Evaluation of MR Image Overlay for Spinal Interventions

Introduction: MR guidance is being used for a host of spine injections and pain management procedures such as discography, periradicular nerve blocks, sympathetic blocks, celiac plexus blocks, sacroiliac joint injections, and facet (zygoapophysial) joint neurotomies. MR imaging provides exceptional depiction of the patho-anatomy in the vertebral column and intervertebral discs and also provides superb visualization of spinal nerves and exiting nerves structures. As a single modality, MR facilitates diagnosis and treatment of degenerative disease of the spine. Disc access is used for diagnostic (discography or aspiration for suspected infection) and therapeutic purposes. Vertebral access is important for biopsy, tumor ablation and augmentation. MR imaging is very sensitive in detecting these tumors, and it is conceivable that MR-guidance could play a major role in the focal ablative treatment of bone involvement from these entities. Imaging is used in five separate and distinct ways: planning, targeting, monitoring, controlling, and assessing treatment response [1]. In all of these cases, placing a needle at a precise location in the body as prescribed on the anatomical images is paramount. We have developed a 2D augmented reality image overlay device to guide needle insertion, thereby making diagnostic high-field magnets available for interventions without a complex engineering entourage. The device consists of an LCD screen that projects an MR image onto a semitransparent mirror, which appears to be floating inside the patient with correct size and position. This optically stable image provides the physician with two-dimensional tomographic vision to guide needle placement procedures from any view point. We present the use of this image overlay system for spinal interventions and an independent validation of the needle placement accuracy by using electromagnetic tracking in a functionally equivalent configuration of the device in the laboratory. The validation system provides a more significant measure of accuracy than that which we can obtain from MR images taken after each insertion attempt, due to paramagnetic artifact caused by the needle and lack of distinct targets; further, cost of scanner time makes large experiments with statistically significant data infeasible.