| Computerized tomography (CT) has become the preferred method for localizing hepatic metastases, especially those which are small and situated deeply within the liver. Given its ability to discriminate between tissue and tumor structures, it would be beneficial to actively use CT images during surgery; however, current surgical methods do not utilize tomographic imaging for guidance in either hepatic resection or ablative procedures. An interactive, image-guided surgical system (IIGS) for use in liver surgery has been developed to display current surgical position and anatomical structures on preoperatively obtained tomographic images, laparoscopic video images, and liver surface renderings. For localization of surgical position, instruments needed for both open and minimally-invasive hepatic resection and ablation procedures were modified in order to track them in physical space using an active optical position sensor. Several registration techniques for use in IIG liver surgery have been investigated and analyzed, including a surface registration to map physical space into tomographic image space and the direct linear transformation (DLT) to project information indicating the location of deep seated tumors onto video images acquired with a laparoscope. In cases where a minimum number of control points are used to calculate the DLT, the transformation is often unstable and the mathematical process can lead to high target registration errors. Mathematical filters developed through the analysis of the DLT parameters were used to eliminate most outlier matrices which produced high target errors. The software for these components of the system, along with methods to display various types of medical images, was developed in Visual C++ 6.0 and bundled as dynamic link libraries in a PC-based image guided surgical system. The utility of the system is demonstrated on experiments conducted with liver phantoms. |
