Magnetic Field Strength Issues in Magnetic Resonance Imaging (MRI)
The purpose of this report is to examine the clinical application of magnetic resonance imaging (MRI) with respect to the issue of magnetic field strength (MFS). The report reviews briefly the basic physics, types and features of MRI units, examines the distribution of MRIs according to MFS, discusses both technical and practical issues related to MFS (safety, patient throughput, costs and clinical performance) and provides a summary of the current status of studies on MFS in MRI.
Magnetic resonance imaging (MRI) was first applied clinically in 1978 and has since been adopted to varying degrees in many countries. The basis for producing an image in clinical MRI applications is the detection of radiofrequency signals that are emitted from nuclei (most commonly hydrogen) exposed to an external magnetic field.
The magnets used in MRI can be of various types and magnetic fields strengths (MFS). MFS, which is the focus of this report, is measured in units called Tesla (T). For clinical applications, MRI units range in MFS from 0.064 T to 2.0 T. Magnets can be of two basic types. Electromagnets have fields generated by circulating currents induced on conductors that are either resistive or superconducting. Resistive electromagnets range in MFS from 0.2 T to 0.4 T and superconducting electromagnets are the only magnets that can practically exceed 0.3 T, but require coolants to maintain low operating temperatures. The second type of magnet is a permanent magnet, which does not require electricity to generate its magnetic field and is generally limited to field strengths below 0.3T.
Distribution of MRIs by MFS
Data on the distribution of MRI units by MFS indicate that, overall, the majority of MRI units (approximately 70%) are either 0.5 T or 1.5 T out of the total number of units installed (MFS less than or equal to 1.5 T), for the countries selected for comparison in this report (Japan, USA, Germany, Spain, France, Italy, United Kingdom, Scandinavia, Austria, Switzerland, Australia, Belgium). When the number of units in Canada (MFS less than or equal to 1.5 T), are compared to other countries, it is evident that Canada has, by far, the largest proportion of 1.5 T MRI units (80%). This may be, in part, because the distribution of MRI units in Canada is limited by the availability of government funding and the majority of MRI units in Canada are located in university hospital settings for both clinical and research work.
In general, since the introduction of MRI, the trend has been from the lower MFS to the mid fields units (less than 0.5 T) and then towards the higher field units (1.5 T) in the late 1980s. More recently, the trend has shifted back to the low and mid field systems as the systems become more widely distributed and the technology improves. To date, Canada illustrates the exception to this generalization, with only 20% of the total number of installed units at a field strength of less than 1.5 T. Based on the number of units on order, it is unlikely that this distribution amongst MFS in Canada will be any different over the next year or two.
One of the factors in MRI upon which there tends to be a focus is the magnetic field strength, partly because of the perception that higher MFS produces better images and the potential for magnetic resonance spectroscopy at field strengths greater than 1.5 T (although magnetic resonance spectroscopy has not yet been clinically proven).
The relationship amongst the variables contributing to image quality is complex, and extends far beyond MIFFS. Image quality will be optimized at a given field strength by adjusting a set of parameters relative to one another, and not by any single variable alone. There is no field strength that is optimal in all situations.
As an imaging modality, MRI offers the benefit of improved tissue contrast, the absence of ionizing radiation and does not generally require contrast agents. Nonetheless, there are some potential hazards associated with MR imaging, and these may vary with MFS. Ferromagnetic objects are potentially dangerous in the presence of MRI. Loose objects may behave as projectiles, metallic implants may be dislocated and intraocular ferromagnetic fragments may move and cause damage to vision. In addition, 1 - 4% of patients will experience claustrophobia from MRI scanning and non-ferromagnetic equipment is required for patients who need respirators. In genreal, the lower magnetic field strength units are safer than higher fields units, although when properly operated, the actual risk to patients is minimal at any field strength.
The time required to obtain an image using MRI is dependent on a number of factors, including pulse repetition time, the number of acquisitions per image and the number of lines in the field of view. As MFS increases, a comparable image can be produced using a smaller number of acquisitions (increasing patient throughput), the pulse repetition time becomes longer (decreasing patient throughput) and the lines in the field of view remain fixed for a given field of view and spatial resolution. Overall, it is not absolutely clear as to whether there will be a net gain in patient throughput as MFS increases because these parameters impact on patient throughput in opposite directions. The overall impact of MFS on patient throughput is complex, and in addition to technical factors, depends on the skill of the staff in operating the system. Studies vary in their conclusions about how MFS affects patient throughput from a modest advantage of 1.5 T over 0.5 T, to no major effect, to an advantage of 0.5 T over 1.5 T.
Costs considered in this report include typical capital costs, costs of siting, operating costs and total annual costs. The intent is to illustrate the relative costs of the units at different MFS. In general, costs of all types tend to increase with MFS. The capital cost of a 1.5 T unit ($2,690,000 - $2,940,000) is about three times that for a 0.064 T unit ($ 925,000 - $935,000). Similarly, siting costs, as well as the space required is significantly greater for a 1.5 T unit ($100,000 - $230,000; 65m2) compared to a unit less than 0.5 T ($45,000 - $110,000; 33m2). Finally, total annual cots were estimated including labor, supplies, acquisition and siting depreciation, service costs, cryogens and power are $1,236,000 - $1,603,000 for a 1.5 T unit, nearly double that for a unit than 0.5 T at $826,000 - $866,000. These figures are meant to be estimates only. In general, costs for an MRI increase with MFS. Nonetheless, costs may vary considerably depending on the specific technology and operational choices that are made.
There have been few rigorous studies done comparing the diagnostic accuracy or impact on patient outcome of MRI to other imaging modalities. Nonetheless, given the tremendous capability of MRI, it is probably reasonable to conclude that MRI is better than other imaging modalities in virtually all body parts, with the possible exception of the abdomen.
When comparing images produced at various MFS, the images produced at higher field strength tend to be considered better on the basis of subjective evaluations by radiologists. This does not mean, however, that these images are necessarily more diagnostically accurate or have any impact on patient management or outcome. There are few studies comparing MRI at different MFS on the basis of these parameters, and it depends somewhat on the body part being imaged. Various studies have recommended higher field strength units for head and spinal imaging. Limited evidence suggests that mid field systems perform equally as well or better than high field systems in body imaging. The higher field units have previously had an advantage in imaging the musculoskeletal system and extremities by achieving thinner slices, but with improving technology, the low field strength units are also considered to be excellent in the application.
Current Status of Studies on MRI Field Strength
There is a paucity of evidence in the literature about the relative merits of any particular MFS over another in MR imaging, and the studies that have been done tend to focus on the image quality rather than the diagnostic accuracy or impact on outcome at various MFS. There are at least two studies about these issues that are currently in progress.
The first is a review paper comparing low field (0.064 T) and high field MRI (1.5 T) in imaging various body parts. Based on preliminary results, the authors, Orrison and Gronemyer (1992), have concluded that: (1) there is no diagnostically significant difference between low field and high field MR imaging although high field images offer better spatial resolution, and, (2) the 0.64 T unit is a satisfactory imaging modality with particular benefits with respect to safety, patient access and cost, although more time is generally required per scan.
The second study is a randomized controlled trail of MR imaging at 0.5 T and 1.5 T being undertaken in London, Ontario by Vellet (personal communication) and collaborators to evaluate patients undergoing MRI of various systems including the central nervous system, the musculoskeletal system, the cardiothoracic and hepatobiliary system. This will be the first randomized field strength study completed using comparable MR technology. Preliminary results from the study support the hypotheses that MRI is superior to other diagnostic modalities and that there is no diagnostic difference between 0.5 T and 1.5 T systems.