MR-guided cholecystostomy: assessment of biplanar, real-time needle tracking in three pigs.

PURPOSE: To demonstrate the feasibility of magnetic resonance (MR)-guided cholecystostomy using active, real-time, biplanar MR tracking in animal experiments. METHODS: Experiments were performed on three fully anesthetized pigs in an interventional MR system (GE open). The gallbladder was displayed in two orthogonal planes using a heavily T2-weighted fast spin-echo sequence. These "cholangio roadmaps" were displayed on LCD monitors positioned in front of the interventionalist. A special coaxial MR-tracking needle, equipped with a small receive-only coil at its tip, was inserted percutaneously into the gallbladder under continuous, biplanar MR guidance. The MR-tracking sequence allowed sampling of the coil (needle tip) position every 120 msec. The position of the coil was projected onto the two orthogonal "cholangio roadmap" images. RESULTS: Successful insertion of the needle was confirmed by aspiration of bile from the gallbladder. The process of aspiration and subsequent instillation of Gd-DTPA into the gallbladder was documented with fast gradient-recalled echo imaging. CONCLUSION: Biplanar, active, real-time MR tracking in combination with "cholangio roadmaps" allows for cholecystostomies in an interventional MRI environment.

Biopsy needle susceptibility artifacts.

Understanding the appearance of thin metallic structures in magnetic resonance imaging is important for evaluating the potential role of MRI in guiding and monitoring percutaneous interventions. As most MR compatible instruments are made from materials with a susceptibility different from water, their visibility is enhanced beyond what is expected on the grounds of displaced water alone. Unfortunately, this artifactually enhanced visibility is not constant, but instead depends on a variety of factors. This article presents computer simulations of the image distortion resulting from magnetic susceptibility differences between a needle and the surrounding tissue. The simulations show not only an artifact size that is dependent on needle composition, orientation, and pulse sequence, but also a corresponding shift of the artifact center away from the actual center of the needle. These effects place limits on the accuracy of MRI-guided needle tip placement.

Human aorta: preliminary results with virtual endoscopy based on three-dimensional MR imaging data sets.

PURPOSE: To determine the feasibility of use of magnetic resonance (MR) imaging data sets to perform virtual intraarterial endoscopy (VIE). MATERIALS AND METHODS: Seven female and 14 male patients (aged 9-86 years [mean, 42 years]) with various pathologic aortic conditions underwent three-dimensional gadolinium-enhanced spoiled gradient-echo MR imaging. With prototype software, VIE postprocessing algorithms (based on ray casting) were applied to the imaging data sets. Findings at conventional angiography were used as the standards of reference. RESULTS: The vessel wall was seen from the inside in each case, and the following pathologic conditions were depicted clearly: the two lumina in a congenial double aortic arch and the single lumen after correction, vessel narrowing in coarctations, characteristics of Leriche syndrome, stenoses, occlusions, and the true and false lumina of dissections. CONCLUSIONS: Limitations of VIE include the image quality of the original data set, the threshold chosen to minimize intraluminal artifacts, and the inherent smoothing of vessel walls.