Login / Signup

Improvement of the insertion axis for cochlear implantation with a robot-based system.

Renato TorresGuillaume KazmitcheffDaniele De SetaEvelyne FerraryOlivier SterkersYann Nguyen
Published in: European archives of oto-rhino-laryngology : official journal of the European Federation of Oto-Rhino-Laryngological Societies (EUFOS) : affiliated with the German Society for Oto-Rhino-Laryngology - Head and Neck Surgery (2016)
It has previously reported that alignment of the insertion axis along the basal turn of the cochlea was depending on surgeon' experience. In this experimental study, we assessed technological assistances, such as navigation or a robot-based system, to improve the insertion axis during cochlear implantation. A preoperative cone beam CT and a mastoidectomy with a posterior tympanotomy were performed on four temporal bones. The optimal insertion axis was defined as the closest axis to the scala tympani centerline avoiding the facial nerve. A neuronavigation system, a robot assistance prototype, and software allowing a semi-automated alignment of the robot were used to align an insertion tool with an optimal insertion axis. Four procedures were performed and repeated three times in each temporal bone: manual, manual navigation-assisted, robot-based navigation-assisted, and robot-based semi-automated. The angle between the optimal and the insertion tool axis was measured in the four procedures. The error was 8.3° ± 2.82° for the manual procedure (n = 24), 8.6° ± 2.83° for the manual navigation-assisted procedure (n = 24), 5.4° ± 3.91° for the robot-based navigation-assisted procedure (n = 24), and 3.4° ± 1.56° for the robot-based semi-automated procedure (n = 12). A higher accuracy was observed with the semi-automated robot-based technique than manual and manual navigation-assisted (p < 0.01). Combination of a navigation system and a manual insertion does not improve the alignment accuracy due to the lack of friendly user interface. On the contrary, a semi-automated robot-based system reduces both the error and the variability of the alignment with a defined optimal axis.
Keyphrases
  • machine learning
  • deep learning
  • high throughput
  • minimally invasive
  • computed tomography
  • patients undergoing
  • magnetic resonance
  • soft tissue