Magnetic resonance imaging is known as an excellent tool for noninvasive diagnosis. With the advent of novel fast imaging technologies and open magnets, MRI is now becoming an imaging tool for guiding various interventional procedures. Excellent soft tissue contrast and the ability to visualize lesions in high resolution are the main reasons for this usage. Currently, various intervention and biopsy procedures on multiple organs including the breast, liver, prostate, and brain as well as the heart are being investigated. Guidance of vascular interventions using MRI is technically very challenging, but some investigational studies have already been completed. Complicated interventions on the spine currently are being guided by magnetic resonance imaging.

Technical Challenges

Jean-Michel Serfaty, MD, a postdoctoral fellow at Johns Hopkins University at the time of the study, manipulates guidewires and catheters under MR guidance to conduct an experimental coronary balloon angioplasty procedure on an animal. Serfaty is now an interventional radiologist at Bichat Hospital, Paris, France.

There are multiple technical challenges in guidance interventions under MRI. First, access to the patient in a conventional 50 to 60 cm-diameter bore-size MRI scanner is rather difficult. Three alternative strategies are being investigated:

1) Magnets with open architecture are an excellent solution to the access problem, but one that usually comes with a decrease in field strength and homogeneity and therefore compromised image quality.  Short magnets solve the access problems in certain procedures such as vascular intervention where access is necessary only to the groin of the patient. As a result of competition between the scanner manufacturers, magnet designs are changing every year and giving more access options to clinicians.

2) Taking the patient partially out of the magnet during intervention and carrying on the imaging operation only for verification of the device position is a simple and very effective solution for many interventions where real-time image guidance is not a must. In these types of operations, some MR-visible markers are placed on the patient and location of the intervention is estimated based on the relative position of these markers with respect to target location.

3) Another very attractive solution to the access problem is to use robots inside the MRI scanner. In this approach, while a robot manipulates the surgical tool such as a needle, MR images are continuously acquired to give feedback control to the robot on the position of the surgical device. The design of the robot is technically very challenging: it should not contain a significant amount of ferromagnetic materials. Although plastics are MRI compatible, their mechanical stability creates engineering challenges. Furthermore, the space limitation causes significant problems in the design of the robots. Sense of touch is critical in most interventional procedures and robots that have haptic feedback capabilities are being developed. Existence of these strategies relieves the patient access problem of interventional MRI. Access to patients in the MRI scanner during an interventional procedure is no longer a serious problem.

Complete MRI-guided intervention in a circumflex artery. a) Placement of the MRI-guiding catheter (arrowhead) in the ascending aorta using the oblique sagittal view. b) Catheterization with the MRI-guiding catheter (arrowhead) of the left main coronary artery and circumflex artery using the oblique coronal view. c) Real-time projection angiography of the circumflex artery (arrowhead) on an oblique coronal view after injection of diluted gadolinium (31 mM) in the MRI-guiding catheter. d) Placement of the MRI guidewire (arrowhead) in the circumflex artery in the oblique coronal view. The balloon angioplasty catheter can be localized and advanced on the MRI guidewire by using a black artifact created by a platinum ring localized in the center of the balloon angioplasty catheter (long arrow). e) Injection of diluted gadolinium-DTPA (31 mM) in the balloon enhances the balloon on the real-time projection angiography images (long arrow). Oblique coronal view. (From: J.M.Serfaty, Yang X, Foo T, Kumar A, Derbyshire A, Atalar E. MRI-guided coronary catheterization and PTCA: A feasibility study on a dog model. Magn Reson Med 2003;49(2):258-263. Reprinted by permission of Wiley-Liss, Inc, a subsidiary of John Wiley & Sons, Inc.)

The second technical challenge is the requirement that the equipment used in interventional procedures is MRI compatible. The high magnetic field of the scanner attracts objects made out of ferromagnetic materials, and the attraction force applied to these objects varies based on the composition of the materials and the shape of the object. It is not uncommon to have accidents as a result of flying objects. Although the solution is straightforward, many existing devices, such as physiological monitoring equipment, are incompatible with MRI. Although there are companies that offer MR-compatible equipment, their high cost related to their low sale volume becomes a significant issue. Some suggest that leaving the MR-incompatible equipment outside the five-gauss line (outside this region, the magnetic field is negligibly low) may be sufficient to solve this problem. In this case, staff training will be a very critical issue. It should be noted that some of these MR-incompatible devices generate high electrical noise that disturbs the MR image quality. This is unacceptable.

Another major technical problem is the visibility of interventional tools, such as catheters, guidewires, and needles. Under x-ray, the devices are visible in high resolution, but soft tissue visualization is very poor. Under MRI guidance, the scenario is just the opposite: soft tissue visualization is excellent, but it is difficult to determine the position of the interventional tools. Many alternatives are being investigated, but the most common and widely used technique is to employ devices that are made out of materials that cause slight image distortion. Although the location of the devices may not appear as precise as they do under x-ray guidance, in many instances, they create enough artifact to do the procedure. An alternative to this technique is coating the catheter with a material that reduces the relaxation constant, or T1, of the surrounding tissue and therefore the catheter appears bright. Both of these approaches are called passive visualization. Active visualization of the interventional devices is achieved by placing a radiofrequency (RF) antenna on the device, which picks up a very high signal around the interventional tool. With this technique, catheter visualization is much easier, but the technique has been criticized for possibly causing excessive heat around the antenna. Researchers have developed designs that minimize this potential risk. While research on interventional device visualization continues, today viable solutions for this problem do exist.

Current uses

Most installed MRI scanners are equipped with software suitable for diagnostic procedures. In the guidance of interventions, real-time image updates are critical. Specialized pulse sequences are necessary to accurately and quickly find the position of the interventional devices while the target is visualized from a monitor inside the MR scanner room in real time. While active research is currently under way, MR scanner manufacturers are now delivering software packages for this purpose.

With the advent of this technology, many interventional procedures became possible and are currently under development in the hands of clinical researchers. Some interventional procedures require further technological development. Here we will touch on a few procedures in clinical practice.

Although MR-guided breast biopsy is not in routine clinical use, the effectiveness of this technique has been shown with multiple studies. Magnetic resonance imaging has very high sensitivity but, unfortunately, its specificity is rather low. If a biopsy sample can be taken from the suspected lesion, the accuracy of the diagnosis will improve significantly. The MR-guided breast biopsy procedure, which is designed to address this issue, is typically applied to  patients who have suspected lesions and negative conventional biopsy results. In this procedure, the breast is gently compressed between two compression plates, one of which has a grid structure. Contrast-enhanced magnetic resonance images depict the position of the breast lesion very accurately. The physician measures the position of the lesion with respect to the grid with the aid of specialized software and calculates needle insertion position and needle insertion depth. After placement of the needle, its position is verified with a quick scan, and the needle is readjusted if necessary and the sample is taken. The procedure typically takes about 30 to 40 minutes. Both the software and hardware that are required to conduct these procedures are commercially available from multiple vendors. The widespread use of this technique is limited by its high cost.

With a similar technique, MR-guided biopsy of the prostate is possible. Biopsy of the prostate currently is carried out under ultrasound guidance. Since ultrasound is practically blind to prostate tumors, six or more biopsy samples are taken from the prostate without targeting any lesion. This untargeted biopsy is known to cause significant false-negative findings and, therefore, significant delays in cancer treatment. It has been shown that MR-guided biopsy can be carried out in a open magnet as well as in a conventional MR scanner with the aid of specialized hardware.

One of the big challenges in brain tumor removal is to define precise boundaries of the tumor and complete the resection of the tumor while sparing normal brain tissue. While visual inspection in brain surgery is the common technique, it has been shown that many tumors are not removed completely because of the difficulties in visualization. Although MR-guided brain surgery is currently conducted in only a handful of research centers, it carries great potential.


As discussed earlier, MRI is a very versatile imaging technique. One of its capabilities is to image the amount of temperature change in the body. It has been observed that frequency of the received MRI signal depends on temperature. Although this dependence is very small (less than one hertz per degree temperature change), with very accurate imaging techniques, it is now possible to image the temperature change in the body within less than one degree accuracy. This capability of MRI creates an opportunity for ablation of tumors with high precision. Various techniques have been developed including ablation with the aid of RF ablation needles. A recently developed alternative is the use of focused ultrasound to ablate lesions that are deep inside the body without a single incision. It has been known for a long time that ultrasonic energy could be focused in the body with the aid of a large transducer. However, because of the significant variations of the speed of sound in the body, the focus location was not known precisely. Imaging of the temperature variation in the body creates a unique opportunity to solve this technical problem. While imaging the tissue using the temperature-imaging technique described earlier, a small energy is applied to the transducer such that it does not cause any damage to the tissue but raises the temperature at the focused region enabling visualization under MRI of the extent and location of the focused region. This enables readjustment of the focus for precise ablation of the tumor of interest without damaging surrounding tissue.

Cryoablation is an alternative to heat therapy. MR can monitor the size of the iceball formed in the body. This technique is thought to be very suitable for ablation of liver tumors.

Accurate visualization also is critical in the delivery of therapeutic agents. One very exciting development is that magnetically labeled stem cells now can be delivered to lesions under MR guidance. With this technique, researchers are investigating the possibility of curing infarcted heart tissue. MR guidance may play a critical role in solving delivery problems. Many other possibilities do exist.

As mentioned, in performing vascular interventions under x-ray guidance, the interventional devices can be visualized very accurately, but soft tissue is visible in the form of shadows. This poor visualization creates many technical difficulties for the vascular interventionalist. With the ability to solve these visualization problems, MRI is offering a serious alternative to x-ray in the guidance of some difficult vascular procedures.

Although atherosclerosis is a disease of the vessel wall, the current treatment techniques are done without seeing the diseased wall. Instead, the treatment is based on the appearance of the vessel lumen. With its ability to visualize high soft tissue contrast, MR guidance offers the possibility of characterization of atherosclerotic plaques. Although it is not yet determined how this new information can be used in the treatment of the disease, some suggest that with the accurate visualization techniques, novel therapies will emerge.

A further limitation of x-ray guidance is the inability to visualize the vessels beyond an occluded region due to the fact that vessel visualization is achieved by injecting contrast agents. Passing an occlusion, therefore, is a very difficult task, a problem that is greatly simplified with MR guidance. Similarly, the portal vein system is invisible under x-ray guidance, because there is no direct access to this system from the vena cava or any other means. Patients with advanced cirrhosis sometimes require a bypass surgery, and with the aid of MRI guidance, this difficult surgery can be done percutaneously.

Research is ongoing in MR-guided vascular interventions as well as other nonvascular procedures. I believe that MRI will become a critical tool in the guidance of many interventional procedures. For those interested in learning more about interventional MRI, the 5th Interventional MRI Symposium will be held on October 15-16, 2004, in Boston. n