Magnetic Susceptibility Distortions for Good Echo Planar Imaging - Essay Example

? By of Learning: Article Critique Paper Modules 3 to 5 discuss various methods to correct for magnetic susceptibility distortions, while maintaining a good EPI image. Consider the following clinical setting: You require a T2-weighted image which covers the whole of the brain, where the field of view and matrix size have been fixed at 23 cm x 23 cm, and 128 x 128 respectively. Introduction Magnetic susceptibility refers to the process y issue tissue affects magnetic field. The existing boundaries between fat and compact bone are normally affected by the transverse magnetic field de-phasing and signal distortion. The consequent of the above is that frequency positions in the wrong location in the phase direction. Different results have been found from different body areas which have varying magnetic field strengths thus such body areas end up having varying processional water frequencies. Blood has been observed to have the highest susceptibility as it has iron contents as one of its components. Compositions based on water also have high susceptibility in comparison with air (McRobbie 2007) (a) List the parameters that could be changed to optimize the EPI image in this case, plus how you would change the parameter (i.e. increase or decrease). (2 marks) Slew rate and the receiver bandwidth Each of the listed parameters has a different way of changing it. For example, the case of slew rate increment, an overall reduction in echo spacing as well as reduction in geometric distortion is required for it to be altered. For receiver bandwidth to be increased, an overall reduction in echo spacing, signal to noise ratio and geometric distortion need to be lowered (Van Der Zwaag, at el. 2012, p. 129) In a normal scenario, MR equipments take their time until the trapezoid gradient gets to the flat top. When this time comes, the data points will then be sampled by the MR equipments in the direction of the frequency. This implies that no data acquisition takes place during the rise time. The rise time refers to the time when the gradient is not yet at its minimum or maximum amplitude. However, by employing ramp sampling method, the data points witnessed during the readout gradient switching can be acquired during the rise time. a) Two different features can be used while employing ramp sampling, and they include; 1. Having the geometric distortion minimized; ramp sampling can be used to reduce the flat top while at the same time keeping the Nx constant. This implies that the echo spacing can be decreased consequently minimizing geometric distortion. 2. Increasing resolution; by keeping the flat top at the same level, more Nx data points can be fitted during readout and thus increasing the overall resolution. b). Using Conjugate Synthesis; Conjugate synthesis is a symmetry property which means that only half of the raw data spaces in the whole MR can be acquired to come up with a Mr image that is complete. The most desirable means to achieving high resolution during a single shot EPI experiments is through having the readout duration on the ky to be as long as possible so as to have only the negative and positive kx values. c). Slew rate enhancement d). Echo train length reduction e) Making use of shimming to minimize distortion d). Employing parallel imaging to minimize echo spacing f) . Making use of Multi shot EPI so as to ensure that there are no cases of phase errors building up. g). Through increasing the TE which will in turn increase the transverse magnetization hence larger T2* and T2 (b) Discuss the chain of consequences if you optimize the image by increasing the receiver bandwidth. (4 marks) Receiver bandwidth generally describes how fast an MR signal can be digitized. In cases where the receiver bandwidth is generally higher, there is a corresponding faster digitization of the produced image. This is because the aspect of receiver bandwidth is inversely proportional to the time taken during the imaging process. The receiver bandwidth generated normally determines the range of frequencies that would be sampled by the available frequency encoding gradient. There are several consequences that would occur as a result of optimizing the image in relation to the receiver bandwidth. For instance, whenever the receiver bandwidth would be increased, geometric spacing will generally reduce together with geometric distortion and signal to noise ratio. On the other hand, there will be no profound effect on spatial resolution as well as scan time. This is because such factors remain constant throughout the increase of the receiver bandwidth. Generally, upon the increment in the receiver bandwidth, there will be an overall reduction in the amount of time at the flat top of the available gradient. Shortening of the time of flat top is the trigger to the reduction of the ESP as well as geometric distortion of the image. Since the SNR is inversely proportional to the root of the received bandwidth, there would be a reduction in SNR as the receiver bandwidth gets increased (Schmiedeskamp, at el., 2010, p. 67) A rBW that is high implies that the process of signal sampling will take place at a faster rate now that the duration between data points has been reduced. In increase in the rBW results in a decrease in the duel time thus a faster signal acquisition. This is because the acquisition of the echo will be faster consequently reducing the gradient of the flat top. In such a scenario, the ESP will reduce and hence the acquisition process will be faster and less phase errors accumulation time. This will in turn reduce the geometric distortion presence. However, when the bandwidth is increased, the noise ends up being sampled thus the SNr decreases. To compensate this signal lack, the following can be done; 1. The coil technology being improved 2. More NEX to be acquired By employing the above named measures, the amplifier voltage has to be increased or alternatively reducing the gradient’s induction coil. The hands have to be left unclasped so as to stop PNS from increasing with the increase in slew rate. An increase in rBW will result in a reduction in the flat top and hence the ESP and geometric distortion will decrease. An increase in rBW will also lead to an increase in the difference between the sampled extreme frequencies. The sampled frequencies have to be sampled so as to come up with the right sample. If this is not done, acquired artefacts will be seen following the aliasing process. The two matrixes TE and rBW have significant effects on the data that has been acquired. Some of the witnessed impacts include; 1. Frequency range detection which are normally digitized in the readout phase 2. Chemical shift increase 3. Data points detection during the readout phase 2- What are the benefits of using circular spiral or square spiral EPI methods? (2 marks) Specific benefits come with the application of either the circular or the square spiral EPI’s. To start with, for the case of the square spiral EPI, the process of data collection begins from the middle and spirals in to a larger K as the time passes. This is majorly an advantage achieved through the use of different timing and amplitudes in the readout gradients which are supported by the square spiral EPI. Another advantage of the spiral EPI’s is that the data collected is acquired in a regularly spaced grid. This is because the data has to be shuffled for it to move from negative to the positive side of the K before applying a 2D Fourier transform for an image reconstruction (Mcdonald, Coronado, Johnson, 2003, p. 78) One of the main advantages of sampling using square spiral is that the acquisition of square spiral k space is normally done on a grid that is regularly spaced. Despite this advantage, the process leads to the shuffling of negative data to end up on the positive run. In such a case, the negative lines will have to be recorded for image reconstruction to take place (McMahon, 2012). As in the data acquisition process in the circular spiral method, there has data has to be further adjusted in relation to the data points spacing which in turn will lead to signal noise and resolution loss or in some cases reconstruction of non-Fourier (Franz Schmitt, 1998, p. 92) One outstanding advantage associated with circular EPI is that the continuous sampling of MR signal as well as the non sampling of the k space edges leads to an overall reduction in the acquisition time by 27% in comparison with Blipped EPI. In Circular spiral, the k space centre becomes entirely sampled and its periphery acquisition is faster. The circular spiral scheme results to a decrease in the resonance effects now that in this case the resonances effects appear more like blurred artifacts and not sharpened artifacts that have been well defined like what is normally witnessed in the blipped EPI case (McMahon, 2012). The circular spiral EPI sampling method is advantageous in that it produces essential gradients that are comprised of sinusoidal waveform of constantly increasing amplitude. Another main advantage of the application and use of the circular spiral EPI is that it produces an animation whenever the frequency and phase gradients are sinusoidal modulated at the same time (McMahon, 2012). 3- Discuss the advantages and disadvantages of using segmentation in EPI. (8 marks) In single shot EPI, all the steps taken during phase encoding are completed in one repletion time. This implies that, Ny represents the echo rain length which is the phase encoding steps number. By employing this approach, it is possible to acquire a data set that is multi-sliced in 20-100ms. In such a situation TR tends to infinite now that only a single excitation pulses set has to be used in every image (McMahon, 2012). Figure 4.1: Diagram of single shot EPI and k -space coverage. Segmented EPI approach has its readout broken down into segments referred to as multiple shots in that; This calls for the need of several repetitions so to have the entire k space covered (Masters, 2006, p. 67) The segmentation degree can be looked at as the complete shots number used in full image acquisition or the echo train length for every excitation. A good example of the above is the acquisition of 256 steps of phase encoding in 32 shots with an echo train length corresponding to 8 (McMahon, 2012). There are both disadvantages as well as advantages associated with the use of segmentation in EPI. To begin with, there are many benefits got through the application of the segmented EPI. For instance, segmentation acts as an added advantage to the EPI model in that it is used in the aspect of resolution increment. This is because it is a trigger to the susceptibility of the brain motion. The aspect of segmentation for the case of EPI places less stress on the gradient in relation to a single shot EPI. This is because a segmented EPI runs very well in conventional systems which cannot in any way support the application of a single shot EPI. Segmentation serves as the solution especially in situations where it is relatively hard to obtain the necessary k-space data prior to the transverse relaxation eliminating the MR signal. Another major benefit of the segmented EPI is that it reduces the overall time in which phase errors accumulate in relation to the single shot EPI which supports longer times for such errors to accumulate. This then ensures that there is an overall reduction on the rate of magnetic susceptibility artifacts. Furthermore, segmented EPI allows for the T1 weighting to be introduced in the system (McRobbie, 2007). Advantage of segmented EPI: 1. Segmentation can be used to increase the quality of the final image as well as increase resolution. 2. The scanning time in segmented EPI is normally reduced and this is normally achieved, Fast Spin Echo technique or in some cases Breath hold technique are employed. 3. Now that the bandwidth in segmented EPI is high in its phase encoding, distortions and chemical shifts effects are greatly reduced. 4. As compared to the single shot EPI, segmentation EPI has less negative effects on the gradients Segmentation EPI plays a significant role when the SNR causes difficulties in obtaining the required k-space data before the MR signal is eliminated using transverse relaxation. This leaves segmented EPI the best choice for conventional systems where single shot EPI is not applicable. 5. Less gradient stress in segmented EPI ensures that less acoustic noise is witnessed 6. As a result of the finite TR, it is easy to achieve T1 weighting 7. Segmented EPI has been credited for its ability to reduce susceptibility artifacts now that it witnesses less build up time. 8. The SNR in segmented EPI is usually increased now that less echo trains are acquired by every RF excitation. 9. Segmented EPI does not require gradient systems that perfume highly. According to research, several limitations have also been associated with the application of the segmented EPI. To begin with, upon the use of the segmented EPI, more time gets consumed. This is because it generally takes longer to perform tasks in comparison to the single shot EPI. For instance, it supports a longer scan time hence slower than the single shot EPI. Due to the longer scan time, the segmented EPI becomes highly vulnerable to motion artefacts in comparison the single shot EPI. Another disadvantage associated with segmentation is that the EPI become more vulnerable to specific artifacts due to the complex coverage of K-space. Segmented EPI Disadvantages 1. The increase in ESP period in segmented EPI results in the appearance of ghosts with multiple high intensity. 2. The approaches takes longer in its image acquisition following the time between segments 3. Segmented EPI approach requires time that possibly used in techniques with resolutions of high temporal like perfusion and diffusion. 4. Now that the K space coverage in segmented EPI is complicated, various artifacts are normally experienced such as discontinuity in the k space leading to ghosts 5. It takes more time to be completed. The time needed for completion is reached at using the formula; TR*Ns * NEX 6. The Mosaic type of segmented EPI is vulnerable to chemical shifts, nyquist Ghosts and susceptibility. 4- What is the difference between SMASH and SENSE reconstruction in parallel imaging? (4 marks) Each of the pixel located in the image files effect SENSE and are normally produced following the single factor and this takes place collectively and in an array resulting from the action of several signals coming from different location, the proton density which assists in the production as well as the information regarding the coil sensitivity. Reference scan has proved to be important especially when it comes coil sensitivity observation as it plays a crucial role in the separation of every signal form each pixel and varying spatial positions. Figure : Reference scan acquired with the body coil, b) a directreconstruction of the information coming from one element of the coil array, and c) final SENS reconstruction Cha Characteristics of SENSE: 1. Corrects Aliasing issues 2. It allows for the provision of single pixel. 3. It utilizes data domain after the imaging process involving Fourier transformation 4. A single pixel in SENSE can be used to solve several equations 5. Fourier transformers are used to control each coil 6. SENSE does not limit its coil arrangements thus making it mostly used in exponential arrangements 7. It needs few steps to have in its phase encoding gradient 8. SENSE cannot be used in imaging of dual oblique cardiac process because it is not effective when it comes to identifying unfolded image 9. Does not produce extra noise as a result of reduced SNR Simultaneous acquisition of spatial harmonics (SMASH): Fast imaging with radiofrequency coil arrays. Magnetic Resonance in Medicine38: SMASH SMASH operates under the phase encoding principle which has it that the phase is dependant on the magnetic shift. The function in this case is regarded as the shift’s gradient in the actual one, the cosine cycle in the field while at the same time the sine function is looked at as the imaginary one. In SMASH, every element in each image seen in the array in made up of a filed view that has been reduced. The area of interest in this case is also sampled. Several variations in the strategies used in reconstruction that take place as a result of the situations that are constantly changing as seen during the image wrapping process. Considering the reconstruction stand witnessed in SENSE with regards to encoding sensitivity and the simultaneous acquisition stands in SMASH during spatial harmonics, the partial phase encoding coming from each element of the coil is observed to be completely different from one scenario to the other. This normally leads to the production of image reconstructions in both SMASH and SENSE. The SENSE and SMASH reconstruction differ in such away that when it comes to parallel imaging, SMASH reconstruction witnesses its imaging process being achieved with the use of data that is raw while in the SENSE case, the process of imaging involves the use of data that is image domain. The other difference that is evident between SMASH and SENSE is that no restrictions are observed when it comes to SENSE with reference to the arrangement of any special element in the phased array. This is normally not the situation in SMASH parallel imaging (Blaimer, M et al., 2004, p. 89) SMASH makes use of a signal combination coming from the coils array so as to copy spatial encoding that is normally assisted by the encoding gradients while on the other hand SENSE makes use of reference scans to determine the coil sensitivity in every available phase array. SENSE operation depends on the fact that every pixel forms an image using a single element but this does not apply in SMASH. In SENSE, the first step is the acquisition of a reference scan related with the coil body after which the information is directly constructed from the coil array’s single element followed by the final reconstruction of the SENSE. In the case of SMASH, the gradient of phase encoding results in a modulation that is sinusoidal in the signal which is taken to be different K space lines. The spatial harmonics that are sinusoidal found in the array continue to operate in the same way thus the final reconstruction in SMASH occurs. The main difference between identical pictures and phased-array picture is that in the case of identical picture, the understanding routine of the coil is used as the program for development. This understanding provides additional development after the slope development. This permits the development procedure to become more identical as several details that are encoded with sensitivity are achieved at the same time with several rings. SENSE approach uses the coils accurate knowledge to understand the details of spatial development. The coils understanding forms a smooth function which can be determined using scans of low resolution and can also be used as additional strategies for development. Through determining the understanding and later using it in changing the encoding matrix in an effort to restore the picture, the various samples needed for Fourier development can be cut down. The strategy in SMASH is the same as what was seen in SMASH as it makes use the coils understanding in the encoding process. Despite of this similarity, SMASH goes a head to make use of these details in ways that are totally different from SENSE. SMASH operation depends on coming up with sinusoids from the sensitivity patterns of the coil. By doing this, it ensures that the losing lines in the k space are filled up by coming up with sinusoids which normally could be produced using gradients. The Parallel Imaging With Localized Sensitivity (PILS) approach can be regarded as being easier as it uses the fact that every coil happens to be vulnerable to different picture FOV regions. This means that different coils pictures can be cut and then copied together in a simple process. This approach can be applied in a unique SENSE situation in cases where the rings understandings are orthogonal to one another (Blaimer, M et al., 2004, p. 89). The PILS approach has so far been appreciated as being eye- catching now that it does not need coils understanding routine measurement, complex reconstruction or rings calibration. This approach also offers disturbances that are of better quality compared to both SMASH and SENSE and has less relics as compared to those witnessed in SMASH. PILS is also durable in the sense that it results in pictures that are truly similar using renovation and purchase programs that are completely separate among the various rings. Various rings can be put at locations which call for attention after which the separation of pictures can take place. Now that individual reconstruction of pictures from each coil is possible, the approach can also be applied in situations where multiple rings have been placed in locations that are far away. Such situations will be made up of things like both legs multiple pictures. The advantage of PILS is that its renovations is easy to compute even when applied in the K space trajectories used are irrelevant. This is not possible in both SMASH and SENSE (Blaimer, M et al., 2004, p. 89) Bibliography FRANZ SCHMITT, A., M. K. S. & TURNER, R. 1998. Echo-planar imaging: theory, technique and application, Springer. JEFF D. WINTERA & TERRY THOMPSONA, A. N. G. 2007. Efficacy of motion artifact reduction in neonatal DW segmented EPI at 3 T using phase correction by numerical optimization and segment data swapping. MASTERS, B. R. (2006). Confocal microscopy and multiphoton excitation microscopy: the genesis of live cell imaging. Bellingham, Wash, SPIE Press MCDONALD, J. A., CORONADO, V. G., & JOHNSON, R. L. (2003). Questionnaire design. [Atlanta, GA], Departmentt of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Division of Reproductive Health. McMAHON, K. 2012. MRES7005. Fast imaging techniques. Queensland, Australia: University of Queensland.   MCROBBIE, D. W. 2007. MRI from picture to proton, Cambridge, UK, Cambridge University Press. SCHMIEDESKAMP, H, NEWBOULD, R D, PISANI, L J, SKARE, S, GLOVER, G H, PRUESSMANN, K P, & BAMMER, R. (2010). Improvements in parallel imaging accelerated functional MRI using multiecho echo-planar imaging. Info:Doi/10.1002/Mrm.22222. Wiley-Blackwell. http://dx.doi.org/10.5167/uzh-45306. VAN DER ZWAAG, W., MARQUES, J. P., KOBER, T., GLOVER, G., GRUETTER, R., & KRUEGER, G. (2012). Temporal SNR characteristics in segmented 3D-EPI at 7T. http://infoscience.epfl.ch/record/175310. Read More