Research  >  Labs & groups  >  X-ray imaging  >  Projects  >  XMR: Fusion of MRI and X-ray modalities

XMR: Fusion of MRI and X-ray modalities

John Bracken, Philip Komljenovic, David Green, Giovanni DeCrescenzo

X-ray fluoroscopy and angiography are essential modalities for visualization of surgical tools and vasculature in real time across a large field of view for minimally invasive percutaneous interventions used to treat patients with heart or vascular disease. However, some interventions, such as those used in the treatment of peripheral chronic total occlusions (CTOs) and percutaneous aortic valve replacement, are still difficult with the X-ray modality alone because soft-tissue contrast is lacking in the X-ray images. Magnetic resonance imaging (MRI) can provide complementary soft tissue contrast information and details about device position relative to the anatomy being treated.

Therefore, we are developing a hybrid X-ray/MRI system (XMR) to improve the safety and efficacy of these difficult procedures by harnessing the complementary strengths of both modalities. The XMR will be composed of an X-ray C-arm positioned close to (about 1 m) the entrance of a 1.5 T MRI scanner. The close proximity of both modalities will facilitate safe, smooth and rapid patient transfer between each modality while simultaneously providing a sufficiently large workspace for the clinical team to work effectively at the MRI scanner entrance. The XMR system is the principal component of the Imaging Research Centre for Cardiac Intervention (IRCCI), the clinical research centre that has been established at Sunnybrook Health Sciences Centre.

Figure 1: Propo sed hybrid X-ray/MRI (XMR) system


Several technical challenges are present in the construction of the XMR system. First, the X-ray C-arm and its components could affect the magnetic field homogeneity of the MRI scanner. Second, the magnetic fringe field present near the entrance of the MRI scanner could affect the performance of the X-ray tube and the flat-panel detector. Previous work1 has shown that a properly modified flat-panel detector will function effectively in the magnetic fringe field. We have also found that the field homogeneity of the MRI scanner is only slightly affected by the presence of the X-ray system and that field nonhomogeneity can easily be corrected with the shimming coils of the MRI scanner. Therefore, we have been addressing the remaining challenge: ensuring effective X-ray tube performance in the fringe field.

We have found that the induction motor of the X-ray tube used to rotate the anode in which X-rays are generated performs effectively in the weak fringe field (about 5–20 mT) that is present about 1 m from the scanner entrance. The reduction of anode rotation frequency in this fringe field is negligible, which means that X-ray tube heat loadability will not be affected by the fringe field.

Figure 2: Anode rotation frequency with the X-ray tube placed in the MRI magnetic fringe field. At a field of 5–20 mT, anode rotation frequency reduction was negligible, so X-ray tube induction motor performance was unaffected.2

We have also investigated the deflection of the electron beam in the X-ray tube caused by the fringe field. This deflection causes the focal spot on the X-ray tube anode to shift, resulting in a change in the size and position of the field of view. An increase in focal spot blurring was also observed. To correct this electron beam deflection, we have designed an active shielding system that determines the strength of the fringe field acting on the electron beam. The shielding system then produces a counter magnetic field to deflect the electron beam back to its original trajectory when no fringe field is present. This negates the shift of the focal spot on the anode.

Figure 3: a) Image of focal spot with no fringe field present on a flat-panel detector. b) Focal spot image deflection in a fringe field of 5 mT. The crosshairs indicate the location of the centre of the focal spot with no fringe field present. c) Focal spot image in a fringe field of 5 mT with active shielding. The shielding system was able to deflect the focal spot back to its original location on the anode.

Currently, we are constructing a complete prototype clinical XMR system for in vivo percutaneous interventions in animal models and ultimately patients.


  1. L. Brzozowski, A. Ganguly, M. Pop, Z. Wen, R. Bennett, R. Fahrig and J. A. Rowlands, “Compatibility of interventional X-ray and magnetic resonance imaging: Feasibility of a closed bore XMR (CBXMR) system,” Med. Phys. 33, 3033-3045 (2006).
  2. Bracken J., Lillaney P., Fahrig R., Rowlands J. A., “Hybrid X-ray/magnetic resonance imaging for percutaneous aortic valve replacement,” RSNA (Chicago, IL), 2007, paper SSQ16-01.
  3. J. Bracken and J. A. Rowlands, “X-ray tube induction motor performance in a 1.5 T MRI fringe field,” Med. Phys. 33, 2667-2667 (2006), AAPM 2006 conference abstract.
  4. R. Fahrig, A. Ganguly, P. Lillaney, J. Bracken, J. A. Rowlands, Z. Wen, H. Yu, V. Rieke, J. M. Santos, K. B. Pauly, D. Y. Sze, J. K. Frisoli, B. L. Daniel and N. J. Pelc, “Design, performance and applications of a hybrid X-ray/MR system for interventional guidance,” Proc. IEEE 96, 468-480 (2008).