Rapid Essay Researchers We offer fast essays with the best content
Nuclear Magnetic Resonance (NMR) was first observed by Isidor Isaac Rabi in 1937 when he bathed a beam of lithium molecules in radio frequency energy and varied their magnetic environment. He demonstrated that low-energy nuclei would absorb energy at a resonant frequency and that this energy was dependant upon the strength of the magnetic environment, B0 (Fig. 1) (Rabi, 1937). This represents the underlying physical principle upon which Magnetic Resonance Imaging (MRI) is based.
The relationship between resonant frequency and magnetic field strength had been described by Larmor in classical mechanics and is defined by the Larmor Equation (Fig. 2) (Pooley, 2005).
The gyromagnetic ratio is a constant specific to different types of nuclei and so the frequency of precession is directly proportional to B0 (Pooley, 2005).
In equilibrium, the statistical distribution between the two energy states favours the lower energy state and it is this difference that creates a net magnetisation that can be measured and forms the magnetic resonance (MR) signal that contributes to the image (McRobbie, Moore, Graves and Prince, 2007).
An early pioneer was Felix Bloch who introduced the T1 and T2 time constants in his mathematical description of NMR (Elster and Burdette, 2001):
dMz/dt = (M0-Mz)T1
dMx,y/dt = Mx,y/T2
The equations predict that following a 900 RF pulse, the net nuclear magnetisation, Mz will gradually spiral back to its equilibrium to align with B0 as shown in Figure 3. This is termed T1-relaxation and is defined as the time for the longitudinal magnetisation to reach 63% of its final value. T2-relaxation is the time taken for the transverse component of the magnetisation, Mx,y to decay to 37% of is final value through de-phasing due to local magnetic variation. (Elster and Burdette, 2001) The two processes occur simultaneously and the times are tissue-specific which allows tissues to be discriminated based upon measurements of the magnetic vector or MR signal (Pooley, 2005). This was first recognised by Raymond Damadian (Fig. 4) who observed differences in the T1-relaxation time of tumours relative to healthy tissue and it prompted him to suggest using NMR to differentiate between cancerous and normal tissue in human beings (Damadian, 1971).
Paul Lauterbur suggested the in vivo study of malignant tumours following the successful imaging of two capillary tubes by using magnetic gradients in conjunction with a back-projection method (Lauterbur, 1973) (Fig. 5).
An extension of the use of gradients was in their application to selectively excite a slice of a larger volume, taking advantage of the principle that nuclei exposed to a magnetic gradient will precess (and resonate) at frequencies dictated by their relative position along that gradient (Fig. 6) (McRobbie et al., 2007).
The spatial encoding of the MR signal through linear magnetic gradients is essential to MR imaging (McRobbie et al., 2007). Uncontrolled variation in the linearity of these gradients can result in distortion and artefact because unknown variations in the gradient strength will alter the precessional frequency of nuclei in an unpredictable manner and result in inconsistent excitation and inaccurate localisation of signal (McRobbie et al., 2007).
As technology improves, so do the achievable gradient amplitudes allowing the generation of smaller field-of-views and thinner slices, which has implications for both image resolution and the detection of more subtle pathology (McRobbie et al., 2007). Increasing amplitudes and slew rates also facilitate some specialised MR techniques such as diffusion-weighted imaging that has proven useful in imaging stroke and malignant disease (McRobbie et al., 2007). The biological effects associated with time-varying magnetic fields relate primarily to peripheral nerve stimulation (PNS) where at low frequencies, induced currents can stimulate nerve and muscle cells (Price, 1999). Uncomfortable and distressing, such stimulation can result in limb movement or even ventricular fibrillation although most PNS can be avoided by keeping the rate of change of the magnetic flux density to less than 20T/s (Medicines and Healthcare products Regulatory Agency, 2007). A secondary effect of the use of gradients is the production of acoustic noise that is sufficiently loud to damage hearing of patients (Westbrook, Roth and Talbot, 2005).
Gradient technology was eventually refined into processes known as phase and frequency-encoding to spatially locate signal within a slice (Kumar, Welti and Ernst, 1975). Gradients are used to change the phase of precessing nuclei in one dimension and the origin of signal in the second is distinguished based on frequency. This process is repeated to fill a matrix known as k-space with raw data that is then Fourier transformed into the final image (Fig. 7) (McRobbie et al., 2007). While each point in k-space contains information relating to the whole of the image, contrast information is primarily central and edge detail is distributed peripherally (McRobbie et al., 2007). It should be noted that scan time is defined as the time to fill k-space and is dependent on the number of phase-encoding steps (Westbrook et al., 2005). Given that the relatively long scan times involved in MRI promote “ghosting” artefact through movement in the phase-encoding direction (Fig. 8), reductions in scan time improve both image quality and the patient experience (Westbrook et al., 2005).
There are a number of principles that can be applied to k-space acquisition that have an impact on image quality and the acquisition time. Techniques such as the partial Fourier method which takes advantage of the symmetry of k-space to acquire just over half of the phase-encoding steps; duplicating the remaining information to maintain the FOV and image resolution at a small cost to the signal-to-noise ratio (SNR) (Yang, Roth, Ward, deJesus and Mitchell, 2010).
Phased array coils were developed to boost the SNR by virtue of overlapping elements that detect less noise, but they can be used with k-space to reduce acquisition time (Glockner, Hu, Stanley, Angelos and King, 2005). In parallel acquisition, the coil elements acquire a subset of k-space and combine them to form the final image offering a reduction in scan time proportional to the number of elements (Glockner et al., 2005). Unfortunately, parallel imaging suffers from a number of artefacts such as ghosting, aliasing and non-random noise distribution, but it remains an effective method of reducing scan time and can be applied to a wide variety of pulse sequences (Glockner et al., 2005). It has particular value in cardiac and vascular studies where the reduction in acquisition time is vital (Glockner et al., 2005). Combining it with non-rectilinear acquisitions of k-space such as spiral sampling, yields advantages in Contrast Enhanced – Magnetic Resonance Angiography (CE-MRA), by facilitating fast scans and acquiring contrast data early to coincide with the administration of intravenous contrast (Westbrook et al., 2005). CE-MRA is a useful technique in imaging blood vessels, offering a high SNR, minimal artefacts, and a relatively high spatial resolution (Glockner et al., 2005). It is an invasive procedure with inherent risks of infection, extravasation and a debilitating, occasionally fatal condition known as nephrogenic systemic fibrosis (NSF) that is derived from the contrast medium (Grobner, 2006).
Gadolinium is described as a paramagnetic contrast agent and works by promoting T1-relaxation through, in part, its strong magnetic moment (Weinmann, Brash and Wesby, 1984). Superparmagnetic contrast agents such as iron oxide promote T2-relaxation and are also used to distinguish magnetically similar but histologically dissimilar tissues (Westbrook et al., 2005). Gadodiamide (commercial name Omniscan) (Fig. 9) is considered a first generation contrast agent and possesses a large magnetic moment but suboptimal tumbling frequency; necessitating a relatively large dose compared to later generations (Juluru, 2009). While Broome et al. (2007) identified dose as a risk factor of NSF, Habibi et al. (2008) has since demonstrated the potential of reducing dose in peripheral MRA at magnetic field strengths of 3T without compromising diagnostic quality, attributing this to the higher SNR, increased speed, coverage and spatial resolution available with higher magnetic fields. Dagia and Ditchfield (2008) supported this by noting that an increase in the T1-relaxation times seen at 3T has a relatively small effect on paramagnetic contrast agents, exaggerating the difference between the contrast agent and unenhanced tissue.
There are a number of techniques that can image blood flow without running the risks described above. In spin echo sequences with pre-saturation pulses, rapidly flowing blood appears dark, discriminating vessels and potentially allowing the identification of any occlusion as a persistent high signal within the vessel (Westbrook et al., 2005). More sophisticated techniques such as Time-of-Flight MRA image blood flow by using rapid pulses of RF to prevent T1-relaxation of stationary tissue while allowing inflowing blood to give signal. The technique is most sensitive to flow perpendicular to the axis of the slice, with any parallel or turbulent flow potentially being saturated, producing an unreliable representation of the vessel. Another disadvantage is high background signal from fat, although this can be minimised by sampling the echo when fat and water are out of phase, taking advantage of a phenomenon known as chemical shift of the second kind.
Different molecular environments around protons produce local variations in the magnetic field so that they have different precessional frequencies. This is demonstrated through a chemically resolved spectrum (Fig. 10). Since fat and water precess at different frequencies, their signal can combine or subtract depending on whether they are in or out of phase respectively (McRobbie et al., 2007).
Chemical shift of the first kind or chemical misregistration is an imaging artefact where the fat and water signal of adjacent structures are mapped to distant pixels depending on the defined receive bandwidth and the field of view (FOV) used (Fig. 11) (McRobbie et al., 2007). An important point to note is that as the precessional frequency of the nuclei is proportional to the strength of the magnetic environment – as defined by Larmor’s Equation; the frequency difference between fat and water (and the severity of chemical misregistration) increases in direct proportion to B0 (Dietrich, Reiser and Schoenberg, 2008). The effect of chemical misregistration can be countered by increasing the receiver bandwidth or by using a fat-suppression sequence (Fig. 12) and given that as B0 increases, so does the frequency difference between fat and water, it becomes easier to selectively eliminate the signal from either one (Dagia and Ditchfield, 2008).
The MR signal increases proportionally with the strength of the main magnetic field because fewer nuclei have sufficient energy to oppose the field, providing a greater proportion of magnetic moments to manipulate (McRobbie et al., 2007). Many applications benefit from this increase in signal, while some are degraded by an associated increase in artefact, like chemical misregistration and magnetic susceptibility (Dagia and Ditchfield, 2008). Magnetic susceptibility is defined as the extent to which an object placed within a magnetic field becomes magnetised itself (Westbrook et al., 2005). It is significant within MRI because the alteration of the local magnetic environment will alter the precessional frequency of nuclei causing signal voids in the absence of resonance and image distortion where the uniformity of spatial-encoding gradients are compromised (Westbrook et al., 2005). While generally considered an undesirable effect, magnetic susceptibility can be advantageous and is exploited in cerebral perfusion imaging, functional imaging and in detecting blood products like hemosiderin (Bitar et al., 2006). As it is worse at higher field strengths, greater temporal resolution or signal strength (or a mixture of the two) to be achieved in functional imaging (Schmitz, Aschoff, Hoffman and Grön, 2005).
Modern MRI scanners typically use superconducting magnets because they offer very stable, high field strengths and good magnetic homogeneity (in conjunction with ferromagnetic blocks and shim coils). This is achieved through the phenomenon of superconductivity where the resistance within a superconductor falls to zero when it is cooled to a specific critical temperature (Fig. 13).
As these systems use a cryogen to keep the coils of wire at a low temperature there is a risk of it boiling off and venting into the scanning room, displacing oxygen and potentially causing asphyxiation (Westbrook et al., 2005).
.
High-strength superconducting magnets are invariably tube-types which may not accommodate patients who are claustrophobic or patients who cannot fit into the scanner due to body habitus or pathology (McRobbie et al., 2007). Open scanners (Fig. 14) effectively address this issue and allow interventional MRI with all the associated ferromagnetic risks. Upright scanners have recently been developed that can image structures such as the lumbar spine in various positions and under various stresses. While using an electromagnet to produce a magnetic field of just 0.6T, preliminary research suggests that this technology offers increased sensitivity to some pathology (Wei et al., 2007).
As B0 increases, so do concerns regarding possible bioeffects associated with powerful, static magnetic fields (Price, 2009). A known effect is the temporary elevation of the T-wave in ECG tracings due to an induced electric potential across blood vessels (Price, 2009), but no mechanism of injury has been proposed relating to this phemomenum and by far the most signifcant risk of harm remains the rotational and translational forces applied to ferromagnetic objects either within or external to a patient (Fig. 15) (Price, 2009). This is epitomised by the tragic death of Michael Colombini in 2001 (Chen, 2001).
By Larmor’s equation, as B0 increases, so does the excitation frequency with a commensurate reduction in wavelength, potentially causing B1 inhomogeneity through the superimposition of RF (Schmitz et al., 2005). Energy deposition can also be a problem since the specific absorption rate (W/kg) is proportional the square of the excitation frequency.
A coil design that has had an impact on image quality and patient safety is the quadrature coil (Fig. 16). Traditional linear coils discard half of the signal data because the Fourier Transform cannot differentiate between positive and negative frequencies. Quadrature coils generate two B1 fields with a 900 phase shift, exciting nuclei with twice the efficiency and resulting in less RF deposition. During the receive stage, the coil detects two independent, pase-shifted signals and combines them to improve the SNR by a factor of 1.4 (Westbrook et al., 2005).
While the SAR is an issue with patients with compromised thermo-regulatory functions, most RF-related injuries are due to the induction of emf in conducting loops (Price, 2009). While conductors such as coils are thermally insulated, inadvertant conducting loops may be formed by the patient. Ocassionally some tattoos may approximate one of these conducting loops with the associated heating effects due to the presence of conducting particles within the ink. (Fig. 17) (Wagle and Smith, 2000).
A method for reducing the RF-related heating is to select a pulse sequence that uses a limited amount of RF. Echo Planar Imaging (EPI) is a highly efficient method of image acquisition which involves one RF pulse followed by rapidly switching phase and frequency-encoding gradients to acquire the whole of k-space within one time-to-repetition, TR (Bitar et al., 2006). While these scans are sensitive to magnetic susceptibility artefact and the rapidly changing gradients produce substantial noise and can stimulate nerve and muscle tissue, the EPI provides superior tissue contrast to standard gradient-echo sequences. Magnetic susceptibility artefact caused by metallic implants can severly degrade a gradient-echo image because of its reliance on magnetic gradients. An effective pulse sequence that mitigates this artefact is the spin echo sequence (Fig. 18). Spins are rephased through the application of a 1800 RF pulse that ompensates for any T2* dephasing. Considered the “gold-standard” in MRI, spin echo sequences have the disadvantage of being relatively slow in acquisition (Bitar et al., 2006).
Fast Spin Echo sequences (Fig. 19) offer reduced magnetic susceptibility and a significant reduction in scan time through the application of multiple 1800 RF pulses to rephase spins and produce an echo in conjunction with multiple phase-encoding steps per TR. As each echo is generated at a different TE, the central lines of k-space are filled around an “effective TE” – contributing overwhelmingly to image contrast at the expense of reduced image resolution since low amplitude signal is used to fill peripheral k-space (Westbrook et al., 2005). Additional contrast issues associated with the repeated 1800 RF pulses include a reduction in the effect of spin-spin interactions in fat that results in high signal from fat on T2-weighted scans, and increased magnetization transfer effects that causes a small loss of signal from tissues like muscle (Westbrook et al., 2005).
Scientific discoveries and improvements in technology have seen MRI emerge as a cornerstone of medical imaging (Westbrook et al., 2005). Its development can be traced through Nobel laureate history and given its complexity, it is easy to imagine that that will continue to be the case.
.
References
Bitar, R., Leung, G., Perng, R., Tadros, S., Moody, A. R., Sarrazin, J., McGregor, C., Christakis, M., Symons, S., Nelson, A. & Roberts, T. P. (2006). MR pulse sequences: What every radiologist wants to know but is afraid to ask. Radiographics, 26(2), 513-537.
Broome, D. R., Girguis, M. S., Baron, P. W., Cottrell, A. C., Kjellin, I & Kirk, G. A. (2007). Gadodiamide-associated nephrogenic systemic fibrosis: why radiologists should be concerned. American Journal of Roentgenology, 188(2), 586-592.
Chen, D. W. (2001, August 01). Small town reels from boy’s M.R.I. death. The New York Times, p.5.
Dagia, C. & Ditchfield, M. (2008). 3T MRI in paediatrics: Challenges and clinical applications. European Journal of Radiology, 68(2)309-319.
Damadian, R. (1971). Tumour detection by nuclear magnetic resonance. Science, 171(976), 1151-1153.
Dietrich, O., Reiser, M. F. & Schoenberg, S. O. (2008). Artifacts in 3-T MRI: Physical background and reduction stratagies. European Journal of Radiology, 65(1)29-35.
Elster, A. D. & Burdette, J. H. (2001). Questions and answers in magnetic resonance imaging. (3rd ed.). Missouri: Mosby, Inc.
Glockner, J. F., Hu, H. H., Stanley, B. S., Angelos, L. & King, K. (2005). Parallel MR imaging: A user’s guide. Radiographics, 25(5),1279-1297.
Grobner, T. (2006). Gadolinium: a specific trigger for the development of nephrogenic fibrosing dermopathy and nephrogenic systemic fibrosis? Nephrology Dialysis Transplantation, 21(4), 1104-1108.
Habibi, R., Krishnam, M. S., Lohan, D. G., Barkhordarian, F., Jalili, M., Saleh, R. S., Ruehm, S. G. & Finn, J. P. (2008). High-spatial-resolution lower extremity MR angiography at 3.0 T: Contrast agent dose comparison study. Radiology, 248(2), 680-692.
Juluru, K., Vogel-Claussen, J., Macura, K. J., Kamel, I. R., Steever, A. & Bluemke, D. A. (2009). MR Imaging in patients at risk for developing nephrogenic systemic fibrosis: Protocols, practice, and imaging techniques to maximize patient safety. Radiographics, 29(1), 9-22.
Kumar, A., Welti, D. & Ernst, R. R. (1975). NMR Fourier zeugmatography. Journal of Magnetic Resonance, 18(1), 69-83.
Lauterbur, P. C. (1973). Image formation by induced local interactions: examples employing nuclear magnetic resonance. Nature, 242(3),190-191.
McRobbie, D. W., Moore, E. A., Graves, M. J. & Prince, M. R. (2007). Picture to proton. (2nd ed.). Cambridge: Cambridge University Press.
Medicines and Healthcare products Regulatory Agency. (DB/2007/03). Safety Guidelines for Magnetic Resonance Imaging Equipment in Clinical Use.
Pooley, R. A. (2005). AAPM/RSNA physics tutorial for residents: Fundamental physics of MR imaging. Radiographics, 25(4), 1087-1099.
Price, R. R. (1999). The AAPM/RSNA physics tutorial for residents: MR imaging safety considerations. Radiographics, 19(6), 1641-1651.
Rabi, I. I. (1937). Space Quantization in a Gyrating Magnetic Field. Physics Review, 51(4) 652-654.
Schmitz, B. L., Aschoff, A. J., Hoffmann, M. H. K. & Grön, G. (2005). Advantages and pitfalls in 3T MR brain imaging: A pictorial review. American Journal of Neuroradiology, 26(10), 22229-2237.
Wei, F., Hong, S., Zou, J., Tow, B., Miyazaki, M., Morishita, Y., Alanay, A., Abitbol, J. & Wang, J. (2007). The effect of lumbar flexion and extension on the central canal with dynamic MRI. The Spine Journal, 7(5), 38-39.
Weinmann, H. J., Brasch, R.C., Press, W.R. & Wesbey, G.E. (1984). Characteristics of gadolinium-DTPA complex: a potential NMR contrast agent. American Journal of Roentgenology, 142(3), 619 -624.
Westbrook, C., Roth, C. K. & Talbot, J. (2005). MRI in practice. (3rd ed.). Oxford: Blackwell Publishing Ltd.
Yang, R. K., Roth, C. G., Ward, R. J., deJesus, J. O. & Mitchell, D. G. (2010). Optimising abdominal MR imaging: Approaches to common problems. Radiographics, 30(1)185-199.
TO HAVE YOUR ASSIGNMENTS DONE AT A CHEAPER PRICE, PLACE THIS ORDER OR A SIMILAR ORDER WITH US NOW.