Manipulator assisted endoscope guidance in functional endoscopic sinus surgery: Proof of concept

Background: Functional endoscopic sinus surgery (FESS) is characterized by single-handed preparation and guidance of the endoscope by the nondominant hand. This results in an additional extension of operation time by up to 15% and ergonomic deficits. The aim of this study is the conception of an automated assistance system for FESS in view of the following questions: (1) Which degree of surgical automation is suitable for FESS? (2) Which design is suitable? (3) What are the properties of the technical system (planning, time, accuracy, precision) of the selected system? (4) Does the system offer potential for a clinical application? Methods: In all 49 FESS were analyzed for surgical workflows. Measurement of the maximum forces within FESS was performed with 40 trials on an anatomical model. Three different mechanical systems were used in ten FESS and evaluated using the ICCAS Human-Machine Evaluation Scale. For realization of automated endoscope guidance an engine-driven and -braked manipulator (PA10-6c, Mitsubishi, Japan) was used. The technical parameters determined were expenditure of time for the preoperative planning of workspace, surgical accuracy and precision of the intraoperative endoscope positioning, maximal forces, and time. Results: Concept-conditioned instrument changes amount to an average of 41.1 and 18.9% (5.21 min) time requirement for each FESS side. Maximum forces on the mucous membrane during a conventional FESS were measured at 9.8 N (5.9-9.8). Usability of the mechanical endoscope holder was estimated in 18 of 20 cases to be inferior to the standard procedure. The time needed for segmenting the intranasal workspace was 15.2 min (10.0-23.0). The maximum deviation of the automatically driven endoscope from a planned position amounted to 0.85 mm (manually 4.64 mm). The maximum force was measured with 1.1 N in the z direction (manually 9.8 N). Automated guidance of the endoscope to an intranasal position needed 7.25 s (6.4-7.9); manually 12.64 s (5.9-43.0). Conclusion: Guidance of the endoscope for FESS by an automated motor-driven system is possible. The conception which is based on workflow analysis favors a system with automatic definition of the workspace and a manual movement of the endoscope. The examined system offers a potential for clinical application. Definition of the automation level and development of a man-machine interface is more important than selection or reconstruction of a special manipulator for endoscope guidance in FESS from a surgical point of view.