In 1993, Paul Lauterbur and Peter Mansfield were awarded the Nobel prize for their discovery that gradients in magnetic fields could be used to create images of bodily structures without harmful exposure to radiation.1 Their work laid the foundation for the development of magnetic resonance imaging (MRI). Since their seminal work in the early 1970s, over 60 million MRI investigations have been performed using approximately 22,000 MRI scanners worldwide.1 MRI has gained widespread use in modern medicine, particularly in the neurosciences. It is not surprising, therefore, that such technology would eventually be applied to neurosurgical operations. What follows is an overview of the current applications of intra-operative MRI (iMRI), the surgical and engineering considerations surrounding its use, and a brief discussion of the future applications for this bold new technology.

Current Applications of Intra-operative MRI

iMRI was initially described in 19932 and consists of small-scale MRI units built in MRI-compatible operative suites. The utility of iMRI lies in its ability to obtain near-realtime brain imaging during neurosurgical procedures. This is valuable for several reasons. First, the brain is susceptible to structural deformations during neurosurgical procedures. Deformations become more pronounced over the course of the operation due to brain retraction and dissection. Deformations distort anatomical landmarks and may mask the location of brain lesions. Near-realtime imaging provides the surgeon with enhanced anatomical orientation. Second, brain lesions such as tumors are frequently infiltrative and resemble the surrounding normal brain tissue. This increases the complexity of the primary goal of surgery—to maximize tumor removal while minimizing injury to normal brain. Intra-operative imaging may differentiate residual tumor from normal brain, thereby optimizing surgical resection and minimizing the morbidity associated with surgery.

Currently, iMRI is only available at a few centers. Before iMRI reaches widespread use, several conflicting considerations must be reconciled. These considerations balance the surgeon’s desire to integrate an unobtrusive iMRI with the engineer’s desire to integrate a scanner that produces high-quality images suitable for surgical use.

A Surgeon's Perspective

To be a useful adjunct to neurosurgical procedures,iMRI must first abide by the Hippocratic doctrine—primum non nocere (“first, do no harm”). Hence, the iMRI must not interfere with the neurosurgical procedure. Current limitations of iMRI are increased operative time1 and decreased surgical access to the patient.2

Neurosurgical procedures are lengthy. Arresting the flow of an operation to obtain imaging can significantly extend the operative time. First, time is required to establish magnetic field homogeneity (i.e. shimming) and to adjust for radiofrequency (RF) effects of the local environment (i.e. tuning).The acquisition of MR images then requires various durations depending on the image sequence and the volume of interest. Beyond image acquisition, many intra-operative scanners require positioning of the scanner, or patient, within the magnetic field. Fixedposition scanners avoid this additional time, but force the surgeon to operate within the physical limits of the magnet. Sharing this confined space of approximately 50cm are the surgical assistant, operative microscope, and surgical equipment. This also constitutes a significant physical constraint on proper head position and surgical approach. Of course, all of the above must exist under sterile conditions.These considerations serve as potential obstacles to the widespread introduction of iMRI into the operating theater.

An Engineer’s Perspective

There are numerous technical considerations that center on the engineering and design of iMRI to produce high-quality images. Image quality is directly proportional to magnetic field strength, signal to noise ratio, gradient strength, time of image capture, field of view, and RF coil features (i.e. size, geometry, quality). Magnet strength, measured in Tesla (T; units of magnetic flux density), is at the top of the list of technical considerations. While conventional clinical MRI units range from 1.5 to 3T, iMRI units are more commonly 0.3–1T. The design of the magnet itself is also critical. Several third-generation magnet designs are currently in production, including ‘double donut’ designs, suspended bore magnets and flat-bed designs. Double donut designs may be mobile or fixed. The most recent mobile version rolls up to the operative table. Once in position, it retracts under the table until needed. Suspended magnets, meanwhile, travel on ceiling tracks to a docking position at the head of the table. Flat-bed designs require a sliding operative table to move the patient into the magnet when imaging is necessary. Additionally, iMRI requires shielding from external RF interference. This mandates numerous adaptations to the operating suite. Routine neurosurgical instruments, such as electrocautery devices and electrophysiological monitoring equipment, are potential sources of RF interference. Finally, the operative suite itself must be shielded with leadlined walls.