Schenectady, NY--With help from the image displayed on the liquid-crystal monitor, the surgeon gently advances the cryoprobe towards its target: a tumor lodged in the patient's prostate gland. The imaging system's computer calculates probe-tip coordinates as well as the probe's projected path. These values, superimposed on the monitor, continually refresh the MR image, giving the surgeon real-time, three-dimensional feedback.
Until now, magnetic resonance imaging (MRI) has been useful primarily for diagnostic purposes, involving images obtained before and after the interventional procedure. The ability to locate a tumor in the prostate gland, liver, or brain, and visually guide the needle or probe to that exact location, existed only on the surgeon's wish list.
The reason? Conventional superconducting MR imagers feature a series of cylindrical and symmetrical coils, housed in a tubular cryostat. Designed to create a highly uniform magnetic field over the entire imaging field of view, this configuration physically blocks a physician's access to the patient lying within the superconducting magnet bore.
"Magnetic Resonance Therapy" (MRT)--first described by Ferenc Jolesz, MD and director of image-guided therapy at Boston's Brigham and Women's Hospital (BWH)--moves MRI from the diagnostic arena into the operating room. The system features an "open" magnet configuration, enabling the surgeon to stand and work directly within the imager. With MRT, the physician not only pinpoints the location of diseased or damaged tissue, but monitors therapy as it is applied.
"What makes this technology so exciting," says Leonard Holman, BWH chairman of radiology, "is its potential to be used in concert with so many other existing technologies." Endoscopy and laparoscopy, in particular, lend themselves to three-dimensional image guidance. So do laser, cryosurgical, and ultrasound delivery systems. The goal, Holman states, is to replace traditional surgical procedures with minimally invasive, and even "trackless" interventions, where there is no body contact at all.
An open magnet. Several obstacles stood in the path of achieving such a goal when Jolesz proposed the idea of magnetic resonance therapy to Morry Blumenfeld, general manager of image-guided therapy programs at General Electric Medical Systems, Milwaukee, WI. The first challenge was the most obvious: design a magnet in which the physician can comfortably stand and work.
Placing the magnet coils in separate, but communicating cryostats proved to be the solution. This open configuration, resembling a pair of doughnuts on edge, generates a uniform magnetic field across a 30-cm-diameter, spherical imaging volume, yet allows a 56-cm gap between the innermost coils for access to the patient.
The open design's wide gap, reports Project Leader Kirby Vosburgh of General Electric Corporate Research and Development Center, Schenectady, NY, was possible due to a key invention allowing a new class of magnet design: Superconducting tape based on niobium-tin, rather than conventional niobium-titanium. Wrapped around the magnets, the new material becomes superconducting at 10K (-263C) instead of 4K (-269C).
|How MR imaging worksOne of the problems surgeons
experience when operating inside the body is tissue differentiation:
Diseased tissue looks the same as healthy tissue.
Because magnetic resonance imaging
provides excellent soft-tissue discrimination, it presents an effective
means of establishing the boundary between the prostate gland, for
example, and a tumor. MRI is also sensitive to temperature, making it
compatible with thermal treatments such as laser- or cryo-surgery.
Magnetic resonance therapy combines
these advantages with real-time feedback. Its operating principle,
however, remains the same as conventional MR imaging: A magnetic field
excites the protons within the targeted tissues to their resonant
frequency. Gradient fields apply a time-variant factor, tuning the
frequencies for spatial differentiation.
Radio-frequency signals, transmitted
and received by a third set of coils, complete the process. Decay signals
from the differently resonating protons create the contrasts that make up
the final image.
"It would have been impossible to achieve a comfortable working gap at 4K," Vosburgh claims. Maintaining such a low temperature requires a liquid helium bath between the magnet's superconducting coils and cryostat walls. Accommodating this type of cooling system, he explains, would significantly shrink the gap.
Fortunately for the team of 40-plus scientists and engineers involved in the project, the slightly higher transition temperature of niobium-tin falls just within the practical limits of mechanical refrigeration. Use of a cryocooler, compared to liquid helium, reduces space requirements allowing a larger opening.
The open magnet prototype, constructed, assembled, and tested in Schenectady, weighs 3,600 kg and generates a field strength of 0.5 Tesla. The first of its kind, it features an outer diameter of 1.84m, and measures 1.48m long.
Distortion-free imaging. Another challenge to building the prototype scanner involved the gradient and radio-frequency (RF) coils. Operating in tandem, these coils provide spatial resolution of different tissue from any view direction. Both sets of coils typically surround the patient, preventing access in the same manner as the main magnet of conventional MR imaging systems.
Computer analysis of various coil designs produced a shielded gradient coil with a large central gap for each of the three principal directions. These coils permit access to the patient within the magnet bore with little reduction in gradient strength, linearity over the field of view, or eddy current performance. A surgical drape, placed directly on the patient over the region to be imaged, contains the RF coils.
"While it was necessary to make major adjustments to the superconducting coil position within the cryostat, and to redesign all three gradient coils as well as the RF coils," Vosburgh reports, "the open system possesses imaging capabilities comparable to those of conventional 0.5T MR imagers."
Heart stress monitoring
Point and image. Besides giving the surgeon direct access to the patient, the open MR scanner allows near real-time control of the MR imaging process. Light-emitting diodes, built into the probe tip, enable the physician to "set up" a surgical procedure directly from within the magnet. Good for "outside the body" position orientation, the system works as follows:
The physician points the probe--a biopsy needle, catheter, or endoscope--in the direction of the patient's tissue target. Video cameras transmit a visual reference of the LED's position to a workstation that calculates probe-tip coordinates and a unit vector along the probe's projected track. An MR image console combines this data with the MR image itself to present a picture of the plane containing the needle's projected path.
If probe orientation is not satisfactory, the physician redirects the probe, obtaining additional images until determining the correct approach. Images update every 1.5 seconds, causing only minimal delays in the procedure.
Once the image viewed on the monitor confirms an appropriate probe orientation, the physician begins advancing the probe along its projected path. A second device-tracking technique, based on MR data, monitors progress via repetitively updated images. Intended for precise positioning "inside the body," this arrangement relies on pulsed gradient fields, generated by one or more small receiver coils. Located on the probe, and measuring less than 1 mm across, the coils detect MR signals from adjacent tissues only, displaying position as a graphic symbol superimposed on the conventional MR image.
'Open' magnet, gradient, and RF
Niobium-tin superconducting tape
Video and MR probe-tracking
Clinical evaluation. Two prototype systems are in the field: One resides at Brigham and Women's Hospital, the other at University of Zurich in Switzerland. Several more systems are scheduled for delivery this year. Input from different institutions employing many physicians--from neurosurgeons to prostate surgeons--will aid in further defining the role of magnetic resonance therapy.
Continued evaluation will also answer technical questions. For example, Vosburgh believes his team could push the technology to 1.0T from the current 0.5T, achieving a better quality MR image. This, however, means shrinking the gap. "What's the trade-off?" he asks.
For now, the operating systems have been used primarily to perform biopsies on tumors in various parts of the body. Reviews have been positive.
"Even with minimally invasive procedures, physicians today are limited in that they can only view the surface of the area on which they are working," says Brigham and Women's Jolesz. "By providing 3-D images of anatomy, MRT will allow physicians to see the areas they are operating on from all angles. In essence, it will help provide a road map that shows the best surgical route."
The hope is that such road maps will soon lead to more effective cancer treatment, and ultimately, its eventual cure.