KIT | KIT-Bibliothek | Impressum | Datenschutz

Analyzing Natural Language Commands during Laparoscopic Surgery for Robotic Camera Guidance

Scheikl, Paul Maria; Nigl, Beija; Kunz, Christian; Mathis-Ullrich, Franziska

Abstract (englisch):
In this work, we investigate the use of natural language commands for adjusting camera positions during laparoscopic surgery in order to equip a robotic assistance systems with an intuitive voice interface. The integration of natural language commands enables the surgeon to flexibly adapt the camera’s position and to provide feedback to a learning assistance system. Based on the transcripts from 20 laparoscopic surgeries the verbal commands were clustered to receive a grounded understanding of the different command types used. These insights then serve as input for supervised classification methods that map natural language commands to actions of the robotic assistance system that guides the camera. The preliminary results show the need for more complex voice command structures than those presently integrated in existing assistance systems.

Open Access Logo

Zugehörige Institution(en) am KIT Institut für Prozessrechentechnik, Automation und Robotik (IPR)
Institut für Anthropomatik und Robotik (IAR)
Publikationstyp Proceedingsbeitrag
Jahr 2019
Sprache Englisch
Identifikator ISBN: 978-3-00-063717-9
KITopen-ID: 1000100501
Erschienen in 18. Jahrestagung der Deutschen Gesellschaft für Computer- und Roboterassistierte Chirurgie e.V. - Tagungsband. Reutlingen, 19 - 21 September 2019. Hrsg.: O. Burgert
Veranstaltung 18th Jahrestagung der Deutschen Gesellschaft für Computer- und Roboterassistierte Chirurgie e.V. (2019), Reutlingen, Deutschland, 19.09.2019 – 21.09.2019
Verlag Hochschule Reutlingen, Reutlingen
Seiten 29-31
Externe Relationen Abstract/Volltext
Schlagworte Laparoscopic Surgery, Robot Assisted Surgery, Natural Language Processing
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page