2 Introduction and Theory
2.1 Larmor precession
2.2 Spine lattice and spin-spin relaxation time
2.3 Weighted imaging
2.4 Pulse sequence
2.4.1 The spin echo pulse sequence
2.4.2 The gradient echo pulse sequence
2.5.1 Slice selection
2.5.2 Frequency encoding
2.5.3 Phase encoding
2.7 Safety and risk assessment
3. Result and discussion
The MRI experiment was done with using oil and water. The safety steps were taken before entering the MRI room. A single spin echoes images were taken with different TR and TE to show the effect in oil and water. Different contrast was showing in the images. Then a multi-echo image was taken with TR 5000 ms is constant. After that a
Different tissues have different relaxation time like water and fat. The T2 is when the magnetic field of nuclei interact with other nuclei magnetic field. When applying an RF plus to flip the protons 90 degrees and they become coherent. When the RF pulse stopped the coherent between the protons start to decay, or another word protons start to relax. The protons will start to repeal each other because of the positive charge and move apart. When they move apart and lose coherent, this called T2 relaxation. The T2 occur in X, Y direction. It can be express in equation
The advantage of MRI compared with other modalities is the high contrast between different tissue. The contrast of MRI happens because of different timing and magnitude of RF pulse. When applying an RF pulse to protons and excitation occur, then a relaxation started before a second RF pulse is applied is called repetition time (TR). Repetition time is the time between first RF pulse and start of second RF pulse and measured in milliseconds. In another word, how quickly we put in the RF pulse. The TR determines the longitudinal relaxation time between first and second RF pulse which can see it connected to T1. The echo time (TE) is the time between the RF pulse and the peak of the signal. In another way to explain is how quickly we listen to the return signal. The TE is measured in milliseconds. TE is connected to T2 which controls time of decay of transverse magnetization (Allisy-Roberts & Williams, 2008).
Now the basic principle is covered, contrast and image weighting can be introduced to understand the idea behind different image scales. There are two majors’ categories that affect the contrast of the image. One is intrinsic parameters which cannot be altered because it is in the tissue, for example, T1, T2 and proton density. Second is extrinsic parameters which can be altered for example TR, TE, and flip angle. The area that emits high signal will appear as white, an area with the low signal will appear dark. The area that between high and low signal will have a different degree of grey colour. There are three important factors that affect the T1 and T2 relaxation. One is the inherent energy of the tissue. Tissue with low inherent energy can absorb more energy from hydrogen nuclei. This factor is important for T1 because of the interaction between tissue and lattice. Second is how close the molecules are to each other. The closer is the molecules to each other the better interaction between them which affect the spin-spin interaction. Last is the efficiency of the exchange rate between molecular and lattice. Now to make it easier, the effect of T1 and T2 in water and fat will be explained in details to show the effect in different tissue. Fat with the structure of hydrogen, oxygen, carbon and lipid. Fat has low inherent energy which means it can accept energy without difficulty from hydrogen nuclei. Furthermore, the efficiency of exchange energy with hydrogen nuclei is high. Because of these factors, the fat regains their longitudinal magnetization very fast, which mean fat has short T1(Westbrook et al., 2011). Now the fat molecular are close to each other, and their interaction is efficient which lead to short T2 time. For water which composed of two hydrogen water and one oxygen. The T1 has high inherent energy and low efficiency in energy exchange because of that it will take longer to regain their longitudinal magnetization. So, the T1 in water is long. T2 of water is long because the space between molecular is large which decrease the efficiency of interaction between hydrogen nuclei. Now T1 weighted image can be obtained by short TE and TR to decrease the effect of T2. A figure (1) demonstrates the effect of TR (Westbrook et al., 2011).
Figure (1) the effect of TR in the images (Westbrook et al., 2011)
From the figure (1) can show that short TR will not allow all the protons to regain their longitudinal magnetization which is important to differentiate between fat and water. If the TR was long, the different contrast would disperse. T2 weighted images can be better explained using figure (2). As shown in the figure (2) short TE will not allow enough decay time for both fat and water which lead to no separation. The longer the TE will allow better separation in T2 weighted images (Westbrook et al., 2011).
Figure (2) show the effect of TE in water and fat (Westbrook et al., 2011)
Inhomogeneities can cause a dephasing because of dephasing a fast loss of coherent magnetization occurs, the phenomena called
T2*.The tissue will not have enough time to produce a good contrast image. To overcome the
T2*a second 180 RF pulse is introduced or a gradient. Adding another pulse to regenerate the signal called the spin echo pulse sequence. Adding gradient to regenerate the signal called gradient echo pulse sequence (Bushberg, 2012).
An excitation pulse used to flip the NMV 90 degrees. Now the NMV is in the transverse plane. When the RF pulse stopped a
T2*dephasing start and free induction decay (FID) signal occur. Figure (3) help to imagine what happen in the spin echo pulse sequence (Westbrook et al., 2011).
Figure (3) Spin echo pulse sequence (Westbrook et al., 2011)
Some hydrogen nuclei will be faster (red color) than others (blue color). The fast hydrogen nuclei will be in the front, by applying a 180 degrees pulse to overcome the
T2*dephasing. Now the leading hydrogen nuclei become in the back and the slow hydrogen nuclei in the front. With some time, the fast nuclei catch the slow one and they become in phase. Because the nuclei in phase a maximum signal is produced. In the spin echo pulse sequence TR is between the first 90 degrees pulse and start of second 90 degrees pulse. The TE is between excitation pulse and the echo signal. A T1 weighted image can be produced by applying one echo pulse with short TE and short TR. By making TE short, will not allow enough time to decay. Moreover, making TR short will not give enough time for either water or fat to recovered. To produce a T2 weighted images, two echoes is use. The first echo has short TE and long TR. The second pulse has long TE and long TR which allow T2 to dominate the signal. The spine echo pulse sequence is the protocol of used in almost all procedure. It has a good image quality, true T2 weighting images and very versatile. However, the scan times is long. A new technique was introduced to decrease the time of scan which called turbo spin echo. The turbo spin echo does not change the TR, NEX or number of phase because that will affect image resolution and that is not desired. In the normal spin echo pulse sequence every TR filled one K-space line. In turbo spin echo multiple K-space line is filled in one TR by using echo train. The echo train happen because multiple 180 degrees pulse is used within one TR. In every pulse a, phase encoding is performed and a line is filled in K-space. Using the turbo technique will decrease the scan time. However, a flow and motion artefact will increase (McRobbie, 2006; Westbrook et al., 2011).
the RF excitation pulse is introduced, but this time it can go at any angle. After the pulse stop an FID signal produce and
T2*dephasing started. This time overcome the dephasing using gradient. A gradient is a magnetic field produce by electricity around gradient coil. Now in the axis there are two magnetic field which lead to high magnetic field in one end and decrease toward the other end. The direction is determined by the direction of the current. Using a gradient echo give some advantage like shorter scan time because it does not relay on 90 degrees angle. Furthermore, shorter TR is needed which lead to shorter scan time. The disadvantage of gradient echo is contained susceptibility artefact and do not compensate for inhomogeneities (Westbrook et al., 2011).
The signal is created and received, but the location of the signal in three dimensions is not recognized. To locate the exact location of the signal, first, a slice selection must have done then axes encoding using different gradients. MRI contains three gradients coil in Z, X, Y direction. When the gradient is switch on the magnetic field in axis change in linear fashion. One end will increase in magnetic strength which lead to the prediction of precession frequency of nuclei which called spatial encoding. Figure (4) demonstrate the effect of gradient on the magnetic field. By using gradients, another encoding can be done like slice selection, frequency encoding, and phase encoding (Westbrook et al., 2011).
Figure (4) the affect of gradient in magnetic field (Westbrook et al., 2011)
When a patient is scanned, a slice of the interest area is taken. The slice selection happens when a gradient in the Z-direction switches on. Because of the gradient, the linear precession frequency is changed. Every point on the axis of the gradient has a certain precession frequency. The nuclei in the interesting slice have a specific precession frequency. Now using an RF pulse with matching frequency of nuclei of the interesting slice. This called slice selection gradient. Other nuclei will not resonate because they have different precession frequency due to the effect of the gradient. The range of frequency that used to cover two points is called bandwidth. The smaller the bandwidth, the thinner slices (McRobbie, 2006; Westbrook et al., 2011).
Now a slice is selected, but more details are needed to provide a diagnostic image. The next step is to locate the signal in the X direction within the selected slice. The frequency encoding gradient is switch on which lead to a change of the strength in a linear fashion on X direction. The nuclei now will have different precession frequency. The frequency gradient work when echo signal received (McRobbie, 2006; Westbrook et al., 2011).
The last part to produce an MRI image is applying phase encoding gradient in the Y direction. The phase encoding gradient is switched on after the excitation pulse. When the gradient is switch on some nuclei in the higher strength of the magnetic field will phase faster and on the other side will be slower. Then the phase gradient will switch off. The nuclei have different phase but same speed. Now an image can be created because every nucleus has different frequency and phase encoding gradient. The nuclei with the high signal will show brighter than nuclei with low signal (McRobbie, 2006; Westbrook et al., 2011).
Now the signals within the interesting slice are recognized because of frequency and phase encoding. The data are stored in K-space. The k-space have a rectangular shape figure (5). The X direction is frequency, and Y direction is a phase. The unit used in K-space is radiant per cm. the concept of K-space is important to understand how data are stored and then transferred to image. Knowledge of K-space is important to understand a different way to decrease the time of the scan by using different techniques. The K-space is divided into the positive and negative line. The data can be negative or positive because of polarity and slope of the phase gradients. To simplify the K-space idea, image the K-space as a chest of drawers. To select a chest of drawers is slice selection (chest of drawers is the K-space). Now we need to choose a drawer to open this will happen when a phase encoding is applied (a drawer is one line in K-space). Every time a new drawer (line) is needed a change of TR is important or data will be in the same drawer every time. Negative phase encoding will lead to negative lines in lower part of K-space. Now applying a frequency encoding is to put data inside the line. Figure (6) show the process with the help of imaging K-space as a chest of drawers to simplify the idea. The data points which is collected is controlled the matrix size. This process is done within one TR. If the TR is long, more slice can be obtained but more scan time is needed. In another hand. Shorter TR lead to less slice can be obtained and shorter scan time. To move from one line to another an amplitude of phase encoding either decrease or increase in every TR. The K-space data is not an image, and the top line does not mean the top of the image. The data in K-space need to be transformed to the image by using a mathematical method called fast Fourier transform. The data in the middle of the K-space contribute contrast and in the edges resolution. The time of scan control by different factors like TR, matrix size and a number of excitation. The shorter the TR and smaller matrix size will lead to shorter scan time. The scan time is important in patients whose claustrophobia, cannot hold still for a long time and time management of MRI units (Westbrook et al., 2011).
Figure (5) K-space dimension (Westbrook et al., 2011)
Figure (6) show the K-space encoding (Westbrook et al., 2011)
Figure (7) show the gradient timing (Westbrook et al., 2011)
Until today there are no approve or data for excited of long-term biological side effect associated with exposure to MRI. Even though this is true, some safety steps should be considered because of other hazards in MRI environment. In 2001 a tragic accident happened when an oxygen tank killed a 6-year-old. After that accident, a government moved, and specialist team produce a guideline for safe MRI environment and practice (McRobbie, 2006; Westbrook et al., 2011).
The major concern in MRI practice is the magnetic field. All materials are divided into two group. The first group is materials that can enter MRI room called MR safe. MR safe materials mean materials will not possess any hazards in MR room. Second groups are MR unsafe materials define as any materials that may lead to hazard in MR room. The different sign should be posted around MR environment to different label area. The area outside the bore of MRI that possess a magnetic field called fringe field. The fringe field can become a dangerous area because of ferromagnetic metal can be projected at very high speed and cause death. Any patients should be given a questioner to list if have any metals or implant inside the body. After that, a metal detector is placed in the door to ensure there is no metal in the patient. For example, an aneurysm clip is not allowed in MR room because they are ferromagnetic (Westbrook et al., 2011).
Not only magnetic field that may cause a hazard to the patients but also the radiofrequency irradiation. The RF can cause heating of tissue which can be measured by a specific absorption rate (SAR). The unit of SAR is watts per kilogram. The SAR is used instead of temperature probe because temperature probe is not MR safe. To monitor the SAR level, patient weight is needed and pulse sequence parameter. Burns can occur because of RF antenna. Surface coils or ECG can produce an electrical current in the conductive loop. Caution should be taken not to allow conduction loop to occur. All wires should be electrically and thermally insulated. Cables from the coil that touch patient skin can lead to burn. Thermal injuries have happened before in MR scan. Some tattoos can lead to overheating and burns in some patients. The FDA announced that medication skin path could lead to burn in MR scan. Some clinic asks female patients to remove their makeup because some types contain a metal component that affects the quality of the image (Westbrook et al., 2011).
A biological effect can happen because of changing magnetic field. Some phenomena like peripheral nerve simulation and magneto- phosphors happened when voltage introduced within the conductor. The nerve, muscle and blood vessels work as a conductor in the human body. The effect of the phenomena in MRI can be explained in Faradays law of induction
= change in magnetic field
= change in time
= change in voltage
Any change of voltage, time or magnetic field will produce current in the conductor. Some patients will experience peripheral nerve simulation which means an involuntary muscle contraction. The involuntary muscle contraction is not good for image quality because of motion artefact. Other patients will experience a problem in their vision that may happen because of retinal phosphate affect by a change in the magnetic field. The experience of a problem with vision called magneto-phosphors. MRI machine is noisy when it operates, and loud noise can be noted by patients inside the MRI bore. Patients should wear earplugs to reduce the noise. The new MRI equipped with noise reduction hardware which adds to the cost of the machine (McRobbie, 2006; Westbrook et al., 2011).
All MR machines are cooled with liquid helium with a very low temperature that reaches 4 Kelvin. The helium can be used as a safety to stop the magnetic field if patiently is in danger. Quenching can be done manually but can damage the machine. If a fire occurs, a magnetic field must be quench before any firefighters enter the room because they carry oxygen tanks. The helium system is connected to outside environment to release the helium outside during quenching. The MR room must contain an oxygen monitor devise to not allow leaked helium to the room without notes to replace air because helium is lighter. The last safety measures that need to be taken if helium is leaked in MR room is the window in control room can be separated from the wall. The separation only happens when the pressure in the room change and the door cannot be opened. In this case, the window will separate from the wall and pressure can be equalized. This is important if a patient is inside the room and need to be evacuated (McRobbie, 2006; Westbrook et al., 2011).
Lab experiments started with the introduction of MR rules and regulation. Then a safety process was done to ensure no harm to anyone who will enter the MRI environment. A brief explanation of workstation of MRI and the steps need to be done in the experiments. After that an introduction of the main theory needed for the experiment. In this experiment, olive oil is used and water to demonstrate the effect of different TE, TR and bandwidth in image contrast. A coil was used around both components and ensure they are placed in the same place in all scans. The oil was used as a replacement for fat. we started with single spin echo using different TR and TE. Figure (8,9) show the effect of different TR and TE in image intensity. The intensity was calculated using MATLAB program. A program was run ten times and mean of intensity was calculated with standard deviation and error. All calculation is attached in the appendix.
Figure (8) single spin echo for oil
Figure (9) Single spin echo for water
As we can see when the TR and TE are short, there are differences in image intensity and contrast. This because there is not enough time between RF and second RF to allow full recovery of both oil and water to B. This will show the water as dark because water does not efficiency exchange energy with surrounding lattice. Water need longer time to go back to it longitudinal magnetization. Water produces a low signal in this type of setting. In another hand, oil has an efficient exchange with lattice and faster regain it longitudinal magnetization. Because of the faster, regain longitudinal magnetization oil produce a high signal and appear light in the image with this setting. When we used the short TR and TE this called T1 weighted image. In other image using longer TR, we can see the contrast are the same, and intensity from graph are close to each other. We allow both oil and water enough time to relax and gain their longitudinal magnetization. The longer the TR, the less different intensity between oil and water which cannot differentiate between them. When the TR is long the dominant imaging will be proton image density, and that can be seen in the figure (9) when the TR is long which will cancel the T1 weighted imaging. When we talk about longer TE which means T2 weighted images. In T2 weighted images, we see oil is dark, and water is light because the magnitude of transverse magnetization for oil is smaller than water. We can recognize it when the TE higher than 60 ms in the graphs.
In the second part of the experiment, we put TR 5000 ms and keep it constant with changing of TE. Multi-echo sequence was used. The intensity was calculated using MATLAB. Then national log of the intensity of both water and oil. Figure (10) show the intensity plotted in the graph.
Figure (10) show the Ln of intensity for both water and oil
By using the slop in the graph, we can calculate
T2. Table (1) show the value for
|water 5000 ms||-0.0002||5000|
|oil 5000 ms||-0.0079||126.5822785|
Table (1) show the T2 calculation
the last part of the experiment, we scan using different bandwidth to see the chemical shift. The figure (11) was created using ImageJ to show the chemical shift in oil. The different resonance frequencies between water and oil create the chemical shift. Fat has slower resonance frequency than water which makes the chemical shift in frequency encode direction. The phenomena happen because of the electron cloud is protection or creating a shield around the nucleus from the magnetic field. Every tissue has a different chemical composition which has different Larmor frequency.
Figure (11) the chemical shift in oil
the chemical shift is related to bandwidth. When the bandwidth is increased the chemical shift decreased. On the other hand, when the bandwidth is smaller, the chemical shift will increase. Furthermore, the chemical shift increases with the increase of magnetic field. The chemical shift can create an artefact in the image because of miss registration
MRI is imaging machine that uses the proton to create an anatomical image by using the phenomena of renounce. The different TR and TE can help to create a different setting to image different part of the body which has different chemical compositions. The different of TR and TE timing will allow the contrast between different tissue. The chemical shift phenomena are happened because of different Larmor frequency in different tissue in the body. The chemical shift happens with different bandwidth and can affect the image. The chemical shift artefact can be solved by using fat suppress imaging.
Allisy-Roberts, P., & Williams, J. (2008). Chapter 1 – Radiation physics. In Farr’s Physics for Medical Imaging (Second Edition) (pp. 1–21). W.B. Saunders. https://doi.org/10.1016/B978-0-7020-2844-1.50005-3
Bushberg, J. T. (Ed.). (2012). The essential physics of medical imaging (3rd ed). Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins.
McRobbie, D. W. (2006). MRI from picture to proton. Cambridge, UK; New York: Cambridge University Press. Retrieved from http://public.eblib.com/choice/publicfullrecord.aspx?p=321131
Westbrook, C., Roth, C. K., & Talbot, J. (2011). MRI in practice (4th edition). Chichester, West Sussex: Wiley-Blackwell.
Young, I. R. (2004). Significant events in the development of MRI. Journal of Magnetic Resonance Imaging, 20(2), 183–186. https://doi.org/10.1002/jmri.20123