Computational models in image guided interventions

Radiofrequency ablation of lung tumors in swine assisted by a navigation device with preprocedural volumetric planning

PURPOSE

To develop an image guidance system that incorporates volumetric planning of spherical ablations and electromagnetic tracking of radiofrequency (RF) electrodes during insertion.

MATERIALS AND METHODS

Simulated tumors were created in three live swine by percutaneously injecting agar nodules into the lung. A treatment plan was devised for each tumor with optimization software to solve the planning problem. The desired output was the minimum number of overlapping ablation spheres necessary to ablate each tumor and the margin. The insertion plan was executed with use of the electromagnetic tracking system that guided the insertion of the probe into precomputed locations. After a 72-hour survival period, animals were killed and histopathologic sections of the tissue were examined for cell viability and burn pattern analysis.

RESULTS

A planning algorithm to spherically cover the tumors and the margin was computed. Electromagnetic tracking allowed successful insertion of the instrument, and impedance roll-off was reached in all ablations. Depending on their size, the tumors and the tumor margins were successfully covered with two to four ablation spheres. The image registration error was 1.0 mm +/- 0.64. The overall error of probe insertion was 9.4 mm +/- 3.0 (N = 8). Analysis of histopathologic sections confirmed successful ablations of the tissue.

CONCLUSIONS

Computer-assisted RF ablation planning and electromagnetically tracked probe insertion were successful in three swine, validating the feasibility of electromagnetic tracking-assisted tumor targeting. Image misregistration caused by respiratory motion and tissue deformation contributed to the overall error of probe insertion.

Normal vector information registration and comparisons with mutual information

We propose a new similarity measure for medical image registration and make comparisons with two kinds of mutual information (MI) methods in the literature. This new measure is computed from the normal vector information (NVI) of the images instead of pixel intensity that the MI methods based. The NVI of an image is extracted from the relationship between pixels and computed from the normal vector of the pixel based on their local isosurface. The NVI method has been proved to be able to successfully register two-dimensional (2D) images. In this paper, we will apply it in three-dimensional (3D) medical images and employ the known-result datasets to quantitatively evaluate the performances of the NVI and MI methods. The visual assessment is employed for the unknown-result clinical image dataset registrations. The results show that the NVI method is ready to be affected by the salt-and-pepper noise; while when the random noise is removed, the NVI method performs no worse than the MI methods.