Computer Modeling of Skull Base Endoscopic Instrument Motion and Surgical Pathway Analysis

Background: Navigation-guided endoscopic surgery of the skull base has received increasing interest due to the excellent outcomes that can be achieved in appropriate cases. As this field continues to develop and refine, new surgical pathways with reduced, less traumatic volumes are being investigated. Critical to this process is defining the boundaries of current and potential approaches, and analysis of the surgeon's use of these pathways to determine optimal dimensions for target approach and manipulation.

The goals of this project were to develop a method of automated 3 dimensional endoscopic pathway tracing, to analyze instrument motion through the pathway, and based on this information to determine whether further pathway optimization could be achieved. For initial evaluation, endoscopic transsphenoidal pituitary surgery was investigated.

Methods: After developing a measurement and analysis protocol, cadaver head CT scans were loaded into the surgical navigation system (Stryker Navigation System II, Flower Mound, TX). Navigation was registered, and the surgical approach was recorded with navigation software (iNtellect Navigation software, Stryker, Kalamazoo, MI) using annotation points. Surfaces of the naris and internal valve were annotated while ethmoidectomy and sphenoidotomy were performed to allow access to the sella. Pathway boundaries were recorded circumferentially at eight depths along the route. The procedures were performed by an experienced skull base surgeon on one side, and a junior resident on the other.

Following data collection, image segmentation software (SegView, Seattle, WA) was used to create a 3D surface model of the cadaver head and the coordinates of the annotation points were overlaid onto the model as a point cloud in a computer-aided design tool (Form-Z, AutoDesSys, Columbus, OH). Meshed boundaries were made from the point cloud representing the boundary of the surgical path.

Results: Annotated points were used to reconstruct the pathway boundaries and plot them on a 3D segmented volume for qualitative analysis (Figs. 1 and 2). The volumes of motion tapered on approach to the internal nasal valve, with increasing range of motion beyond the nasal valve followed by more constant volume along the distal approach. The data mapping initial anatomic dimensions was similar for experienced and novice surgeons. However, there was increased deviation in the distal surgical pathway where the course of the pathway was chosen by the surgeon and not limited by strict narrow anatomic confines; the experienced surgeon's instruments traveled through a more direct, lower volume course.

Conclusions: This study demonstrates a novel approach for mapping and analysis of pathways as illustrated for endoscopic pituitary surgery. The data correlated well with anatomic measurement, confirming validity. The pathway construct effectively illustrates the motion of surgical instruments relative to anatomic structures. Qualitative analysis of pathway geometry and instrument range of motion was performed, confirming expected differences in outcomes with surgical experience. Based on these results, we have begun analysis in live surgery. We expect this to enhance operative planning and access while minimizing tissue damage. Potential future uses include evaluation of surgical safety, competency and efficiency.

Fig. 1

Fig. 2