserved. 14 MR Basics: Module 6

16 hold scan time with increased signal (bright) from fat tissue. The image on the right is an EPI T2*-weighted image with an 18-second breath-hold scan time and no signal from fat tissue. Removing the fat signal enables better visualization of the abdominal anatomy. 94. Ultrafast EPI Single-shot EPI pulse sequences can be used in ultrafast imaging applications, such as for claustrophobic or combative patients or for patients who are unable to remain still during conventional MR imaging. The sequences allow the MR technologist to take a snapshot of the entire brain in a short period of time. The 25 axial T2*-weighted brain slices shown on this slide were obtained in 1 minute and 18 seconds. Although the 25 slices were acquired quickly, ultrafast EPI imaging has many trade-offs, such as loss of signal and resolution, and increased susceptibility artifacts. Ultrafast sequences have specific purposes and should not take the place of other imaging pulse sequences. 95. EPI Sequences for Stroke Acute stroke may be detected by using EPI sequences known as diffusion-weighted imaging and perfusion-weighted imaging. These sequences provide physiological details about the brain. Normal brain tissue has freely diffusing water, which appears dark on images. In the event of a stroke, water becomes restricted, thereby appearing bright. The contrast of diffusion-weighted images is controlled by selecting the appropriate b-value. A higher b-value increases the background suppression. Perfusion-weighted imaging evaluates regional blood flow in patients with suspected stroke. MR images are acquired rapidly and repeatedly during intravenous (IV) contrast administration. An apparent diffusion coefficient (ADC) map is a postprocessing technique that is reconstructed from a diffusion-weighted imaging sequence with at least two b-values. If an area appears hyperintense on the diffusion-weighted image and hypointense on an ADC map, the stroke is real. If the area is hyperintense on both the EPI and the ADC map, it is not a stroke, but rather the result of T2 shine-through, and there may be a diffusion-perfusion mismatch on the images. The perfusion abnormality is larger than the diffusion abnormality indicating an area of at-risk tissue. 96. Diffusion-weighted Imaging Look at these two images. The image on the left is a diffusion-weighted scan. The bright signal demonstrates a stroke. The image on the right is a perfusion map. The dark areas around the right lateral portion of the brain demonstrate restricted blood flow due to a stroke. 97. Dynamic Studies Dynamic scans are motion studies. A kinematic joint study examines the patient s joint as he or she moves it in a series of positions. The images are linked together and played back on a monitor in a movie format known as a cine loop so the radiologist can evaluate joint movement. An example of a kinematic joint study is an examination that demonstrates the temporomandibular joint, or TMJ. A TMJ bite block or plate is placed between the patient s front teeth to open the jaw to different positions. Images are acquired with the patient s jaw open various distances (for example, 2.0, 1.5, 1.0 and 0.5 cm). The MR technologist then links the images to create the cine loop, which shows the jaw opening and closing and demonstrates 2012 ASRT. All rights reserved. 15 MR Basics: Module 6

17 the mandibular condyle and its position. The radiologist also can see if the temporomandibular disk slips or is intact. Cine applications are used to display the beating heart and blood flow, and cardiac gating can show the heart at different cardiac phases. For example, physicians can rule out an aortic dissection by observing blood flow from the heart through the aortic arch and down the aorta. 98. Scalped Image The images on this slide represent the same axial slice of the abdomen using a spin-echo pulse sequence, a fast spin-echo sequence and an EPI pulse sequence. All of the images used a TR of 2,500 and a TE of 80. The scan time for the conventional spin-echo was 8 minutes 30 seconds; the fast spin-echo sequence was 3 minutes 50 seconds; and the EPI was 2 minutes 3 seconds. Remember that one of the trade-offs of EPI technique is loss of signal and resolution. Notice the lack of definition and detail in the image on the right. 99. Magnetic Susceptibility Artifacts Each tissue has an inherent T2 decay time. Exposing tissues to extrinsic factors such as inhomogenities in the main magnetic field, chemical shift and magnetic susceptibilities on or inside the patient can accelerate the normal T2 decay time. Magnetic susceptibilities associated with patients include items such as braces, metal teeth fillings, metal from surgical implants, cosmetics and hair products. In addition, street clothing or accessories may contain metallic fibers, buttons, zippers or ink dye that can alter the magnetic field Magnetic Susceptibility Artifacts EPI and gradient-echo pulse sequences are more prone to artifacts from internal or external metallic objects. Using these sequences to scan an area containing implanted metal such as braces, dental work, screws, plates or wires can greatly reduce image quality and render the images nondiagnostic. The patient in this image was wearing an elastic hairband that had a metal component. The metal created a large inhomogeneity artifact that renders the image useless Geometric Distortion When off-resonant water protons accumulate, a phase shift occurs during the readout gradient, causing a geometric distortion. As the readout gradient time increases, so does the distortion. Geometric distortion is most noticeable at tissue/water interfaces such as the sinuses. Minimizing the echo spacing decreases geometric distortion. Other methods of reducing geometric distortion are: Using multishot rather than single-shot EPI. Performing a localized shim over the region of interest. Using thinner slices. Using shorter TEs Knowledge Check The technologist is able to manipulate a number of parameters that impact scan time, signal-tonoise ratio and spatial resolution. It s important to understand the relationship between these parameters. The following exercise helps demonstrate these relationships. Select a parameter from the drop-down menu. Drag the slider bar to increase or decrease the parameter. Notice how SNR, resolution and scan time change ASRT. All rights reserved. 16 MR Basics: Module 6

18 103. Conclusion This concludes Module 6 of MR Basics Pulse Sequences. You should now be able to: List parameters related to tissue characteristics that affect image quality, such as spin density and T1 and T2 relaxation. Identify other parameters that affect image quality, such as repetition time, echo time, inversion time and flip angle. Apply pulse sequence principles to magnetic resonance (MR) imaging. Explain the concepts of image formation in MR imaging. Describe image contrast appearance according to image weighting Bibliography 105. Acknowledgements 106. Development Team 107. Module Completion 2012 ASRT. All rights reserved. 17 MR Basics: Module 6

19 Bibliography Westbrook C, Roth CK, Talbot J. MRI in Practice. 3rd ed. Malden, MA: Blackwell Publishing Inc; Woodward P, Freimarck RD, eds. MRI for Technologists. New York, NY; McGraw-Hill Inc; Schrack T. Echo planar imaging [applications guide]. Milwaukee, WI: General Electric Company; Accessed January 31, ASRT. All rights reserved. 18 MR Basics: Module 6

20 Acknowledgements Subject Matter Experts Cathy Dressen, M.H.A., R.T.(R)(MR) Special Thanks to GE Healthcare Project Management and Instructional Design Charlotte Hendrix, Ph.D., SPHR Narration Vince Ascoli Lead Designer Nicholas Price Module Production Sharon Krein Jake Rumanek 2012 ASRT. All rights reserved. 19 MR Basics: Module 6

21 MR Parameters Actions and Associated Trade-offs In MR imaging, there are no hard-and-fast rules. MR technologists choose a pulse sequence based on the area of interest, the condition of the patient and other clinical factors or protocols. For the new MR technologist, this aspect can be very frustrating, but in reality it is what makes the modality interesting and challenging. All MR technologists must be well-versed in the ramifications of various choices. The benefits and limitations of these choices are called trade-offs. This chart summarizes MR parameters and their associated trade-offs. MR Parameters Actions and Associated Trade-offs Parameter Action Benefit Limitation TR Increase Increased SNR Increased scan time Increased number of slices Decreased T1 weighting TR Decrease Decreased scan time Decreased SNR Increased T1 weighting Decreased number of slices TE Increase Increased T2 weighting Decreased SNR TE Decrease Increased SNR Decreased T2 weighting NSA/NEX Increase Increased SNR Direct proportional increase in scan time NSA/NEX Decrease Direct proportional decrease in scan time Decreased SNR Decreased signal averaging Slice Increase Increased SNR Decreased spatial resolution thickness Increased coverage of anatomy Increased partial volume averaging Slice Decrease Increased spatial resolution Decreased SNR thickness Decreased partial volume Decreased coverage of anatomy averaging FOV Increase Increased SNR Decreased spatial resolution Increased coverage of anatomy Decreased chance of aliasing FOV Decrease Increased spatial resolution Decreased SNR Increased chance of aliasing Decreased coverage of anatomy Matrix Increase Increased spatial resolution Increased scan time Decreased SNR if pixel is small Matrix Decrease Decreased scan time Decreased spatial resolution Increased SNR if pixel is large Receive Increase Decreased chemical shift Decreased spatial resolution bandwidth Decreased minimum TE Receive bandwidth Decrease Increased SNR Increased chemical shift Increased minimum TE Large coil - Increased area of received signal Decreased SNR Increased aliasing if using small FOV Increased chance of artifacts Small coil - Increased SNR Decreased area of received signal Decreased chance of artifacts Decreased aliasing if using small FOV FOV = field of view; NSA/NEX = number of signal averages/number of excitations; TE = echo time; TR = repetition time; SNR = signal-to-noise ratio ASRT. All rights reserved. 20 MR Basics: Module 6