Proctor Observations for Madigan Endoscopic Sinus Surgery Simulator v.2.0

Proctor Observations

Haptics

Overall, there was tremendous improvement in the haptics in this version of the simulator. During the procedures, the subjects who had used the previous version of the simulator noticed a dramatic improvement during Injection and initial Dissection.

Collision between the instrumented endoscope and forceps was still often confused for haptic feedback from the system. During many procedures, the subjects would mistake this interaction to be the haptic feedback hindering them from entering an open area with the forceps. Therefore, falsely indicating to them that the haptics were failing during the latter part of dissection in the Anterior and Posterior Ethmoid Cells. Because of this interaction, many of the residents who were subjects in the previous version of the simulator did not see any improvement in the haptics and one of the staff considered the haptics to be more of a hindrance than a mechanism to better simulate a real procedure. Some subjects preferred to have the haptics inactivated during dissection of the Ethmoid Cells, allowing them greater, but unrealistic, freedom of movement with the forceps and endoscope. When the haptics were inactivated, the subjects were allowed to penetrate the virtual anatomy with the tip of the instrumented forceps. This allowed them the clearance from the encoders on the instrumented forceps, which was needed to obtain a normal view of the virtual anatomy with the endoscope.

It was unanimous that the Ethmoid Cells were better represented graphically in this version, but further testing was needed to create a more realistic haptic representation. It was also suggested, by many of the subjects, that there be 6 Degrees of Freedom (DOF) haptics placed on both the forceps and the endoscope and that the algorithms for the haptics include distinctions for bone vs. cartilage.

Voice Recognition

A number of subjects had incredible difficulty using the voice recognition system to initiate the commands needed during the procedure.

Familiarity with the Virtual Anatomy

Subject's familiarity with the virtual anatomy played a major roll in recognition of active areas for dissection. The Navigation and Injection sub tasks were not affected by the subject's familiarity. Active areas of dissection for the graphical model of the anatomy needed to be shown to most subjects during their trials on the simulator. Most of the subjects had preexisting knowledge of where they needed to dissect on real anatomy, and would attempt to dissect those areas on the simulator, but, restrained by the ability of the Infinite Reality, not all anatomy could be made dissectable. The primary anatomical structures which were difficult to distinguish were the Lamina Papyrecea and the Ethmoid Cells. Although, overall, the subjects thought that the new anatomical models were a more realistic representation of the actual anatomy, there was still a need for more distinction between active areas and inactive areas. Distinguishing between these areas (say, with differences in color or shading) would introduce another unrealistic aspect of the simulator and would, therefore, decrease the immersion of the students during training. Waiting for increased graphical rendering capabilities of workstations, allowing the entire anatomical model to become dissectable (as in real OR cases), would be a more viable solution.

Most of the subjects complained of too short of a viewing frustum for the endoscope. This could have contributed to their need for the proctor's help in identifying the active areas of dissection. This is a simple change in the simulator, which could be implemented for future training sessions.

Ability to "Push Through" the Virtual Anatomy During Trials

The shaft of the forceps and the entire instrumented endoscope were not fitted with 6 DOF haptic feedback, because of the cost of such a system, which would inhibit their penetration of the virtual anatomy. Because of this, the subjects were able to unrealistically pass the viewpoint of the endoscope, the shaft of the endoscope, and the shafts of the virtual tools through the virtual anatomy. In the previous version of the simulator, this was seen, by most, as beneficial, forcing the residents to exercise greater hand eye coordination in maneuvering the endoscope and tools within the virtual anatomy. Because of the significant improvements made in the graphical model, haptics and tools in this version of the simulator, the benefit of exercising greater hand eye coordination was not weighted as heavily as the possibility of the simulator obtaining full operating room realism. The majority of the repeat subject's comments were concentrated on how each aspect of the simulator differed from the real operating room rather than what limited operating room skills residents could gain from using this simulator. This would indicate that the subjects were more directly correlating the simulated procedures to the actual OR procedures.



last updated: 10/17/98