Disclaimer: The information contained within the Grand Rounds Archive is intended for use by doctors and other health care professionals. These documents were prepared by resident physicians for presentation and discussion at a conference held at Baylor College of Medicine in Houston, Texas. No guarantees are made with respect to accuracy or timeliness of this material. This material should not be used as a basis for treatment decisions, and is not a substitute for professional consultation and/or peer-reviewed medical literature. MRI: Physics And Utility In Head And Neck Surgery Purcell and Bloch discovered the phenomenon of nuclear magnetic resonance in 1946. It was soon recognized that the phenomenon could be used to delineate the electronic structure of molecules, but it was not until the 1970s that it was employed in imaging. The first diagnostic human magnetic resonance images were produced in 1977. MRI is very useful in the head and neck area, particularly for its ability to discriminate variations in soft tissue. Lesions of the tongue, the parotid, oropharynx, and cerebellar pontine angle are well visualized with MRI. However, it will not delineate bony erosion and a CT may be necessary for evaluation for bone adjacent to a lesion. A CT is also more advantageous in imaging the paranasal sinus, as any inflammation of mucosa or thickened secretions will appear as an area of high signal intensity on MRI. MIR can also be used to evaluate blood flow. Flow can be detected and mapped by MR and magnetic resonance angiography is, in many cases, replacing traditional angiography. MRI is based on the measurement and alteration and spin properties of hydrogen ions induced by the application of withdrawal of magnetic fields. In essence, a magnetic field is applied to induce spin alignment of protons in a given volume of tissue. The field is then varied and the responses of the protons are measured and reconstructed into images through multiple Fourier transforms. An understanding of the product of magnetic resonance imaging necessitates a comprehension of basic electromagnetism and particle physics that underlie the imaging process Electromagnetic radiation is an energy form that propagates through the space-time continuum at a velocity of 3 x 108 meters per second. Electromagnetic (EM) radiation comprises visible light as well as radio-, micro-, x- and gamma rays. EM radiation is composed of perpendicular sine waves of electric and magnetic fields, which is why the concepts of electricity and magnetism are virtually inextricable. It is on the plane perpendicular to these fields that the energy of the proton travels. The electron field, the electric field, the magnetic field, and the direction of the propagation of the wave are always mutually orthogonal. The energy of a given wave is the frequency of the sine wave multiplied by Planck constant H. The current is defined as the movement of the positive charge. It is a vector quantity having both magnitude and directionality. Whenever a charge is in motion, the magnetic field is induced perpendicular to the direction of the current. Amperes log is the magnitude of the field for a given current. The direction field is predicted by Fleming's right-hand rule. If the thumb of the right hand points along the current flow, then the curled fingers will give the orientation of the circular magnetic field lines of force. When current is run through a loop, the net effect is to create a magnetic field which runs parallel to the long axis of the loop. The right-hand rule can again be utilized to determine the relative directionality of the current in the magnetic field. This time, the thumb is the induced field, and the fingers curve in the direction of the current within the loop. By convention, the lines of force run from north to south outside the loop and south to north inside the loop. Multiple current loops placed in series create a solenoid. The induced magnetic field is n times the magnitude induced by a single loop where n is the number of turns in a solenoid. A much more powerful magnetic field may, therefore, be induced without changing the current if the configuration of the wire is changed to that of a multi-turn solenoid. It is for this reason that solenoids are used in magnetic resonance imaging. The solenoid is responsible for creating the magnetic environment that stimulates the nuclear magnetic resonance. Now I will move on the principles behind the behavior of protons in this environment. Particularly important concepts are those of spin and precession. Spin and precession give rise to the measured quantities of T1, T2, and proton density. These three parameters were given different emphasis or weights in the creation of the image depending on the type of tissue being imaged. For example, a T1-weighted image is one in which T-1 signals are primarily used in the calculations; however, proton density in T2 usually also figure into the algorithm, however, with less mathematical emphasis. Spin is an inherent property of subatomic particles. When the spinning particle is charged as a proton or an electron, a small magnetic field is induced. This is the spin magnetic movement of the particle. A proton spinning on its axis may have its spin magnetic movement oriented south to north. If the same proton were to have the opposite spin, the spin magnetic movement would be of the same magnitude within the anti-parallel direction. When placed within an externally derived magnetic field, the spin magnetic movement is most stable when it is aligned with the lines of force of magnetic field B. There are two possible configurations - one in which the poles of the spin magnetic movement in the external field line up north-north to south-south, and one in which the poles line up north-south/north-south. The north-south/north-south configuration is the lower energy of the two. The particles contain either the high or the low energy state. The energy difference between the two states is quantified as delta E. The energy that is required for a single proton to transition from low to high energy is small enough, but it is available as thermal energy at room temperature. Protons at room temperature constantly flip from state to the other. The quantum of energy necessary to push the spinning charged particle from its low energy spin state to its high energy spin state comes from a photon of light at the appropriate frequency. The correct frequency can be determined for any given particle if one knows the energy difference between the two states, delta E. E= H nu where H is again Planck constant and nu is the frequency of the sine wave of the electromagnetic energy. Similarly, if the particle transitions from the high energy state to the low energy state it will give off a proton of the same frequency. The electromagnetism is radiated into the immediate environment which, for MRI purposes, is referred to as the lattice. All known particles have mass. The mass of a proton is 1.6 x 10-27 kilogram. A spinning proton, therefore, has a movement of inertia, the effect of which is to resist change of the particle's angular momentum. Macroscopically, this phenomenon is illustrated by a gyroscope or by a spinning top. If the top, for example, spins rapidly enough, the force of gravity is unable to overcome the movement of inertia and the top remains upright. As friction takes its toll, however, the top spins more and more slowly. Its angular momentum decreases, its angle of inertia decreases and eventually gravity overcomes it and the top falls. The top demonstrates another very important concept when applied to magnetic resonance. The top is a spinning mass being acted upon by a constant external force, gravity. That external force causes the top to tip off its original axis. Now the top is still spinning around its axis, but the axis itself is at an angle to the original orientation and the axis rotates around the vector of its original orientation. This is called the precession. The top is a spinning mass being acted upon by a constant external force. This is what causes it to precess. Now, if we substitute a magnetic field for the force of gravity, and our charged spinning particle substitutes for the spinning top, we can see how a proton in an external field behaves similarly. The proton spins on its axis and the axis precesses around the vector of the applied magnetic field. This state is a steady state of protons within the MR unit before another radiofrequency pulse is applied. The magnetic movement and the movement of inertia for a given particle are inherent properties of the particle and are combined into one property called the gyromagnetic ratio. The gyromagnetic ratio basically quantifies the intensity of the particle to resist change in its orientation within the magnetic field. The energy required to change a particle's energy state is dependent on the strength of the applied field B, and a gyromagnetic ratio, gamma, of the affected particle. Correlating this energy equation, with E= H nu, allows us to calculate the inherent frequency at which a given particle will precess in a static magnetic field. This frequency is called the Larmor frequency and is inherent to a particle in a static field. For a proton in a 1.5 Tesla magnetic field, the Larmor frequency is 63 MHz. This corresponds to a radiofrequency pulse which is required to excite the proton and a pulse of the same frequency in the radio band would, therefore, be emitted by a proton when it goes from the high energy to the low energy state. All the concepts that we have covered so far apply to all atoms in the periodic table. Protons are very convenient as examples because they are incredibly simple particles, but their simplicity also makes them useful in magnetic resonance imaging. Larger atoms consist of a positively charged nucleus surrounded by a variable cloud of electrons. The net magnetic movement for an atom with protons and electrons depends not only on the charged state of the atom, but also on the orbital configuration of the electrons. This presents a variable orbital magnetic movement and, therefore, a variable gyro magnetic ratio and Larmor frequency. Atoms with more variability are less likely to be in a coherent state unless a very strong magnetic field is applied. Protons have the advantage of relative simplicity, a small movement of inertia and a natural abundance in tissue. These properties make protons the ideal particles for magnetic resonance. The radiofrequency used in magnetic resonance imaging is a relatively low frequency and, therefore, carries lower energy. Visible light is slightly higher energy and x-ray and gamma rays are much higher than that. The x-radiation used in CT and traditional x-ray films are relatively high in energy and, therefore, can impart more damage to tissues that they encounter. MRI units operate within a field of 0.5-3.5 Tesla. As a frame of reference, refrigerator magnets have a slightly weaker magnetic field at around 0.1 Tesla. The earth's magnetic field is much smaller and averages 0.0001 or 10-4 Tesla. Now that we have achieved a steady state, we are finally ready to create a signal. The steady state can be represented by a proton spinning on its axis and precessing around the Z axis, which is defined as the direction of the original magnetic field. An electromagnetic pulse is added to the steady state in the X-Y plane. This energy induces change in the configuration of the magnetic movement and moves it away from the more stable configuration and precesses, therefore, at a larger angle on the Z axis. Note that the magnetic movement is now larger in the X-Y plane and smaller in the direction of Z versus steady state. Remember that energy has been added to the system, and the proton is now in a less stable and higher energy state. The two most commonly used pulses or radiofrequencies with the appropriate magnitude and orientation and displaced magnetic movement is 90 degrees or 180 degrees. These pulses are, therefore, termed 90 and 180 degree pulses. In the 90 degree pulse, the magnetic movement is displaced to the X-Y plane by the radiofrequency pulse when the pulse and the magnetic movement returns to its low energy, the Z-oriented configuration. In so doing, it emits the radiofrequency signal. This is called spin relaxation. The 180 degree pulse is similar in concept; however, the pulse simply deviates the vector of the axis by a full 180 degrees. The time between excitatory pulses is called the repetition time as represented by TR. The time between the excitatory pulse and resultant signal is called the echo time and is represented as TE. Another pulse can be applied at any time after the first pulse. These are fundamental parameter that can be varied to create different pulse sequences for highlighting various types of tissues. We have seen that when stimulating pulse is turned off, the magnetic vector returns to its equilibrium configuration. As it does, it gives off a signal which is readable by a radiofrequency receiver in the MR unit. The simplest parameter that the MRI unit can read and quantify is the proton spin density. This corresponds to the number of protons in the given volume of tissue that give off the appropriate signal. An image can be formed with this most basic of signals. Areas of higher densities of free protons are areas that resemble water in their chemical make up. Notice the CSF and the orbits weakly light up in the proton density mode. The other common types of signal, namely T1 and T2, are derived from breaking the same signal down into component functions. The return of the particle to its baseline configuration is composed of two separate functions. The return of the Z vector from 0 (zero) to its equilibrium state, and also the loss of the X-Y component induced by the added pulse. The return of the Z component of the magnetic field is called spin-lattice relaxation, because it emits energy into the surrounding medium or into the lattice. Spin-lattice relaxation is also known as T1. It is also called T1 recovery. The second function is the decay of the X-Y component in the induced magnetic field. It is due to the dispersion of precessional frequencies along the population of protons. A 90 degree pulse is applied, magnetic movement is precessing in the X-Y plane and the signal and the X-Y component disappears while the Z component is reappearing (which correlates again to the T1 aspect). This is called spin-spin relaxation. It is also known as T2 or T2 decay. T1 recovery and T2 decay occur essentially independently and so are measured separately; however, the Z component of the magnetic field cannot fully recover until the X-Y component is no longer displacing it downward. For this reason, T1 is always longer than or equal to T2. The protons in various types of tissue experience different environments due to the types of the bonds in which they are involved, the charges they bring atoms and the nature of the molecules to which they are attached. Protons in different environments will have different recovery times due to all of these factors and, therefore, T1 and T2 will differ. T1 for proton in pure water is 3 seconds. It is half that in CSF, lower still for muscle, and even lower than that for fat. This explains why these tissues all have different appearances on MRI. Various tumor types also have characteristic MR signals. While it is presently not feasible for a computer to convert these characteristic appearances into differential diagnoses, an experienced radiologist can use this information to make an educated guess as to the type of tumor revealed. Notice that a fatty tumor is the brightest or highest intensity object on an image on T1, and on T2 it is significantly darker. CSF, on the other hand, resembles water and is dark on T1 and bright on T2. Brain and muscle have intermediate properties and, therefore, have intermediate signal intensities. Bone does not have a readable signal due to the type finding of proton in solid bone. It shows black on both T1 and T2. Vessels with blood flowing in them appear as a complete signal void. This is because the blood that is stimulated by the radiofrequency pulse then has moved out of the picture frame by the time the signal is captured. These are a couple of tricks for determining whether images more heavily weighted on T1 and T2. T1-weighted images show water as low intensity and dark and fat is high intensity. T2 weighting reverses this. Water is high intensity and fat is low intensity. The repetition time is different and tends to be longer in T2 weighted series, so comparing the TR for images can give a clue as to the weighting.. There are several variables that can be manipulated to change the relative intensities of various tissues. The magnetic field gradient allows for differentiation in three dimensions and helps to sharply delineate borders between tissue types. There are various protocols that manipulate these parameters and specific time sequences. The radiologist knows in advance what type of lesion he or she is looking for. Often a sequence that maximizes visualization of that particular tissue type can be chosen. Fast spine echo, or simply spin echo, is a mode which uses 90 degree and 180 degree radiofrequency pulses, one on top of each other. It does not wait for the protons to re-equilibrate. Rather, it changes their angles with respect to whatever configuration they have attained during the shorter TR. This is performed multiple times in one pulse and creates a series of signals which are called echoes. Fast spin echo significantly cuts down on scan time. It has tissue resolution comparable to a T2 image. For this reason, fast spin echo is rapidly coming in to wider use. One drawback to fast spin echo is that unlike in regular T2 weighted images, fat remains very bright. This can make it difficult to differentiate the borders of lesions that light up on T2. For this reason, the fat suppression mode was created. Fat suppression is performed by the computer which performs a Fourier transform on the signal and attenuates the mathematical function associated with fat signals. This leaves other tissue types in relative relief. Contrast media are used in MRI for their ability to delineate breakdown of normal tissue barriers such as the blood brain barrier. The salient feature of MRI contrast agents such as gadolinium is that they are paramagnetic. This means a weak magnetic field can be temporarily introduced into a material that is normally nonmagnetic. This brings us to magnetic resonance angiography or MRA. MRA images of vessels are based on the velocity of blood flow. Basically a gradient of magnetic fields is used so the blood will enter the field between the pulse and the response and so give a signal. The signal strength is different depending on where in the gradient the blood originated. Therefore, it is dependent on the velocity of the flow. Two big advantages to MRA over conventional angiography are that contrast is not necessary and there is no ionizing radiation involved. There are several contraindications to the performance of MRI. Electromagnetic force can cause serious malfunction to cardiac pacemakers and cochlear implants. Some older vascular clips and ossicular prostheses contain paramagnetic nails and unless the clips or prostheses are known to be MR stable, another modality should be used. Shrapnel may be dislodged by the magnet and thereby cause tissue damage. This applies to metal shards that may lodge in the eye of a welder or metal grinder. If shards are suspected, the patient should be cleared by an ophthalmologist prior to MR imaging. Orthopedic prostheses are not contraindications to MR. They are made of non-paramagnetic material such as titanium. They may, however, induce artifacts. MR is not contraindicated in pregnancy since there is no ionizing radiation. Finally, claustrophobia is a relative contraindication. However, with the advent of the open MRI, this is becoming less of a problem. MRI is currently being investigated for use in the operating room. This method approaches real-time imaging, updating anatomy as surgery proceeds. While the present applications of intraoperative MRI are primarily neurosurgical, descriptions of such varied procedures as cholecystectomy, breast biopsy and endoscopic sinus surgery have been published, primarily out of Boston. A small unit with a low Tesla magnet can fit at the head of an operating table. Images can be created during the procedure and when not in use, the unit can be retracted underneath the table. Compared with the traditional fluoroscopy, which also approximates real time, MR is more accurate and does not expose the surgeon or the patient to ionizing radiation and is three-dimensional. Special considerations include the usual contraindications for MR. There is also need for shielding of the room's magnetic field with the use of non-paramagnetic instruments and anesthesia equipment. Most items can be made of titanium, which is non-paramagnetic. Case Presentation A.S. is a 3-month-old girl, otherwise healthy, who presented to the Texas Children's Hospital Emergency Room with a left-sided post-auricular mass. The patient's mother reported that the mass had suddenly appeared 10 days prior, and had not changed size in that interval. As the mass was not resolving, she decided to have her child evaluated. The mother denied any history of fussiness, fever, ear-pulling, upper respiratory infection or otitis, or ear drainage during the time surrounding the appearance of the mass or since. The child had no known medical problems, was the product of a normal spontaneous vaginal delivery and had had no need to be hospitalized since the time of her birth. On physical examination, the child was well developed and well nourished. She was in no apparent distress, and took a bottle readily and without difficulty. Her length and weight were age-appropriate. She had evidence of fluid behind her left tympanic membrane on otoscopy, and there was a 3cm mass just behind the left auricle. The mass was soft, non-fluctuant, non-tender, non-erythematous and the overlying skin appeared normal. CT with contrast of the brain with temporal bone cuts revealed a large, non-enhancing postauricular mass with a necrotic-appearing center. The outer table of the mastoid was eroded, and the mass appeared to be contiguous with the fluid-filled middle ear via a passage though the mastoid. A vague area of enhancement in the occipital region, which appeared to follow the Sagittal Sinus was noted by the radiologist and the differential diagnosis of this was given as thrombosis versus artifact from the adjacent bone. The child was taken to the operating room, where she underwent drainage of a left subperiosteal abscess and a simple mastoidectomy. The inner bony table of the mastoid was intact. Post-operatively, MRI of the brain was performed to rule out intracranial abscess or thrombosis based on the prior CT. MRI revealed good blood flow throughout, as evidenced by flow void, and no signs of inflammation or abscess intracranially. The patient did well postoperatively and was discharged home a few days later in good condition. Bibliography Bigelow DC, Eisen MD, Smith PG, Yousem DM, Levine RS, Jackler RK, et al. Lipomas of the internal auditory canal and cerebellopontine angle. Laryngoscope 1998;108:1459-1468. Carlton RR, Adler AM. Principles of Radiographic Imaging Albany, NY: Delmar; 1996. Curtin HD, Hirsch WL Jr. Imaging of acoustic neuromas. Otolaryngol Clin North Am 1992;25:553-605. Dunniway HM, Welling DB. Intracranial tumors mimicking benign paroxysmal positional vertigo. Otolaryngol Head Neck Surg 1998;118:429-436. Fried MP, Hsu L, Topulos GP, Jolesz FA. Image-guided surgery in a new magnetic resonance suite: Preclinical considerations. Laryngoscope 1996;106:411-417. Fried MP, Topulos G, Hsu L, Jalahej H, Gopal H, Lauretano A, et al. Endoscopic sinus surgery with magnetic resonance imaging guidance: Initial patient experience. Otolaryngol Head Neck Surg 1998;119:374-380. Hornak JP. The Basics or MRI. http://www.cis.rit.edu/htbooks/mri/bmri.htm. Lalwani AK. Meningiomas, epidermoids, and other nonacoustic tumors of the cerebellopontine angle. Otolaryngol Clin North Am 1992;25:707-728. McElveen JT, Saunders JE, Meisler WJ, Grist TM, Orest BB. Magnetic resonance angiography: technique and skull-base applications. Am J Otol 1991;12:323-327. Mitchell MR. Magnetic Resonance Imaging & Computerized Tomography: Referring Physician's Guid. Simi Valle, CA: Optimized MR Services, Inc.; 1991. Schwartz RB, Hsu L, Wong TZ, Kacher DF, Zamani AA, Black PM, et al.. Intraoperative MR imaging guidance for intracranial neurosurgery: Experience with the first 200 cases. Radiology 1999;211:477-488. Smith RC, Lange RC. Understanding Magnetic Resonance Imaging. Boca Raton: CRC Press; 1998. Som, Curtin. Head and Neck Imaging. St. Louis: Mosby; 1996. Sunders SR. MRI: A Conceptual Overview. New York: Springer-Verlag; 1997. Valvassori GE, Mafee MF, Carter BL. Imaging of the Head and Neck. New York: Thieme Verlag; 1995. Vogel TJ, Balzer J, Mack M, Steger W Differential Diagnosis in Head and Neck Imaging. New York: Thieme; 1999. Westbrook C, Kaut C. MRI in Practice. Oxford: Blackwell Science; 1998. Grand Rounds Archive | Department Home page BCM Public | BCM Intranet | Privacy Notices | Contact BCM | BCM Site Map | ©2001-2006 Baylor College of Medicine
|